WEBVTT

cue-1
00:00:00.080 --> 00:00:04.550
Any one of them is enough to find that
OpenAI breached the charitable trust. Now,

cue-2
00:00:04.550 --> 00:00:10.220
first, they enriched the investors,
Microsoft, and insiders like

cue-3
00:00:10.780 --> 00:00:15.399
Brockman and apparently quite a few more.
We know Sutskever has now seven billion

cue-4
00:00:15.400 --> 00:00:20.500
dollar interest, and who knows who else
might have other interests. Um, Professor

cue-5
00:00:20.500 --> 00:00:24.320
Schizer told you that a basic principle of
the nonprofits is their revenues and

cue-6
00:00:24.320 --> 00:00:28.560
assets have to be for the benefit of the
nonprofit. And the defendants violated

cue-7
00:00:28.560 --> 00:00:34.340
that basic principle by entering into this
twenty-twenty deal with Microsoft under

cue-8
00:00:34.340 --> 00:00:39.520
terms that were unfair to the nonprofit.
It's that simple. All right? The ten

cue-9
00:00:39.520 --> 00:00:42.840
billion dollar Microsoft investment was
ten times its initial investment.

cue-10
00:00:44.060 --> 00:00:48.780
At that point, it had invested thirteen
billion dollars, and that totally dwarfed,

cue-11
00:00:48.780 --> 00:00:55.380
totally dwarfed charitable contributions
and weakened OpenAI in its negotiating

cue-12
00:00:55.380 --> 00:01:02.320
position with, um, excuse me, with, um,
Microsoft. You remember I asked,

cue-13
00:01:02.320 --> 00:01:02.540
um,

cue-14
00:01:03.140 --> 00:01:05.580
Professor Schizer, wh-what if they raise,

cue-15
00:01:06.200 --> 00:01:06.400
you know,

cue-16
00:01:07.560 --> 00:01:07.980
the money,

cue-17
00:01:08.840 --> 00:01:11.600
if they raise the thirteen billion
dollars? But I asked him, wh-what if they

cue-18
00:01:11.600 --> 00:01:12.340
raised half of it?

cue-19
00:01:13.280 --> 00:01:16.320
And I don't have the question right in
front of me, but it was a question along

cue-20
00:01:16.320 --> 00:01:20.020
those lines. And, and the idea is you at
least have some bargaining power there. He

cue-21
00:01:20.020 --> 00:01:25.020
testified that there are charities in,
in, in this country, in fact, one right up

cue-22
00:01:25.020 --> 00:01:29.820
the road at Stanford, that can raise
three billion dollars in, in the course of

cue-23
00:01:29.820 --> 00:01:30.830
a year. He mentioned...

cue-24
00:01:38.500 --> 00:01:42.640
Um, there we go. He, he mentioned one that
was, as he put it, essentially a sou-

cue-25
00:01:42.640 --> 00:01:46.130
soup kitchen. I think it was called Feed
America. And he said that they raised over

cue-26
00:01:46.130 --> 00:01:50.120
four billion dollars a year. Now you
might, you might say, "Well, gee, you

cue-27
00:01:50.120 --> 00:01:50.250
know,

cue-28
00:01:51.140 --> 00:01:54.940
Brockman and Altman, you know, how are
they gonna raise that money?" Well, I have

cue-29
00:01:54.940 --> 00:01:58.100
to tell you, um, Stuart Russell's message

cue-30
00:01:59.080 --> 00:02:00.539
is so powerful, right?

cue-31
00:02:01.460 --> 00:02:02.560
The way that AI--

cue-32
00:02:03.440 --> 00:02:05.960
It, it does create good things. I'm not
saying that it can't create good things.

cue-33
00:02:05.960 --> 00:02:08.580
It's the way that it can have such a
negative effect

cue-34
00:02:09.340 --> 00:02:14.290
on our lives. One would think that people
would be knocking down the door to donate

cue-35
00:02:14.290 --> 00:02:20.900
to an organization that's dedicated to the
safe development as a nonprofit. So I

cue-36
00:02:20.900 --> 00:02:23.880
don't know what they did. I don't know if
they had firms or evidence that they did

cue-37
00:02:23.880 --> 00:02:30.440
that, but, um, that was a problem. And,
um, the constraints on the earlier deals,

cue-38
00:02:30.440 --> 00:02:30.920
the cap

cue-39
00:02:31.600 --> 00:02:35.460
and the limitations on the IP rights,
those, um, those were gone too.

cue-40
00:02:36.140 --> 00:02:40.900
Under the terms of the twenty-twenty-three
deal, OpenAI would need to earn roughly

cue-41
00:02:40.900 --> 00:02:44.840
two hundred and seventy-five billion
dollars in profit,

cue-42
00:02:46.040 --> 00:02:49.400
excuse me, for the nonprofit to, um,

cue-43
00:02:50.640 --> 00:02:52.100
receive... Remember, uh,

cue-44
00:02:52.720 --> 00:02:57.060
Professor Schizer used the term residuals?
There was a, a term residuals. And I- at

cue-45
00:02:57.060 --> 00:03:02.060
the time he was testifying, I, I, I looked
over and I saw some recognition of his

cue-46
00:03:02.060 --> 00:03:06.900
faces. There, there were these TRAs, uh,
these are the amounts that, that investors

cue-47
00:03:06.900 --> 00:03:12.260
would get, including employees, including
the nonprofit, including Microsoft.

cue-48
00:03:12.860 --> 00:03:17.840
And anything that was paid off, uh, beyond
these target

cue-49
00:03:18.460 --> 00:03:22.180
redemption amounts, the target amounts
that they would get, would go into the,

cue-50
00:03:22.180 --> 00:03:23.060
into the nonprofit.

cue-51
00:03:23.920 --> 00:03:26.860
But in the twenty-twenty-three agreement,

cue-52
00:03:28.540 --> 00:03:28.740
at--

cue-53
00:03:29.430 --> 00:03:33.620
when twenty-twenty-five came along, the
residents, excuse me, the residuals

cue-54
00:03:34.280 --> 00:03:38.740
would become much smaller. And we-- he
showed this because-- And, and by the way,

cue-55
00:03:38.740 --> 00:03:43.340
he, he did not represent that this was a
dollar for dol-- you know, exactly precise

cue-56
00:03:43.340 --> 00:03:45.700
calculation. He was showing you what
happens.

cue-57
00:03:46.360 --> 00:03:48.740
And, and can we just go back to the last
slide? So

cue-58
00:03:50.720 --> 00:03:53.540
it makes sense, right? That if you're, if
you're getting some amount of money, and

cue-59
00:03:53.540 --> 00:03:56.720
that would be represented by the blue dots
and well as the blue,

cue-60
00:03:57.940 --> 00:03:58.440
and then

cue-61
00:03:59.100 --> 00:04:03.920
everybody else gets more money two years
later, the residuals go up by twenty

cue-62
00:04:03.920 --> 00:04:04.400
percent,

cue-63
00:04:05.040 --> 00:04:08.060
you're gonna get a smaller amount for the
residuals. So the nonprofit is getting

cue-64
00:04:08.060 --> 00:04:08.800
less money.

cue-65
00:04:09.540 --> 00:04:14.500
And that twenty percent is extraordinary.
I asked Satya Nadella, "Twenty percent?" I

cue-66
00:04:14.500 --> 00:04:14.660
said,

cue-67
00:04:15.380 --> 00:04:17.100
"That doubles. That doubles

cue-68
00:04:18.220 --> 00:04:18.980
in less than four years."

cue-69
00:04:20.160 --> 00:04:25.380
And, and he agreed. So that was an
extraordinary term and an extraordinary

cue-70
00:04:25.380 --> 00:04:29.670
give on the part of the nonprofit when
they were negotiating this deal.

cue-71
00:04:31.300 --> 00:04:37.500
Um, Microsoft also got ri-broad rights to
all of OpenAI's intellectual property up

cue-72
00:04:37.500 --> 00:04:42.480
to the time that AGI was achieved. It got
the right under this agreement to put

cue-73
00:04:42.480 --> 00:04:44.060
twenty engineers,

cue-74
00:04:44.860 --> 00:04:44.870
um,

cue-75
00:04:45.660 --> 00:04:49.150
and eventually up to ten percent of
OpenAI's engineering population,

cue-76
00:04:49.800 --> 00:04:52.230
to help Microsoft commercialize,

cue-77
00:04:53.060 --> 00:04:53.360
um,

cue-78
00:04:54.460 --> 00:04:58.640
that intellectual property and earn
revenue. And this is the trend that you're

cue-79
00:04:58.640 --> 00:05:04.120
seeing at this point in time, right? That
they are commercializing OpenAI. It's no

cue-80
00:05:04.120 --> 00:05:04.570
longer

cue-81
00:05:05.240 --> 00:05:10.370
a research lab doing safe AI. I mean, uh,
when Sam Altman testified and he said,

cue-82
00:05:10.370 --> 00:05:11.540
"Oh, these are, you know,

cue-83
00:05:12.460 --> 00:05:14.840
researchers, and they need a safe
environment to,

cue-84
00:05:15.440 --> 00:05:18.840
um, do their work, and you can't manage
them like a business." Well,

cue-85
00:05:19.540 --> 00:05:19.760
um,

cue-86
00:05:20.660 --> 00:05:24.800
Microsoft, we'll talk about it in a couple
of minutes, was throwing out at them, you

cue-87
00:05:24.800 --> 00:05:25.200
know, "Hey,

cue-88
00:05:25.820 --> 00:05:29.140
you know, you gotta make this hundred
million dollar revenue goal if you wanna

cue-89
00:05:29.140 --> 00:05:33.480
get another investment for us." They were
pushing them for the early release of a

cue-90
00:05:33.480 --> 00:05:35.770
paid version of ChatGPT, ChatGPT.

cue-91
00:05:36.500 --> 00:05:37.260
But, um,

cue-92
00:05:38.640 --> 00:05:41.580
any-- in any event, meanwhile, the
nonprofit raised

cue-93
00:05:42.660 --> 00:05:43.000
zero,

cue-94
00:05:43.960 --> 00:05:45.900
zero dollars in twenty twenty-two.

cue-95
00:05:47.680 --> 00:05:48.320
And at that time,

cue-96
00:05:49.240 --> 00:05:53.140
OpenAI negotiated a deal with Microsoft.
What leverage do we have?

cue-97
00:05:54.500 --> 00:05:56.560
Now, the nonprofit received

cue-98
00:05:57.240 --> 00:06:01.320
five million dollars of donations in
twenty twenty-three, and that's point oh

cue-99
00:06:01.320 --> 00:06:05.596
five percent of Microsoft's ten billion
dollar investment.uh, in the OpenAI

cue-100
00:06:05.596 --> 00:06:09.296
for-profit. The 2023 deal in which
Microsoft

cue-101
00:06:09.936 --> 00:06:13.716
and all of the OpenAI dependents at the
expense of the nonprofit.

cue-102
00:06:14.476 --> 00:06:17.096
Let's talk about that enrichment. What was
the enrichment?

cue-103
00:06:18.256 --> 00:06:19.596
As I mentioned, Greg Brockman,

cue-104
00:06:20.516 --> 00:06:24.446
uh, never invested a nickel. Not only did
he never invest, he told people he was

cue-105
00:06:24.446 --> 00:06:26.536
going to invest to try to raise money, as
we saw,

cue-106
00:06:27.196 --> 00:06:28.796
and promised people he did that with,

cue-107
00:06:29.436 --> 00:06:30.276
and, um,

cue-108
00:06:31.516 --> 00:06:34.476
and never invested in it. And he now has a
stake worth $30

cue-109
00:06:35.136 --> 00:06:38.236
billion. By the way, I thought it was
interesting when I asked Sam Altman,

cue-110
00:06:38.996 --> 00:06:44.496
he said, "You know, Greg Altman's never
given that hundred thousand dollars."

cue-111
00:06:45.556 --> 00:06:49.216
He told us, of course, he was good with a
billion dollars, but he ended up with

cue-112
00:06:49.216 --> 00:06:52.756
thirty billion. Not that a billion dollars
would have been a problem with him. But,

cue-113
00:06:52.756 --> 00:06:56.756
and I asked Altman, I said, "Doesn't that
bother you? The guy's never given any

cue-114
00:06:56.756 --> 00:06:57.356
money. He's never..."

cue-115
00:06:58.026 --> 00:07:01.316
"You know, not really. You know, Greg's
made many contributions." I'm

cue-116
00:07:01.316 --> 00:07:06.136
paraphrasing. "Greg's made many
contributions to OpenAI." I mean, the

cue-117
00:07:06.136 --> 00:07:06.596
arrogance,

cue-118
00:07:07.416 --> 00:07:11.796
the, uh, lack of sensitivity, the failure
to account for

cue-119
00:07:14.016 --> 00:07:17.596
just kind of decency was really, really
appalling.

cue-120
00:07:18.696 --> 00:07:21.076
Remember, Brockman's plan from the
beginning

cue-121
00:07:21.696 --> 00:07:22.036
was

cue-122
00:07:23.216 --> 00:07:24.636
in his diary, "Okay,

cue-123
00:07:25.576 --> 00:07:28.356
financially, what will take me to one
billion dollars?

cue-124
00:07:29.636 --> 00:07:34.596
It would be nice to be making the
billions." We could put this out.

cue-125
00:07:36.556 --> 00:07:36.776
He said,

cue-126
00:07:37.526 --> 00:07:39.276
"We've been thinking that maybe

cue-127
00:07:40.136 --> 00:07:42.056
we should just flip to a for-profit.

cue-128
00:07:42.676 --> 00:07:48.196
Making the money for us sounds great and
all." That's what he was saying at that

cue-129
00:07:48.196 --> 00:07:48.496
time.

cue-130
00:07:51.316 --> 00:07:52.996
Those were Brockman's true thoughts

cue-131
00:07:54.076 --> 00:07:59.576
as he was discussing a path forward with
Elon. Um, he also enriched himself through

cue-132
00:07:59.576 --> 00:08:02.676
self-dealing and transactions. You heard
about Cerebras.

cue-133
00:08:03.596 --> 00:08:07.596
Um, and that was something that he
invested in and made money on. He's held

cue-134
00:08:07.596 --> 00:08:14.416
equity in, um, in, in, um, in Cerebras for
some period of time. OpenAI gave a ten

cue-135
00:08:14.416 --> 00:08:15.936
billion dollar deal to

cue-136
00:08:16.836 --> 00:08:20.056
Cerebras. Um, and he's not only the only
one making those kinds of billions of

cue-137
00:08:20.056 --> 00:08:20.436
dollars.

cue-138
00:08:21.096 --> 00:08:23.996
Um, as I mentioned, Ilya Sutskever has a
seven billion dollar,

cue-139
00:08:24.736 --> 00:08:28.836
uh, stake. I mentioned Joshua Atian, who
testified yesterday with the twenty

cue-140
00:08:28.836 --> 00:08:33.196
million. And, and Professor Schizer
testified, and this is why they're going

cue-141
00:08:33.196 --> 00:08:38.136
to bring up this email from early on when
they talked about that people that work

cue-142
00:08:38.136 --> 00:08:40.536
there can get full startup-like equity,
okay?

cue-143
00:08:41.616 --> 00:08:44.316
Nothing wrong with people making money.
There's nothing proper wrong with people

cue-144
00:08:44.316 --> 00:08:48.536
making money, uh, working for a nonprofit,
especially if they're doing work that's

cue-145
00:08:48.536 --> 00:08:50.436
challenging work that's going to require
people,

cue-146
00:08:51.256 --> 00:08:52.796
um, from the marketplace to come.

cue-147
00:08:54.216 --> 00:08:56.165
But Professor Schizer was very clear on
this,

cue-148
00:08:56.816 --> 00:09:01.296
that just because you're there and you're
doing an important job that might be

cue-149
00:09:01.296 --> 00:09:02.046
compensated

cue-150
00:09:02.796 --> 00:09:07.336
at a higher level outside of a nonprofit
doesn't mean that you're entitled to make

cue-151
00:09:07.336 --> 00:09:09.876
that same amount in a nonprofit.

cue-152
00:09:10.536 --> 00:09:13.366
He testified that, and he, and he gave the
example,

cue-153
00:09:14.096 --> 00:09:18.156
universities with large endowments,
Stanford or Berkeley for that matter, you

cue-154
00:09:18.156 --> 00:09:22.466
know, have these large endowments. Donors,
you know, donate money,

cue-155
00:09:23.096 --> 00:09:27.076
and the people that manage that money make
quite a lot of money themselves. They're

cue-156
00:09:27.076 --> 00:09:28.036
investment managers.

cue-157
00:09:28.796 --> 00:09:34.016
Those same people, if they were working at
a hedge fund, not a nonprofit like

cue-158
00:09:34.016 --> 00:09:38.976
Stanford or like Berkeley, would, um,
would, would, would be making more.

cue-159
00:09:39.796 --> 00:09:42.476
Well, they're, they're, they're
compensated well, but they're there in

cue-160
00:09:42.476 --> 00:09:43.896
part to serve the mission.

cue-161
00:09:44.816 --> 00:09:46.036
And, you know, here

cue-162
00:09:46.836 --> 00:09:48.456
it was just completely flipped on its
head.

cue-163
00:09:50.356 --> 00:09:50.576
Now,

cue-164
00:09:51.256 --> 00:09:53.556
um, the nonprofit funds

cue-165
00:09:54.336 --> 00:09:58.496
are supposed to, um, go to advancing the
mission,

cue-166
00:09:59.196 --> 00:10:04.036
and they're not supposed to make, uh,
their employees and for-profit partners

cue-167
00:10:04.036 --> 00:10:04.936
Microsoft rich.

cue-168
00:10:05.656 --> 00:10:06.396
What about Sam Altman?

cue-169
00:10:07.656 --> 00:10:11.496
Sam Altman, who told the United States
Senate that he has no equity in OpenAI.

cue-170
00:10:12.276 --> 00:10:14.176
All right? That was a lie, as I said
before.

cue-171
00:10:14.856 --> 00:10:17.116
And, um, he said that, um,

cue-172
00:10:17.876 --> 00:10:22.956
he has, you know, admitted in this, uh,
tech, this trial that he has a stake in

cue-173
00:10:22.956 --> 00:10:27.516
OpenAI through his former company Y
Combinator, which is sort of a company

cue-174
00:10:27.516 --> 00:10:27.996
that helps

cue-175
00:10:28.936 --> 00:10:30.016
companies that are starting up.

cue-176
00:10:30.736 --> 00:10:31.136
And

cue-177
00:10:32.196 --> 00:10:38.096
he also told you that former board member,
uh, Tasha McCauley, that he intent-

cue-178
00:10:38.096 --> 00:10:41.596
intended someday to take additional
equity. There's an email that he tells

cue-179
00:10:41.596 --> 00:10:46.036
Elon the same thing, where he says,
"Someday I may take equity." Now, maybe

cue-180
00:10:46.036 --> 00:10:49.696
that someday is the day that this trial's
over, or maybe it's some other point down

cue-181
00:10:49.696 --> 00:10:53.736
in the future. I don't have any evidence
that I can tell you, but I can just tell

cue-182
00:10:53.736 --> 00:10:59.136
you that what human nature is and what
past behavior often demonstrates as a

cue-183
00:10:59.136 --> 00:11:00.596
predictor of future performance.

cue-184
00:11:01.496 --> 00:11:03.676
Um, so you be the judges, but,

cue-185
00:11:04.316 --> 00:11:10.196
um, Sam Altman has no equity stake. But
what he does have is the ability to, as I

cue-186
00:11:10.196 --> 00:11:13.555
asked him in the interview, I said,
"You're the CEO. You have the ability to

cue-187
00:11:13.556 --> 00:11:17.236
influence contracts, the business that
they do, business partners that they,

cue-188
00:11:17.236 --> 00:11:18.176
that, that they work with."

cue-189
00:11:18.876 --> 00:11:24.476
And, uh, you heard about his self-dealing
being on both sides of deals between

cue-190
00:11:24.476 --> 00:11:30.296
OpenAI and on the one side and companies
that he's invested in on the other. Um,

cue-191
00:11:30.296 --> 00:11:33.816
his stake in those companies, there's an
exhibit that lists, um,

cue-192
00:11:34.776 --> 00:11:37.456
let me look at this here. What's the,
what's the exhibit number that has the

cue-193
00:11:37.456 --> 00:11:41.136
disclosures? It was a, what's called an
interrogatory where they submit, both he

cue-194
00:11:41.136 --> 00:11:44.536
and Brockman submitted under oath this
document saying, "These are the businesses

cue-195
00:11:44.536 --> 00:11:46.276
that we, uh, have interest in."

cue-196
00:11:46.996 --> 00:11:51.736
Um, anyway, I'll get it for you in a
minute. But, um, he, so his, so he, we'll

cue-197
00:11:51.736 --> 00:11:55.736
produce this to you, but his stake in
Helion Energy, remember he was explaining

cue-198
00:11:55.736 --> 00:11:59.316
how Helion, how energy is so important to
this company because

cue-199
00:12:00.156 --> 00:12:02.935
they gotta have it compute, and they gotta
have, you know, these computers need to

cue-200
00:12:02.936 --> 00:12:07.106
be powered, and that's why you hear all
the stories about energy and, and, and

cue-201
00:12:07.836 --> 00:12:12.956
such now. His stake in Helion Energy is
worth one point six five billion

cue-202
00:12:12.956 --> 00:12:16.120
dollars.1.6 trillion dollars. And he,

cue-203
00:12:17.440 --> 00:12:20.900
w-he was until recently chairman of the
board of directors of Helium.

cue-204
00:12:21.760 --> 00:12:25.580
Um, Helium does not have a working product
right now,

cue-205
00:12:26.260 --> 00:12:32.000
but it's got a deal with OpenAI that is
sort of a pathway to future transactions

cue-206
00:12:32.000 --> 00:12:36.460
once it gets a working product. And
Siobhan Zillis testified that when that

cue-207
00:12:36.460 --> 00:12:40.780
deal was announced, it was, quote, "Out of
left field," was a phrase that she used,

cue-208
00:12:40.780 --> 00:12:42.240
and it raised concerns to the board.

cue-209
00:12:42.880 --> 00:12:47.700
Uh, Sam Altman claims that he recused
himself from the transaction, but he

cue-210
00:12:47.700 --> 00:12:48.280
admitted,

cue-211
00:12:48.920 --> 00:12:51.700
uh, that he participated in discussions
about that deal.

cue-212
00:12:52.660 --> 00:12:56.620
Now, uh, another one of his large-- So he
said, "I'm recused," but yet then he went

cue-213
00:12:56.620 --> 00:12:58.720
ahead and negotiated, uh, anyway.

cue-214
00:12:59.940 --> 00:13:03.020
I think he gave the ex-the excuse that he
didn't really know what was going on or

cue-215
00:13:03.020 --> 00:13:07.440
something like that. Well, that's why we
don't allow people with conflicts to be

cue-216
00:13:07.440 --> 00:13:11.940
part of negotiations. Another large
investment he had was in Reddit.

cue-217
00:13:12.660 --> 00:13:13.760
Um, you may know Reddit,

cue-218
00:13:14.400 --> 00:13:18.460
uh, from the internet. Um, it's a social
media company. In twenty twenty-four,

cue-219
00:13:19.100 --> 00:13:23.160
uh, Altman shares in Reddit were worth
about eight hundred million dollars.

cue-220
00:13:24.100 --> 00:13:26.220
Um, he admitted there was an obvious
conflict

cue-221
00:13:26.820 --> 00:13:32.710
when, um-- between his Reddit investment
and his, um, position at OpenAI, the CEO

cue-222
00:13:32.710 --> 00:13:35.360
of O-OpenAI, because they did a deal
together.

cue-223
00:13:36.060 --> 00:13:40.960
And it didn't stop him from negotiating
both sides of that deal. And the story

cue-224
00:13:40.960 --> 00:13:44.240
that we heard was, well, there was some
litigation or threat of litigation. The

cue-225
00:13:44.240 --> 00:13:47.340
thought would be that I would be the
person that could solve it. He has a

cue-226
00:13:47.340 --> 00:13:49.940
fiduciary duty to act for

cue-227
00:13:50.900 --> 00:13:56.160
OpenAI, the nonprofit, act in their best
interest, not in Sam Altman's personal

cue-228
00:13:56.160 --> 00:13:57.760
interest, nor in Reddit's best interest.

cue-229
00:13:58.420 --> 00:14:01.700
And Professor Schizer, under the terms of
the deal that they have, by the way, it's

cue-230
00:14:01.700 --> 00:14:05.480
two hundred million dollars between OpenAI
and Reddit. Professor Schizer explained

cue-231
00:14:05.480 --> 00:14:08.680
that these conflicts of interest like
this, these are a problem because a

cue-232
00:14:08.680 --> 00:14:12.950
nonprofit's executives and board members
have this duty of loyalty. Do you remember

cue-233
00:14:12.950 --> 00:14:17.120
he talked about duty of loyalty? There
was a chart that had three bans on it.

cue-234
00:14:17.120 --> 00:14:20.880
Duty of loyalty was one of them. And that
means they need, they need to focus on the

cue-235
00:14:20.880 --> 00:14:21.750
nonprofit's business.

cue-236
00:14:22.620 --> 00:14:28.320
In addition to negotiating with both
sides, uh, he also, um, had this OpenAI

cue-237
00:14:28.320 --> 00:14:33.390
startup fund that he didn't tell the board
about. And, um, we heard the testimony of

cue-238
00:14:33.390 --> 00:14:35.210
the company's one director who said they
were surprised to

cue-239
00:14:36.040 --> 00:14:41.520
learn about this. OpenAI, um, also now has
converted to a public benefit

cue-240
00:14:41.520 --> 00:14:45.520
corporation. We'll hear a lot about that.
It was in twenty twenty-five and is

cue-241
00:14:45.520 --> 00:14:49.630
exploring something called an initial
public offering, which is where a company

cue-242
00:14:49.630 --> 00:14:50.280
can sell stock,

cue-243
00:14:50.920 --> 00:14:55.080
um, and sell shares to the investing
public. The PBC,

cue-244
00:14:55.920 --> 00:15:00.820
which it's now converting, there's no
caps, no caps on investors and

cue-245
00:15:00.820 --> 00:15:02.000
distributions made in the PBC.

cue-246
00:15:04.220 --> 00:15:05.900
No caps to the investors'

cue-247
00:15:06.740 --> 00:15:08.620
payments under the, uh, PBC.

cue-248
00:15:09.960 --> 00:15:14.690
Um, second, OpenAI failed to fulfill its
mission to open source technology.

cue-249
00:15:14.690 --> 00:15:16.460
Remember, that was something that was
important.

cue-250
00:15:17.260 --> 00:15:18.060
And, um,

cue-251
00:15:19.040 --> 00:15:24.880
by twenty twenty-three, that went out the
window. Um, OpenAI did give Microsoft

cue-252
00:15:24.880 --> 00:15:28.420
interest in commercializing product, okay?

cue-253
00:15:29.240 --> 00:15:33.940
Giving Microsoft these IP rights, the
know-how. What if Microsoft, uh--

cue-254
00:15:33.940 --> 00:15:37.740
M-Microsoft had this huge stake, and you
heard, you heard Mira Murati. She was one

cue-255
00:15:37.740 --> 00:15:40.200
of the witnesses that testified by, um,

cue-256
00:15:40.940 --> 00:15:41.980
deposition in, in--

cue-257
00:15:42.720 --> 00:15:46.140
And by the way, I don't remember what the
judge's instructions said this, but I

cue-258
00:15:46.140 --> 00:15:49.200
couldn't make those people come over
because they live too far away here from

cue-259
00:15:49.200 --> 00:15:54.040
us. But I can rotate their deposition and
have them testify by video. I apologize

cue-260
00:15:54.040 --> 00:15:57.240
for that because I know it's not always
the most interesting, but they came and

cue-261
00:15:57.240 --> 00:16:00.540
gave truthful testimony, and that's what
you heard. It's just the same as if

cue-262
00:16:00.540 --> 00:16:05.260
someone-- um, a witness there. And, and
she was the point person on OpenAI's

cue-263
00:16:05.260 --> 00:16:10.059
relationship with Microsoft, and she
testified that OpenAI agreed to give

cue-264
00:16:10.060 --> 00:16:11.280
Microsoft six months

cue-265
00:16:11.900 --> 00:16:17.340
or one year of exclusivity on GPT language
products. That's a big deal when it's a

cue-266
00:16:17.340 --> 00:16:19.420
new technology being developed like that.

cue-267
00:16:20.560 --> 00:16:24.160
So the failure to open source breached the
mission. And then on safety,

cue-268
00:16:24.760 --> 00:16:29.800
OpenAI failed, uh, it breached the mission
as well too. It was clear from Satya

cue-269
00:16:29.800 --> 00:16:34.900
Nadella's testimony that in early twenty
twenty-three, OpenAI was pressured by

cue-270
00:16:34.900 --> 00:16:36.760
Microsoft to release

cue-271
00:16:37.480 --> 00:16:39.240
a version, a paid version

cue-272
00:16:40.060 --> 00:16:44.380
of ChatGPT. You'll recall I showed him
before the end of his testimony,

cue-273
00:16:45.000 --> 00:16:50.840
three text messages between Satya Nadella
and, and Sam Altman. And these are

cue-274
00:16:50.840 --> 00:16:54.660
exhibits two sixty-nine, two seventy, and
two seventy-two, and again, you'll have

cue-275
00:16:54.660 --> 00:16:54.920
these.

cue-276
00:16:55.760 --> 00:16:59.340
In exhibit two seventy, he says, "The
sooner, the better for getting paid

cue-277
00:16:59.340 --> 00:17:00.780
ChatGPT set up."

cue-278
00:17:01.560 --> 00:17:06.940
Um, and so he's clearly pressuring Altman.
And it's common sense that speed,

cue-279
00:17:07.660 --> 00:17:08.940
speed doesn't equal safety.

cue-280
00:17:09.860 --> 00:17:14.880
Um, as Professor Schizer explained,
there's often a tension between profit

cue-281
00:17:14.880 --> 00:17:19.580
maximization in a, in a for-profit
subsidiary and adherence to the mission,

cue-282
00:17:19.580 --> 00:17:20.620
loyalty to the mission,

cue-283
00:17:21.220 --> 00:17:22.200
service to the mission.

cue-284
00:17:24.660 --> 00:17:26.240
We saw how it played out here,

cue-285
00:17:27.200 --> 00:17:27.360
uh,

cue-286
00:17:28.080 --> 00:17:29.820
in the testimony of Helen Toner

cue-287
00:17:30.460 --> 00:17:34.600
and Tasha McCauley. Those were two other
witnesses who testified by video. You

cue-288
00:17:34.600 --> 00:17:37.560
might remember they were both directors of
the company.

cue-289
00:17:38.520 --> 00:17:41.620
And, um, they testified

cue-290
00:17:42.520 --> 00:17:45.199
that, that, that Sam Altman lied to them,

cue-291
00:17:45.880 --> 00:17:51.360
and both Toner and McCauley testified that
it was related to, uh, pushing

cue-292
00:17:52.040 --> 00:17:56.670
for commercialization that led to the
concerns then for safety.

cue-293
00:17:58.000 --> 00:17:59.520
Concerns for safety.

cue-294
00:18:00.340 --> 00:18:03.860
Uh, you also heard testimony from Rosie
Campbell. Remember, that was a woman who

cue-295
00:18:03.860 --> 00:18:10.540
came and testified toward the end of our
case. And, um, she testified that, uh,

cue-296
00:18:10.540 --> 00:18:14.350
safety took a back seat at the end of her
tenure in twenty three-- excuse me, twenty

cue-297
00:18:14.350 --> 00:18:18.930
twenty-three and twenty twenty-four.Um,
and then OpenAI put up a, a whole bunch of

cue-298
00:18:18.930 --> 00:18:22.780
witnesses then at, at the end of the
trial to say that, "Oh, everything's fine

cue-299
00:18:22.780 --> 00:18:26.740
today." Well, that's not good enough,
right? It's not good enough.

cue-300
00:18:27.540 --> 00:18:30.370
We don't know whether it's fine or not. We
didn't really have a trial on whether it

cue-301
00:18:30.370 --> 00:18:30.880
was fine.

cue-302
00:18:31.540 --> 00:18:32.820
We heard the witnesses say that,

cue-303
00:18:33.520 --> 00:18:37.300
and the question is: what was it like in
2023 when they were in breach?

cue-304
00:18:39.100 --> 00:18:40.940
Don't let that distract you

cue-305
00:18:41.620 --> 00:18:43.600
from all of the safety failures that
occurred.

cue-306
00:18:48.940 --> 00:18:53.720
Recall all of those witnesses who
testified. There were several. There was

cue-307
00:18:53.720 --> 00:18:55.490
Satya Nadella at Microsoft, including,

cue-308
00:18:56.180 --> 00:18:56.440
um,

cue-309
00:18:57.380 --> 00:19:00.840
the, um, the chairman of the nonprofit,

cue-310
00:19:01.480 --> 00:19:03.310
um, Nat, uh, Taylor.

cue-311
00:19:04.760 --> 00:19:08.870
They testified on cross-examination, not
when it wasn't by their side, testified on

cue-312
00:19:08.870 --> 00:19:14.910
cross-examination that the OpenAI
nonprofit was a nutshell for many

cue-313
00:19:14.910 --> 00:19:17.170
years. In November,

cue-314
00:19:17.960 --> 00:19:20.499
uh, Michael Wetter, who testified
yesterday,

cue-315
00:19:21.160 --> 00:19:26.120
was, uh, deposed. He was a sworn statement
taken under oath, you know, where lawyers

cue-316
00:19:26.120 --> 00:19:28.000
get to question. He was deposed,

cue-317
00:19:29.020 --> 00:19:32.340
uh, as the corporate representative of
Microsoft,

cue-318
00:19:33.200 --> 00:19:35.060
and, um, he

cue-319
00:19:36.960 --> 00:19:41.270
knew as well as the, the, the corporate
representative of... Excuse me, I

cue-320
00:19:41.270 --> 00:19:41.640
misspoke.

cue-321
00:19:42.690 --> 00:19:45.880
The corporate representative from OpenAI

cue-322
00:19:47.020 --> 00:19:49.120
was deposed in November,

cue-323
00:19:49.180 --> 00:19:49.920
testifying,

cue-324
00:19:50.620 --> 00:19:52.660
and you saw the testimony under petition.

cue-325
00:19:54.160 --> 00:19:55.550
No full-time employment.

cue-326
00:19:57.040 --> 00:20:00.500
So they're jumping up and down saying,
"Oh, this is the richest nonprofit. It's

cue-327
00:20:00.500 --> 00:20:01.760
this great thing.

cue-328
00:20:03.060 --> 00:20:05.280
It's we- we're doing these wonderful
works."

cue-329
00:20:06.420 --> 00:20:10.940
They had no full-time employees. They only
hired full-time employees about a month

cue-330
00:20:10.940 --> 00:20:13.010
before this trial started, believe it was
six weeks.

cue-331
00:20:14.380 --> 00:20:15.060
What does that tell you?

cue-332
00:20:15.780 --> 00:20:19.940
What kind of people do that? What is, what
is their priority?

cue-333
00:20:21.040 --> 00:20:24.180
They claim, "Well, it's the mission." They
say, "Oh, well, we didn't have liquid

cue-334
00:20:24.180 --> 00:20:24.519
money."

cue-335
00:20:27.180 --> 00:20:30.060
I think it was all that I asked, I'm not
sure, but I did ask the question:

cue-336
00:20:30.700 --> 00:20:32.280
Couldn't you borrow against the asset?

cue-337
00:20:32.920 --> 00:20:36.320
Could-- If you have this asset, you know,
and, and, and you, and for whatever

cue-338
00:20:36.320 --> 00:20:39.280
reason, you can't sell it or there'd be
not an opportunity to sell it, it's worth

cue-339
00:20:39.280 --> 00:20:42.900
all this money. Couldn't you borrow some
money so you could actually do the work?

cue-340
00:20:42.900 --> 00:20:46.720
And what are they doing? They're doing
work that's making grants. They're not

cue-341
00:20:46.720 --> 00:20:47.619
doing

cue-342
00:20:47.680 --> 00:20:48.420
developing,

cue-343
00:20:49.060 --> 00:20:50.380
you know, AI

cue-344
00:20:52.300 --> 00:20:57.059
safely for the good of mankind and, you
know, making it open source with fact.

cue-345
00:20:57.060 --> 00:20:59.140
Well, they're not doing that.

cue-346
00:21:01.100 --> 00:21:01.980
Yeah, it has money.

cue-347
00:21:03.720 --> 00:21:05.260
How much money you think it might have

cue-348
00:21:06.100 --> 00:21:09.860
if they didn't breach the charitable
trust? If they did what they were supposed

cue-349
00:21:09.860 --> 00:21:12.840
to do, if they went out and raised the
money, if they didn't give away the store

cue-350
00:21:12.840 --> 00:21:13.470
to Microsoft,

cue-351
00:21:14.080 --> 00:21:20.340
it would have even more money. And hearing
Stuart Russell talk about the risk of AI,

cue-352
00:21:20.340 --> 00:21:23.120
God, I wish they would've done that.

cue-353
00:21:23.820 --> 00:21:25.740
God, I wish they would've done that. I
wish they would've had

cue-354
00:21:26.440 --> 00:21:32.140
a nonprofit with billions and billions and
billions of more dollars because the risk

cue-355
00:21:32.140 --> 00:21:33.860
is so high to us. But instead,

cue-356
00:21:34.620 --> 00:21:38.860
they're gonna trumpet the fact that, "Oh,
we have two hundred billion dollars." If

cue-357
00:21:38.860 --> 00:21:40.020
you go and rob a bank

cue-358
00:21:40.700 --> 00:21:42.740
and you take a million dollars from the
bank,

cue-359
00:21:43.560 --> 00:21:47.600
it's not a defense to say, "Oh, I left,
you know, a hundred million dollars in the

cue-360
00:21:47.600 --> 00:21:49.220
bank. I only took, you know, a million."

cue-361
00:21:50.460 --> 00:21:51.680
And that's essentially what they're
saying.

cue-362
00:21:52.900 --> 00:21:54.140
So I think that whole

cue-363
00:21:54.940 --> 00:21:58.280
argument and the testimony you heard is
nothing more than a farce.

cue-364
00:22:02.880 --> 00:22:03.760
You might recall that

cue-365
00:22:05.020 --> 00:22:06.340
Sam Altman's, uh,

cue-366
00:22:07.600 --> 00:22:11.520
pod-- uh, the, um, or podcast interview
that I, that I played.

cue-367
00:22:13.360 --> 00:22:14.780
They had governance

cue-368
00:22:15.880 --> 00:22:19.620
in place. The nonprofit board, right, is
supposed to oversee the for-profit.

cue-369
00:22:20.620 --> 00:22:23.400
It was interesting 'cause that's still the
case in this governance situation you

cue-370
00:22:23.400 --> 00:22:26.360
have. You remember when I asked him the
question, "You know, you sit on the

cue-371
00:22:26.360 --> 00:22:30.540
nonprofit board. Nonprofit board governs a
for-profit board. Um,

cue-372
00:22:32.300 --> 00:22:36.740
and that nonprofit board is supposed to,
you know, provide the safety." And I asked

cue-373
00:22:36.740 --> 00:22:39.800
him, "You know, are you the CEO of both
places and you're on the board of both

cue-374
00:22:39.800 --> 00:22:41.480
places? Would you fire yourself?"

cue-375
00:22:42.180 --> 00:22:46.680
Sounds like a silly question, but it's not
a silly question. It's a very pertinent,

cue-376
00:22:46.680 --> 00:22:47.600
important question

cue-377
00:22:48.360 --> 00:22:52.940
because that nonprofit board is supposed
to oversee the profit from the board. And,

cue-378
00:22:52.940 --> 00:22:56.960
and of course, he wouldn't fire himself.
He dissembled like he, he has in other,

cue-379
00:22:56.960 --> 00:22:59.180
other, you know, instances, trying to
dissemble,

cue-380
00:22:59.900 --> 00:23:04.240
um, wh-when, you know, he didn't really
answer my question. We went back and forth

cue-381
00:23:04.240 --> 00:23:05.360
a couple of times, but, you know,

cue-382
00:23:05.980 --> 00:23:06.340
it, it's--

cue-383
00:23:07.360 --> 00:23:08.420
there's no governance there

cue-384
00:23:09.260 --> 00:23:10.900
that's meaningful. And you know how we
know, too?

cue-385
00:23:11.520 --> 00:23:12.080
Because

cue-386
00:23:12.740 --> 00:23:16.560
that's what the nonprofit board had in
twenty twenty-three, right? The, the

cue-387
00:23:16.560 --> 00:23:18.420
ability to hire and fire the CEO.

cue-388
00:23:19.540 --> 00:23:20.160
And what did they do?

cue-389
00:23:21.260 --> 00:23:21.810
They fired

cue-390
00:23:22.580 --> 00:23:23.180
the CEO,

cue-391
00:23:24.780 --> 00:23:25.600
but then they rehired him,

cue-392
00:23:26.860 --> 00:23:27.720
and they fired him

cue-393
00:23:28.500 --> 00:23:29.040
for lying,

cue-394
00:23:30.100 --> 00:23:34.380
and not only for lying, but lying about
safety issues. Safety issues.

cue-395
00:23:35.280 --> 00:23:35.930
Is that governance?

cue-396
00:23:37.700 --> 00:23:41.060
He admitted, he admitted in that tape that
I played

cue-397
00:23:41.840 --> 00:23:42.030
that,

cue-398
00:23:42.660 --> 00:23:48.040
these were his words, not mine, "Although
the board had the legal authority to fire

cue-399
00:23:48.040 --> 00:23:48.200
me,

cue-400
00:23:48.900 --> 00:23:49.700
in practice,

cue-401
00:23:50.520 --> 00:23:51.680
it didn't quite work,

cue-402
00:23:52.880 --> 00:23:56.920
and that is its own kind of governance
failure."

cue-403
00:23:57.600 --> 00:23:58.480
He admitted that.

cue-404
00:24:00.900 --> 00:24:03.720
I don't know if he thought that that would
someday be played in court.

cue-405
00:24:06.100 --> 00:24:07.360
Maybe it was an honest thought.

cue-406
00:24:09.260 --> 00:24:11.390
Now, Professor Schizer explained that
management,

cue-407
00:24:12.080 --> 00:24:17.720
and in particular the CEO, has to be
honest and forthright and forthcoming and

cue-408
00:24:17.720 --> 00:24:20.820
candid with the board.Satya Nadella

cue-409
00:24:21.500 --> 00:24:24.620
testified to the same thing. You remember
I asked him,

cue-410
00:24:25.320 --> 00:24:26.040
um, about

cue-411
00:24:26.840 --> 00:24:30.700
if he had a senior executive that was
dishonest and not providing information?

cue-412
00:24:31.360 --> 00:24:35.819
And, and he said, "No, I'd fire him." And,
and I asked him a, a, a question that was

cue-413
00:24:35.819 --> 00:24:38.720
along these lines, and I could reword,
reword it a little bit, it was like, "And,

cue-414
00:24:38.720 --> 00:24:42.300
and he'd get him out of the building that
day?" And, and he agreed with me. He was

cue-415
00:24:42.300 --> 00:24:44.190
very fast on his answer that, "Yeah, I'd
fire him."

cue-416
00:24:45.260 --> 00:24:47.980
And you heard the same thing from Brett
Taylor,

cue-417
00:24:48.760 --> 00:24:54.340
the current, um, chair of the nonprofit,
and then Zyko Coulter, who was the safety

cue-418
00:24:54.340 --> 00:24:54.660
guy.

cue-419
00:24:55.380 --> 00:24:57.700
Three of them all said that 'cause it
makes, it's common sense.

cue-420
00:24:58.640 --> 00:25:02.360
It's common sense. Of course you would
fire somebody. What did they do?

cue-421
00:25:03.360 --> 00:25:07.320
They had a CEO who was not providing
information, who was deceiving them on

cue-422
00:25:07.320 --> 00:25:07.920
safety.

cue-423
00:25:10.080 --> 00:25:11.320
They fired him and they brought him back.

cue-424
00:25:13.140 --> 00:25:18.100
So having that right today is meaningless.
It really is. It's not about the

cue-425
00:25:18.100 --> 00:25:23.360
contract. It's not about what the, what
the co- what, what is in that document.

cue-426
00:25:23.360 --> 00:25:27.420
It's about how it's carried out, and we
saw how it was carried out. We saw what

cue-427
00:25:27.420 --> 00:25:29.760
they think, we saw what they care about.
They're motivated,

cue-428
00:25:30.380 --> 00:25:33.220
and they're motivated by money. Microsoft
and

cue-429
00:25:35.000 --> 00:25:36.140
Altman documents.

cue-430
00:25:38.080 --> 00:25:42.040
So in about eight minutes, it will have
been two hours for the jury-

cue-431
00:25:42.040 --> 00:25:42.820
-Okay
-... so-

cue-432
00:25:42.820 --> 00:25:45.350
-Okay
-... you should stop at a-

cue-433
00:25:45.350 --> 00:25:46.900
-Um-
-... at a convenient spot.

cue-434
00:25:46.900 --> 00:25:47.100
Okay.

cue-435
00:25:47.900 --> 00:25:50.540
Why don't we stop right now then, if
that's okay? I mean, I, I wanna, I have a

cue-436
00:25:50.540 --> 00:25:54.920
little bit more that I wanna do, but, uh,
if we can stop now, um, to take a break.

cue-437
00:25:56.120 --> 00:25:58.820
-Is that all right?
-We'll stand in recess with the jury for 15

cue-438
00:25:58.820 --> 00:25:59.640
minutes. Okay.

cue-439
00:26:00.740 --> 00:26:01.820
Please rise for the jury.

cue-440
00:26:23.540 --> 00:26:25.760
E- either way, you can, you can leave it
there

cue-441
00:26:27.280 --> 00:26:28.900
and then take it later. Doesn't

cue-442
00:26:28.900 --> 00:26:38.680
matter.

cue-443
00:26:38.680 --> 00:26:42.840
Just so that, um, just so I can manage
time, Mr. Mullow, what do you have?

cue-444
00:26:43.540 --> 00:26:44.900
-How much time do you have?
-Um,

cue-445
00:26:45.800 --> 00:26:46.880
for this?

cue-446
00:26:51.200 --> 00:26:52.400
-Oh.
-Less time.

cue-447
00:26:56.560 --> 00:26:58.100
All right. Well then we shouldn't have
broke.

cue-448
00:26:58.920 --> 00:26:59.040
Okay.

cue-449
00:26:59.720 --> 00:27:03.480
Then we should have finished your piece to
then take a break.

cue-450
00:27:07.940 --> 00:27:09.480
No, I just didn't... I, I didn't...

cue-451
00:27:10.340 --> 00:27:11.750
You've given me a lot of paper,

cue-452
00:27:12.940 --> 00:27:16.240
and so I thought perhaps you had another
45 minutes.

cue-453
00:27:18.720 --> 00:27:19.020
All right.

cue-454
00:27:21.060 --> 00:27:23.520
So we'll just have a couple minute
transition.

cue-455
00:27:24.280 --> 00:27:24.380
Uh,

cue-456
00:27:26.420 --> 00:27:27.120
all right, go ahead.

cue-457
00:27:30.900 --> 00:27:35.280
Transcript page 59 at 9:50 AM, Mr. Mullow
told the jury that, um,

cue-458
00:27:36.340 --> 00:27:41.460
page 59, 9:50 AM, Mr. Mullow told the jury
that we are not asking for money.

cue-459
00:27:41.460 --> 00:27:43.860
-I heard that.
-Um, and I'm troubled by that. I think

cue-460
00:27:43.860 --> 00:27:48.420
that's not true. I think it's not legally
possible, and I thought it was off limits,

cue-461
00:27:48.420 --> 00:27:48.980
and I

cue-462
00:27:50.540 --> 00:27:51.950
wanted to call to the court's attention,

cue-463
00:27:52.730 --> 00:27:55.780
uh, insofar as we think it's, um,
prejudicial and unfair.

cue-464
00:27:59.680 --> 00:27:59.790
I, I,

cue-465
00:28:00.560 --> 00:28:04.240
-could you please what you just said?
-You said that you told the jury that you

cue-466
00:28:04.240 --> 00:28:05.600
are not asking for money,

cue-467
00:28:06.620 --> 00:28:08.100
-and it-
-Oh, I did, I did say something along those

cue-468
00:28:08.100 --> 00:28:10.200
-lines, yes.
-You said that explicitly.

cue-469
00:28:10.200 --> 00:28:13.280
-Yes, yes.
-And that is actually not true.

cue-470
00:28:13.280 --> 00:28:17.480
-But we're not asking them for money.
-You are asking them for... You are asking

cue-471
00:28:17.480 --> 00:28:18.760
for money in this lawsuit.

cue-472
00:28:19.700 --> 00:28:24.400
Otherwise, I will not be having an
evidentiary hearing on Monday for billions

cue-473
00:28:24.400 --> 00:28:28.380
-of dollars of disgorgement.
-We're not asking money for ourselves.

cue-474
00:28:28.380 --> 00:28:33.460
It, no, you, you and I have a disagreement
about what and whether you can do.

cue-475
00:28:34.290 --> 00:28:36.540
-Okay.
-You need to retract that statement-

cue-476
00:28:36.540 --> 00:28:36.800
-Okay
-...

cue-477
00:28:38.780 --> 00:28:39.900
or you need to, um,

cue-478
00:28:41.020 --> 00:28:46.710
drop your claim for billions of dollars
and drop your claims for, uh,

cue-479
00:28:47.560 --> 00:28:50.940
-all the other phase two litigation.
-Okay.

cue-480
00:28:52.220 --> 00:28:56.040
-So you'll clean it up.
-But it is a cleanup. I mean, that's what

cue-481
00:28:56.040 --> 00:29:00.480
I, I, I mean, I just want you to know that
was where that came from with me. So I

cue-482
00:29:00.480 --> 00:29:02.500
appreciate it. I will, I will take care of
it.

cue-483
00:29:02.500 --> 00:29:05.460
But, Your Honor, I don't, I don't wanna
belabor this, but it's, it's not a minor

cue-484
00:29:05.460 --> 00:29:09.620
issue. It's something we spent a long time
on and it, a claim for restitution is

cue-485
00:29:09.620 --> 00:29:11.440
-exactly a claim for money, and-
-I-

cue-486
00:29:12.180 --> 00:29:13.280
-And, and-
-So what, what-

cue-487
00:29:13.920 --> 00:29:18.360
In addition, uh, we'd like the ability to
at least make, address this issue briefly

cue-488
00:29:18.360 --> 00:29:22.580
in our argument as well by way of, um,
response. I said one sentence.

cue-489
00:29:22.580 --> 00:29:24.460
It was, it was, uh,

cue-490
00:29:25.600 --> 00:29:27.440
uh... What did you say about the bank
robbery?

cue-491
00:29:28.720 --> 00:29:30.420
One sentence could be a lot.

cue-492
00:29:32.520 --> 00:29:35.440
-I don't think that one was, but-
-I, I disagree.

cue-493
00:29:36.300 --> 00:29:39.020
-You slipped it in nicely.
-I will, I will say

cue-494
00:29:40.880 --> 00:29:43.660
-I take that as a compliment. [laughs]
-I, I don't know. I mean-

cue-495
00:29:43.660 --> 00:29:46.270
It wasn't, it wasn't an attempt to slip
anything in. It was in the moment-

cue-496
00:29:46.270 --> 00:29:48.620
It was substantially contrary to something
that we've talked about

cue-497
00:29:49.720 --> 00:29:52.200
... so. Sure. I'm s- I, I, I will say
that...

cue-498
00:29:52.920 --> 00:29:53.280
I'm sorry.

cue-499
00:29:54.100 --> 00:29:56.940
I, I was in the moment. I said what I
said. I will correct it.

cue-500
00:29:57.880 --> 00:29:58.340
I will correct it.

cue-501
00:29:59.140 --> 00:30:01.480
We'll hear the correction, and then I'll
let you know.

cue-502
00:30:01.480 --> 00:30:03.660
-Thank you, Your Honor.
-All right. You've got, uh,

cue-503
00:30:04.440 --> 00:30:05.180
11 minutes.

cue-504
00:30:06.260 --> 00:30:10.660
May I have lead counsels come up please
for the certifications? I need lead

cue-505
00:30:10.660 --> 00:30:11.680
counsels to sign.

cue-506
00:30:14.160 --> 00:30:18.460
On the exhibits, I need partners to sign
this, not associates.

cue-507
00:30:19.860 --> 00:30:21.310
He said he needs lead counsel to sign the

cue-508
00:30:23.632 --> 00:30:26.552
Just that he needs to leave council to
sign certifications.

cue-509
00:30:27.592 --> 00:30:31.172
Only because.

cue-510
00:30:37.432 --> 00:30:42.092
What do we want to do about, what do we
want to do about this sound that? It

cue-511
00:30:48.932 --> 00:30:52.812
sounds like it's like mumble. It's like
low.

cue-512
00:30:54.352 --> 00:30:56.892
My question is, when he speaks, it

cue-513
00:30:58.652 --> 00:31:00.352
sounds like, you know,

cue-514
00:31:01.512 --> 00:31:03.832
like mumble. Like, there's no
articulation.

cue-515
00:31:04.572 --> 00:31:11.552
-It's like too low a mumble.
-Yeah. Yeah. Oh, sorry. Make it a

cue-516
00:31:11.552 --> 00:31:16.432
little louder. You're not taking a lot of
the location. But we may not have a party

cue-517
00:31:16.432 --> 00:31:16.732
in these.

cue-518
00:31:21.862 --> 00:31:25.732
-Thank you.
-Thank you. Let's do the sign language.

cue-519
00:31:25.732 --> 00:31:30.392
-Nothing.
-Thank you.

cue-520
00:31:35.152 --> 00:31:37.432
Is there anything else you need us to
sign?

cue-521
00:31:37.432 --> 00:31:39.372
-That's it. Thank you.
-Okay. Good. That's it.

cue-522
00:31:40.492 --> 00:31:47.222
For the plaintiff, I mean, Mr.

cue-523
00:31:47.222 --> 00:31:48.432
Molnar. Okay.

cue-524
00:31:49.252 --> 00:31:50.972
-Oh, about who said-
-Or...

cue-525
00:31:53.812 --> 00:31:57.232
Okay. Um, so it needs to sign it. Can I
bring it back?

cue-526
00:32:14.612 --> 00:32:15.052
I

cue-527
00:32:20.552 --> 00:32:21.792
can turn the screen now.

cue-528
00:32:22.572 --> 00:32:23.452
Have you been using it?

cue-529
00:32:24.652 --> 00:32:25.332
Turn with.

cue-530
00:32:26.372 --> 00:32:28.272
Towards Mr. Savitt for the counsel.

cue-531
00:32:29.892 --> 00:32:31.152
The question, though.

cue-532
00:32:54.152 --> 00:33:05.052
Yeah.

cue-533
00:33:05.052 --> 00:33:06.852
All right.

cue-534
00:33:07.532 --> 00:33:10.812
-Thank you.
-Thank you.

cue-535
00:33:10.812 --> 00:33:12.752
That was cool.

cue-536
00:33:14.432 --> 00:33:15.332
Thank

cue-537
00:33:15.332 --> 00:33:41.561
you.

cue-538
00:33:41.561 --> 00:33:44.572
Sorry. You're welcome.

cue-539
00:33:44.572 --> 00:36:23.472
[clears throat]

cue-540
00:36:35.162 --> 00:36:36.942
Yeah, we can use that. But we're gonna
turn the podium

cue-541
00:36:37.622 --> 00:36:38.282
toward the screen.

cue-542
00:36:48.402 --> 00:39:04.661
[background noise]

cue-543
00:39:04.661 --> 00:39:07.502
No, that's famous.

cue-544
00:39:27.022 --> 00:40:16.522
[background noise]

cue-545
00:40:16.522 --> 00:40:22.872
We're gonna

cue-546
00:40:22.872 --> 00:40:27.462
have to

cue-547
00:40:27.462 --> 00:40:39.532
[background noise]

cue-548
00:40:39.532 --> 00:40:39.532
.

cue-549
00:40:40.692 --> 00:42:06.782
[background noise]

cue-550
00:42:07.712 --> 00:42:09.182
I'm sorry. Yeah.

cue-551
00:42:38.322 --> 00:42:38.392
[background noise]

cue-552
00:42:48.252 --> 00:42:53.672
-Court is back in session
-All right, folks, settle down

cue-553
00:42:53.672 --> 00:42:55.152
-I copy
-Let's call in the jury

cue-554
00:43:01.312 --> 00:43:08.272
Hey, what I was gonna suggest

cue-555
00:43:08.272 --> 00:43:11.292
that I say, I hope this is okay because
that's why I'm previewing it with you,

cue-556
00:43:11.932 --> 00:43:16.902
is that, um, I made a statement that, um,
that we're not asking for money,

cue-557
00:43:17.732 --> 00:43:22.632
that the verdict form will not ask you for
money, but that money is implicated in

cue-558
00:43:22.632 --> 00:43:24.002
the case. I think that's--

cue-559
00:43:24.912 --> 00:43:26.472
Given the one statement that I made

cue-560
00:43:28.832 --> 00:43:32.432
Your Honor, we strongly feel that it would
be safer and preferable for the court to

cue-561
00:43:32.432 --> 00:43:36.632
give the instruction because we fear there
is a risk of real confusion and further

cue-562
00:43:36.632 --> 00:43:39.071
damage to the process if it's left to
counsel.

cue-563
00:43:40.892 --> 00:43:42.792
-All right, I'm gonna say that.
-Okay.

cue-564
00:43:44.132 --> 00:43:45.132
Let's-- We'll call 'em in.

cue-565
00:43:46.712 --> 00:43:49.912
And one of the things, Judge, you had said
I'd been going almost two hours.

cue-566
00:43:49.912 --> 00:43:52.512
According to at least our clock, it was
nine-oh-two.

cue-567
00:43:52.512 --> 00:43:55.172
The jur-- I said the jury has been sitting
there-

cue-568
00:43:55.172 --> 00:43:56.632
-Oh
-... for almost two hours.

cue-569
00:43:56.632 --> 00:43:58.752
-Okay.
-They started at eight thirty.

cue-570
00:43:58.752 --> 00:43:58.952
Okay.

cue-571
00:43:59.672 --> 00:44:12.572
[clears throat]

cue-572
00:44:12.572 --> 00:44:17.772
Oh,

cue-573
00:44:17.772 --> 00:44:20.152
one, two, three.

cue-574
00:44:21.832 --> 00:44:22.642
One, two, three. One, two, three.

cue-575
00:44:24.152 --> 00:44:27.912
Well, I just, um, we, we seem to be, uh...

cue-576
00:44:30.652 --> 00:44:30.852
Yeah.

cue-577
00:44:32.052 --> 00:44:34.142
Okay. Well.

cue-578
00:45:19.512 --> 00:45:22.532
-Okay.
-Okay, we're testing the mics. You may all

cue-579
00:45:22.532 --> 00:45:23.052
be seated.

cue-580
00:45:23.912 --> 00:45:24.872
Testing, testing.

cue-581
00:45:31.252 --> 00:45:37.172
Is that working for you? Great. Okay. So
before, um, Mr. Mullow finishes off here,

cue-582
00:45:37.172 --> 00:45:43.652
I do want to, uh, make sure that one thing
is, is clear. He made a statement, uh, in

cue-583
00:45:43.652 --> 00:45:47.252
his earlier remarks that they were not
asking for money.

cue-584
00:45:47.992 --> 00:45:50.452
The verdict form will not have, um,

cue-585
00:45:51.152 --> 00:45:56.832
a, a line for money, but money is in fact
implicated in this case. Okay? So

cue-586
00:45:57.672 --> 00:46:00.952
just wanna make sure everybody's clear
about that. Go ahead, Mr. Mullow.

cue-587
00:46:06.282 --> 00:46:09.382
-[echoing]
-Right there. Mr. Mullow, can you move that

cue-588
00:46:09.382 --> 00:46:12.812
mic over to the side so we don't have
feedback? Thank you.

cue-589
00:46:13.421 --> 00:46:14.312
Proceed when you're ready.

cue-590
00:46:15.972 --> 00:46:17.812
You'll see the verdict form when you get
the verdict form.

cue-591
00:46:18.432 --> 00:46:20.592
So we'll give the instructions. Um,

cue-592
00:46:21.392 --> 00:46:21.612
so

cue-593
00:46:24.892 --> 00:46:29.752
basically, we've covered the idea of, you
know, what the evidence related to the, to

cue-594
00:46:29.752 --> 00:46:29.892
the

cue-595
00:46:30.492 --> 00:46:32.532
breach of, um, charitable trust claim.

cue-596
00:46:33.332 --> 00:46:33.462
Um,

cue-597
00:46:35.232 --> 00:46:39.592
twenty twenty-three is when it happens, as
I mentioned, enriching investors and

cue-598
00:46:39.592 --> 00:46:42.692
insiders, um, closing off technology,

cue-599
00:46:43.832 --> 00:46:50.122
failing to prioritize safety, and failing
to follow nonprofit, proper nonprofit

cue-600
00:46:50.122 --> 00:46:52.272
procedures consistent with

cue-601
00:46:53.392 --> 00:46:54.112
what they should have done.

cue-602
00:46:54.972 --> 00:46:58.312
So let me talk about the other two claims,
and I'm not gonna do for a long time

cue-603
00:46:58.312 --> 00:46:59.352
because there's some overlap.

cue-604
00:47:00.111 --> 00:47:02.872
Um, the second claim is the unjust
enrichment claim,

cue-605
00:47:03.472 --> 00:47:06.052
and that's against the same OpenAI
defendants.

cue-606
00:47:06.772 --> 00:47:11.452
Um, the first element of that claim is,
there's two elements as well. First

cue-607
00:47:11.452 --> 00:47:13.052
element is receipt

cue-608
00:47:13.872 --> 00:47:16.752
of a benefit, and the benefit includes any
advantage.

cue-609
00:47:17.452 --> 00:47:23.652
So, um, it can be money, but it need not
be just money. Um, the, this includes the

cue-610
00:47:23.652 --> 00:47:26.532
thirty-eight million dollars that Elon
gave OpenAI.

cue-611
00:47:27.192 --> 00:47:31.532
Elon provided the majority of OpenAI seed
funding, as I said, and that's certainly a

cue-612
00:47:31.532 --> 00:47:31.972
benefit.

cue-613
00:47:32.772 --> 00:47:38.162
Um, so just the financial contributions
are, are, are enough. But he provided, and

cue-614
00:47:38.162 --> 00:47:41.672
you heard testimony about this from
multiple witnesses throughout the trial,

cue-615
00:47:41.672 --> 00:47:45.552
he provided much more than that. He gave
his name and reputation.

cue-616
00:47:46.172 --> 00:47:51.412
Um, he recruited key people. He leveraged
his connections, like we heard, with, uh,

cue-617
00:47:51.412 --> 00:47:53.852
Satya Nadella and with, uh, Jesse Wang,

cue-618
00:47:54.672 --> 00:47:59.092
uh, from, uh, um, you know, to get, to get
com- compute, to get computers.

cue-619
00:47:59.872 --> 00:48:02.322
And, um, he provided

cue-620
00:48:03.012 --> 00:48:07.212
strategic guidance, um, to the-- and, and
taught the co-founders, as he put it,

cue-621
00:48:07.212 --> 00:48:08.172
everything that he knew.

cue-622
00:48:08.852 --> 00:48:11.772
Um, you heard about the reputation, you
heard about the recruiting,

cue-623
00:48:12.552 --> 00:48:16.792
um, his industry connections. Obviously,
he was a very, very prominent person in

cue-624
00:48:16.792 --> 00:48:18.392
two thousand and fifteen in the tech
world.

cue-625
00:48:19.252 --> 00:48:24.672
Um, and so I'm not gonna go over all of
that again. You know what those, um, what,

cue-626
00:48:24.672 --> 00:48:25.082
what those

cue-627
00:48:26.092 --> 00:48:32.132
incidents were, what they involved, and,
and what he did. Um, he also, though, he

cue-628
00:48:32.132 --> 00:48:34.872
also, notwithstanding what, what you're
hearing

cue-629
00:48:35.692 --> 00:48:36.112
now,

cue-630
00:48:36.992 --> 00:48:43.772
um, he also really did engage with and
mentor and help with the strategy,

cue-631
00:48:43.772 --> 00:48:45.852
um, of OpenAI when it first began.

cue-632
00:48:46.512 --> 00:48:53.304
Uh, you might, you might remember that he,
um...Ilya

cue-633
00:48:53.304 --> 00:48:55.624
Sutskever said that he met with him
regularly,

cue-634
00:48:56.564 --> 00:49:01.004
and you, you might remember 'cause they
got put up a couple of times, exhibits

cue-635
00:49:01.004 --> 00:49:06.824
ninety-eight and ninety-nine, those were
emails from Ilya and the, and from Brad

cue-636
00:49:06.824 --> 00:49:07.264
Brockman,

cue-637
00:49:07.884 --> 00:49:09.724
and, um, you see 'em right here,

cue-638
00:49:11.404 --> 00:49:15.824
basically praising Elon, thanking him,
telling him it was an honor to work with

cue-639
00:49:15.824 --> 00:49:21.204
him. And so any suggestion-- And, and, and
by the way, these are dated January of

cue-640
00:49:21.204 --> 00:49:25.964
two, twenty eighteen. So this is, you
know, a couple years into the venture at

cue-641
00:49:25.964 --> 00:49:26.484
this point.

cue-642
00:49:27.224 --> 00:49:27.384
Um,

cue-643
00:49:29.084 --> 00:49:30.444
any suggestion that,

cue-644
00:49:31.144 --> 00:49:34.644
you know, he wasn't around, he was an
absentee, I think, I think, um,

cue-645
00:49:36.744 --> 00:49:37.264
we heard

cue-646
00:49:37.904 --> 00:49:42.404
argument that he, he-- or I'm sorry, maybe
it was one of the OpenAI witnesses. So,

cue-647
00:49:42.404 --> 00:49:42.524
you know,

cue-648
00:49:44.084 --> 00:49:45.924
he, he, he came by once in a while or

cue-649
00:49:47.004 --> 00:49:51.024
met with people, you know, over... That's
not true. I mean, you don't-- Ilya

cue-650
00:49:51.024 --> 00:49:51.424
Sutskever

cue-651
00:49:52.184 --> 00:49:53.644
is one of the premier,

cue-652
00:49:54.254 --> 00:49:55.984
premier artificial intelligence

cue-653
00:49:56.643 --> 00:50:00.044
scientists in the world. I mean, there's
no question about it. It was a big get for

cue-654
00:50:00.044 --> 00:50:01.604
them to get him to come from Google,

cue-655
00:50:02.304 --> 00:50:05.784
and for him to write an email like this
really says something. So,

cue-656
00:50:06.544 --> 00:50:10.984
um, he did a lot. He did a lot. And the
second element is the unjust retention of

cue-657
00:50:10.984 --> 00:50:15.704
that benefit. In considering whether a
benefit is unjustly retained,

cue-658
00:50:16.324 --> 00:50:20.684
you can consider whether the defendants
knew or had reason to know that the

cue-659
00:50:20.684 --> 00:50:25.724
plaintiff was, that the benefit was to be
used for charitable purposes, and that it

cue-660
00:50:25.724 --> 00:50:26.044
wasn't.

cue-661
00:50:26.664 --> 00:50:26.964
Okay?

cue-662
00:50:27.624 --> 00:50:31.684
And that's exactly what happened here. The
defendants had every reason to know that

cue-663
00:50:31.684 --> 00:50:37.984
Elon's contributions were charity. It's a
nonprofit. And, you know, we looked at the

cue-664
00:50:37.984 --> 00:50:42.264
emails where Elon insisted on the
nonprofit structure. They are on these

cue-665
00:50:42.264 --> 00:50:48.124
emails. Um, we saw how Elon, uh, told his
co-founders that he did not want to fund

cue-666
00:50:48.124 --> 00:50:48.704
OpenAI,

cue-667
00:50:49.404 --> 00:50:53.144
uh, if it was going in the wrong
direction. He insisted that they commit to

cue-668
00:50:53.144 --> 00:50:58.684
the nonprofit. So I don't think there's
much doubt about that. And, um, they were

cue-669
00:50:58.684 --> 00:51:01.924
enriching not only themselves, but
obviously their commercial partner,

cue-670
00:51:01.924 --> 00:51:02.544
Microsoft.

cue-671
00:51:04.144 --> 00:51:06.424
So they've unjustly retained those
benefits.

cue-672
00:51:07.204 --> 00:51:11.004
Uh, heard about, again, Brockman's
ownership worth thirty billion.

cue-673
00:51:11.804 --> 00:51:16.004
Altman is getting the benefits that he's
getting from his role at OpenAI. The third

cue-674
00:51:16.004 --> 00:51:19.613
claim, final claim that we'll bring is
aiding and abetting against Microsoft, and

cue-675
00:51:19.613 --> 00:51:22.864
to establish that Microsoft aided and
abetted OpenAI,

cue-676
00:51:23.544 --> 00:51:24.904
Altman, and Brockman

cue-677
00:51:25.564 --> 00:51:30.244
in a breach, um, uh, OpenAI, Altman, and
Brockman's breach of charitable trust.

cue-678
00:51:32.144 --> 00:51:34.484
We have to prove that Microsoft had
knowledge

cue-679
00:51:35.244 --> 00:51:36.944
that Elon made his donations

cue-680
00:51:37.604 --> 00:51:43.644
for specific charitable purposes, and that
OpenAI defendants were going to breach

cue-681
00:51:43.644 --> 00:51:47.924
the charitable trust by using Elon's
donations for another purpose.

cue-682
00:51:48.964 --> 00:51:53.764
That Microsoft gave substantial assistance
or encouragement to open A-- to the

cue-683
00:51:53.764 --> 00:51:59.344
OpenAI defendants in breaching the
charitable trust created by Elon. And that

cue-684
00:51:59.344 --> 00:52:04.604
Microsoft's conduct was a substantial
factor, excuse me, in harming Elon.

cue-685
00:52:06.084 --> 00:52:07.184
So knowledge, excuse me.

cue-686
00:52:14.804 --> 00:52:15.524
So, um,

cue-687
00:52:16.704 --> 00:52:17.584
knowledge has two parts.

cue-688
00:52:18.224 --> 00:52:20.364
Uh, first, Microsoft, uh, knew

cue-689
00:52:21.004 --> 00:52:25.704
that Elon made his donations for a
specific charitable purpose. They did. And

cue-690
00:52:25.704 --> 00:52:30.524
second, Microsoft knew that open A-- the
OpenAI defendants were going to breach the

cue-691
00:52:30.524 --> 00:52:34.604
charitable trust by using Elon's
donations for another period, purpose,

cue-692
00:52:34.604 --> 00:52:35.004
excuse me.

cue-693
00:52:36.324 --> 00:52:42.764
Microsoft was aware of what OpenAI was
doing every step of the way. Um, it helped

cue-694
00:52:42.764 --> 00:52:45.024
Altman and Brockman violate OpenAI's
mission,

cue-695
00:52:45.684 --> 00:52:49.384
and, um, that's aiding and abetting, it's
just pure and simple. Back in two thousand

cue-696
00:52:49.384 --> 00:52:55.393
and fifteen, Microsoft knew that Elon,
uh, publicly advocated for AI safety. This

cue-697
00:52:55.393 --> 00:52:59.164
is something you heard the testimony from
the lips of the, of the CEO of the

cue-698
00:52:59.164 --> 00:53:00.324
company, Satya Nadella.

cue-699
00:53:00.944 --> 00:53:01.764
Um, and

cue-700
00:53:02.504 --> 00:53:04.444
he, he said that right here in, in this
courtroom.

cue-701
00:53:09.564 --> 00:53:13.604
He j- his testimony was, his quote, "I
knew that he was concerned about AI

cue-702
00:53:13.604 --> 00:53:13.934
safety,

cue-703
00:53:14.624 --> 00:53:20.184
and I a-- and he also, I think, had deep
concerns on AI being cornered by one

cue-704
00:53:20.184 --> 00:53:20.584
company.

cue-705
00:53:21.304 --> 00:53:22.714
And so therefore, that's why

cue-706
00:53:23.384 --> 00:53:26.364
he was, you know, personally dedicating
himself to build AI."

cue-707
00:53:27.204 --> 00:53:27.444
Um,

cue-708
00:53:28.064 --> 00:53:32.864
and the day after OpenAI launched, Nadella
emailed the link about the announcement.

cue-709
00:53:32.864 --> 00:53:34.664
Remember that, that we saw earlier in
this?

cue-710
00:53:35.284 --> 00:53:38.904
Um, he emailed that to a bunch of senior
executives at, um, at

cue-711
00:53:39.824 --> 00:53:41.324
Microsoft. Um,

cue-712
00:53:43.824 --> 00:53:46.824
and the announcement clearly says-- Let me
just put it up really quickly.

cue-713
00:53:52.424 --> 00:53:55.704
Okay, we don't have to bring it up, as
you've seen it before. But it basically,

cue-714
00:53:55.704 --> 00:54:00.604
you know, clearly says that, um, it's a
nonprofit organization, and that, and that

cue-715
00:54:00.604 --> 00:54:02.584
Elon is one of the funders and founders.

cue-716
00:54:03.184 --> 00:54:10.024
Um, they knew that the profits, um, um,
from their donations and the

cue-717
00:54:10.024 --> 00:54:15.404
assets, um, had to be used to advance the
mission. Nadella himself being a trustee

cue-718
00:54:15.404 --> 00:54:16.964
of a prominent nonprofit.

cue-719
00:54:17.784 --> 00:54:19.584
Um, and he testified that, um,

cue-720
00:54:20.544 --> 00:54:23.584
he knew that E-Elon was a major donor.
That was his testimony.

cue-721
00:54:25.204 --> 00:54:28.744
Microsoft executives, Kevin Scott and
Michael Walter, both testified that they

cue-722
00:54:28.744 --> 00:54:30.224
knew Elon Musk was involved

cue-723
00:54:30.844 --> 00:54:34.104
in OpenAI's founding and was an early
donor. Um,

cue-724
00:54:34.804 --> 00:54:39.584
and in two thousand and sixteen, there's
an internal presentation that Microsoft

cue-725
00:54:39.584 --> 00:54:42.054
does when they're thinking about doing
more,

cue-726
00:54:42.844 --> 00:54:45.163
um, compute with OpenAI,

cue-727
00:54:45.164 --> 00:54:49.164
and it's defined in this deck, which is,
uh, plaintiff's exhibit two two nine,

cue-728
00:54:50.104 --> 00:54:55.604
uh, as a high pro-profile nonprofit
organization, and it identified Elon as a

cue-729
00:54:55.604 --> 00:55:01.740
founder.You heard Kevin Scott, who
testified yesterday, tell you that he knew

cue-730
00:55:01.740 --> 00:55:06.620
nothing about OpenAI's open source
practice, but his email says something

cue-731
00:55:06.620 --> 00:55:06.920
different.

cue-732
00:55:07.560 --> 00:55:12.560
I'll direct you to exhibit two two nine,
where, um, it's clear that

cue-733
00:55:13.220 --> 00:55:13.999
he does know that.

cue-734
00:55:14.659 --> 00:55:18.140
Microsoft's executives were told in
January of two thousand and eighteen that

cue-735
00:55:19.540 --> 00:55:24.660
OpenAI had been sharing their work openly
for their basic tenet. That's what he

cue-736
00:55:24.660 --> 00:55:30.340
calls it in the email, two two nine. Basic
tenet, which he says is open.

cue-737
00:55:31.240 --> 00:55:31.440
Um,

cue-738
00:55:32.580 --> 00:55:34.680
Microsoft knew that-- about the intent to
breach.

cue-739
00:55:35.600 --> 00:55:40.100
They were interested in supporting the
high pro-profile nonprofit. They knew that

cue-740
00:55:40.100 --> 00:55:43.460
it was a nonprofit and that they had to
follow their nonprofit vision.

cue-741
00:55:44.080 --> 00:55:47.540
Uh, again, that statement that I said to
him, "Enough with the charity and on with

cue-742
00:55:47.540 --> 00:55:49.000
the business," and he agreed to that.

cue-743
00:55:49.680 --> 00:55:54.140
Um, Kevin Sco-Kevin Scott, this is kind of
interesting, you know, he spotted the

cue-744
00:55:54.140 --> 00:55:59.680
risk that this would be a problem for
Microsoft to get too, you know, deeply

cue-745
00:55:59.680 --> 00:56:04.080
into it in the first investment in, um,
two thousand and eighteen. And he wrote

cue-746
00:56:04.080 --> 00:56:04.700
this email

cue-747
00:56:05.360 --> 00:56:06.740
saying, "I wonder if the

cue-748
00:56:07.060 --> 00:56:10.930
OpenAI donors are aware of these plans."
And then,

cue-749
00:56:11.680 --> 00:56:17.160
um, basically, that notion of building a
nonprofit on the back of, um, uh, this

cue-750
00:56:17.160 --> 00:56:21.479
thing on the back of a nonprofit. And
Microsoft did some diligence, satisfied

cue-751
00:56:21.480 --> 00:56:25.980
themselves that the billion-dollar
investment cap profit or two billion

cue-752
00:56:25.980 --> 00:56:27.660
dollar investment cap profit,

cue-753
00:56:28.620 --> 00:56:29.820
no twenty percent

cue-754
00:56:30.560 --> 00:56:35.080
increase like the twenty twenty-three deal
had, no broad, broad,

cue-755
00:56:35.720 --> 00:56:40.380
um, obtaining of IP rights like the twenty
twenty-three deal had, that he satisfied

cue-756
00:56:40.380 --> 00:56:44.520
himself that, uh, or they satisfied
themselves that, that that was okay. And

cue-757
00:56:44.520 --> 00:56:49.260
like I said, there's nothing wrong with
having a for-profit or a partnership with

cue-758
00:56:49.260 --> 00:56:54.050
a for-profit and nonprofit, but there are
limits. And, you know, they exceeded those

cue-759
00:56:54.050 --> 00:56:55.060
limits, and they had

cue-760
00:56:56.000 --> 00:56:57.940
features of the deal that made it

cue-761
00:56:58.840 --> 00:57:01.360
no longer subservient to the nonprofit.

cue-762
00:57:02.040 --> 00:57:02.240
Um,

cue-763
00:57:02.860 --> 00:57:06.400
things changed with the ten billion dollar
investment. We've talked about that.

cue-764
00:57:07.860 --> 00:57:08.919
And, um,

cue-765
00:57:14.820 --> 00:57:17.940
in their, in their diligence, they
reviewed,

cue-766
00:57:18.660 --> 00:57:20.080
uh, the Delaware certificate

cue-767
00:57:20.720 --> 00:57:24.480
that I showed you before and put up a
number of times, which says, "No part of

cue-768
00:57:24.480 --> 00:57:28.800
the, of the income or assets of the
corporation shall ever inure to the

cue-769
00:57:28.800 --> 00:57:32.560
benefit of any private person." Private
person includes Microsoft.

cue-770
00:57:33.320 --> 00:57:34.780
Person can mean corporation.

cue-771
00:57:35.960 --> 00:57:36.180
Um,

cue-772
00:57:36.880 --> 00:57:38.720
and they also reviewed the IRS documents,

cue-773
00:57:39.600 --> 00:57:43.280
and they also indicated that they were
supposed to-- they intended to make

cue-774
00:57:43.280 --> 00:57:44.480
research freely available.

cue-775
00:57:45.260 --> 00:57:48.800
So it's all there. It's all in the
documents that, that you will have for

cue-776
00:57:48.800 --> 00:57:49.540
your deliberations.

cue-777
00:57:53.800 --> 00:57:54.920
Substantial assistance.

cue-778
00:57:56.020 --> 00:57:56.110
Um,

cue-779
00:57:58.160 --> 00:58:01.800
you remember at the end of Satya Nadella's
testimony, I asked him some questions

cue-780
00:58:01.800 --> 00:58:02.920
about, you know, had,

cue-781
00:58:03.720 --> 00:58:06.940
had Microsoft done a thirteen billion
dollar deal the year before,

cue-782
00:58:07.800 --> 00:58:10.160
year before that, year before that? He
said, "No."

cue-783
00:58:11.220 --> 00:58:14.040
He didn't use the word never, but he said,
he said no, certainly.

cue-784
00:58:14.660 --> 00:58:17.220
And he, he called it a really big deal,
really significant,

cue-785
00:58:17.830 --> 00:58:19.740
um, in his testimony. So

cue-786
00:58:20.480 --> 00:58:23.820
this was a big-- Thirteen billion. He was
saying, "Well, we took a risk." Well,

cue-787
00:58:23.820 --> 00:58:27.220
yeah, they took a risk, but Microsoft
doesn't take a risk unless they're

cue-788
00:58:27.220 --> 00:58:30.530
convinced that they're gonna get a return
on their investment. They can't. They're--

cue-789
00:58:30.530 --> 00:58:34.820
They have a fiduciary duty. Just like the
people involved in the nonprofit have a

cue-790
00:58:34.820 --> 00:58:41.040
fiduciary duty to the nonprofit's mission,
Microsoft has a duty, a fiduciary duty to

cue-791
00:58:41.040 --> 00:58:46.980
maximize th-the shareholder value, to do
the best for the shareholders, to create a

cue-792
00:58:46.980 --> 00:58:50.720
profit. So they were very calculated.
They, they weren't just doing this to

cue-793
00:58:50.720 --> 00:58:54.120
experiment, to see how is this gonna work.
They thought they'd make a lot of money,

cue-794
00:58:54.120 --> 00:58:55.769
and they drove a hard bargain to do that.

cue-795
00:58:56.460 --> 00:59:00.000
Um, Mira Murati testified about how they
set a hundred million dollar revenue goal

cue-796
00:59:00.860 --> 00:59:03.040
and, um, how important that was.

cue-797
00:59:03.900 --> 00:59:04.070
Um,

cue-798
00:59:04.740 --> 00:59:04.960
so

cue-799
00:59:06.740 --> 00:59:07.690
through this, um,

cue-800
00:59:08.720 --> 00:59:11.820
twenty twenty-three investment, the
safeguards that they

cue-801
00:59:12.780 --> 00:59:15.600
thought were in place for the first two
deals were gone,

cue-802
00:59:16.500 --> 00:59:18.820
and that's where the problem arises.

cue-803
00:59:19.740 --> 00:59:20.430
So, um,

cue-804
00:59:21.600 --> 00:59:21.990
you heard

cue-805
00:59:22.960 --> 00:59:24.080
Mr. Nadella saying,

cue-806
00:59:24.820 --> 00:59:28.940
um, "I don't want to be IBM and OpenAI to
be Microsoft."

cue-807
00:59:29.600 --> 00:59:29.800
And,

cue-808
00:59:30.420 --> 00:59:32.360
uh, you'll recall the testimony for
yourself.

cue-809
00:59:33.040 --> 00:59:33.200
Um,

cue-810
00:59:33.960 --> 00:59:39.120
I might characterize it differently than,
uh, Mr. Savitt might or, or, or, or, or my

cue-811
00:59:39.120 --> 00:59:41.760
friends over at Microsoft might. But I
think if you

cue-812
00:59:43.020 --> 00:59:45.980
think back on that testimony, it was
pretty clear that they didn't want to be

cue-813
00:59:45.980 --> 00:59:48.740
left in the dust. That's what they were
concerned about.

cue-814
00:59:49.440 --> 00:59:50.520
Um, and

cue-815
00:59:51.300 --> 00:59:52.450
the other thing that

cue-816
00:59:53.160 --> 00:59:56.400
he was very clear about, and he and I went
back and forth on this. You remember it.

cue-817
00:59:56.400 --> 01:00:00.240
I, I, I really would have liked you to
look at this in exhibit two fifty-nine.

cue-818
01:00:00.240 --> 01:00:04.240
Remember he said we went back and forth
about IP rights, and he said, "Know-how."

cue-819
01:00:04.940 --> 01:00:10.180
He and I went back and forth on that. And
it's clear that they were pushing. He's,

cue-820
01:00:10.180 --> 01:00:14.120
you know, he said, "We need to control our
destiny," you know, and that meant

cue-821
01:00:14.120 --> 01:00:14.640
getting

cue-822
01:00:15.280 --> 01:00:16.820
the rights, getting

cue-823
01:00:17.480 --> 01:00:19.840
what they could. So, um, it

cue-824
01:00:20.500 --> 01:00:22.240
far exceeds what was done before.

cue-825
01:00:23.100 --> 01:00:26.100
And of course, you have the eye-popping
returns that we talked about.

cue-826
01:00:32.520 --> 01:00:32.720
And

cue-827
01:00:33.800 --> 01:00:37.300
the other thing they did is when they
fired Sam Altman, we talked about how that

cue-828
01:00:37.300 --> 01:00:38.500
was a governance failure

cue-829
01:00:39.180 --> 01:00:40.940
for OpenAI.

cue-830
01:00:43.420 --> 01:00:44.100
Microsoft

cue-831
01:00:44.860 --> 01:00:45.260
is not

cue-832
01:00:45.960 --> 01:00:47.200
held to that, that same

cue-833
01:00:47.820 --> 01:00:52.360
standard because they're not-- they don't
have that same, same duty as the OpenAI

cue-834
01:00:52.360 --> 01:00:55.500
directors in terms of the governance of
OpenAI, the nonprofit.

cue-835
01:00:56.180 --> 01:00:59.520
But they certainly helped them,

cue-836
01:01:00.140 --> 01:01:03.720
not just through their commercial terms.
But you heard about what happened the

cue-837
01:01:03.720 --> 01:01:07.992
weekend that Sam Altman was fired.Talked
to Satya Nad-Nadella and Kevin Scott

cue-838
01:01:07.992 --> 01:01:10.672
almost right away. Was on the phone with
them all weekend.

cue-839
01:01:11.352 --> 01:01:14.632
Um, offered them, offered jobs to Altman
and Bachman.

cue-840
01:01:15.752 --> 01:01:17.352
Over the course of the weekend, offered,

cue-841
01:01:18.492 --> 01:01:19.892
with the help of, um,

cue-842
01:01:20.632 --> 01:01:24.032
Mira Murati and, and, and some others,
jobs to other

cue-843
01:01:24.892 --> 01:01:29.411
OpenAI employees. They were basically
saying, you know, "We'll take, we'll take

cue-844
01:01:29.412 --> 01:01:29.772
you all,"

cue-845
01:01:30.512 --> 01:01:34.692
at the cost of about twenty-five billion
dollars. And it's a big expense, and it

cue-846
01:01:34.692 --> 01:01:38.452
was-- It came as a year, at a year that
Microsoft had just laid off a bunch of

cue-847
01:01:38.452 --> 01:01:39.532
people the year before.

cue-848
01:01:40.212 --> 01:01:44.092
So this was a real commitment that they
were making 'cause they didn't wanna see

cue-849
01:01:44.092 --> 01:01:48.812
OpenAI disintegrate. They didn't wanna see
the people get hired by Google or get

cue-850
01:01:48.812 --> 01:01:51.662
hired by, um, any, any other

cue-851
01:01:52.972 --> 01:01:55.832
company that would be developing AI. It's
a big deal.

cue-852
01:01:56.532 --> 01:02:01.252
Very big deal. And in the end, you
remember too, um, I showed him the text

cue-853
01:02:01.252 --> 01:02:07.892
message chain between he and, um, Altman
about directors. There were actually, uh,

cue-854
01:02:07.892 --> 01:02:12.552
Kevin, Kevin Scott was on it, and Bret
Taylor was on for part of it. And they

cue-855
01:02:12.552 --> 01:02:13.172
were, they were

cue-856
01:02:14.112 --> 01:02:18.932
offering names for the nonprofit, and you
heard Professor Schizer talk about that,

cue-857
01:02:18.932 --> 01:02:19.852
talk about, um...

cue-858
01:02:20.452 --> 01:02:21.672
There's one where he says, "No,

cue-859
01:02:22.332 --> 01:02:23.412
uh, strong no."

cue-860
01:02:24.192 --> 01:02:24.972
And they also

cue-861
01:02:25.652 --> 01:02:29.432
produced, uh, certain names, offered up
certain names. And

cue-862
01:02:30.212 --> 01:02:34.452
in the end, when Altman decides, you know,
"This is what I'm willing to do, how

cue-863
01:02:34.452 --> 01:02:38.492
about this board?" And he proposes one of
the people was Larry Summers,

cue-864
01:02:39.412 --> 01:02:40.612
that Nadella didn't know.

cue-865
01:02:41.432 --> 01:02:44.572
And Nadella said, "Can I call him
tonight?" And Altman said, "Yes."

cue-866
01:02:45.292 --> 01:02:45.472
So

cue-867
01:02:46.092 --> 01:02:49.972
the for-profit partner, the commercial
partner in the for-profit

cue-868
01:02:50.692 --> 01:02:51.172
entity,

cue-869
01:02:52.312 --> 01:02:58.552
was exercising control over the nonprofit.
I mean, giving Microsoft the right to

cue-870
01:02:58.552 --> 01:03:02.212
approve or disapprove a director... Now,
we heard testimony from Micro-- or we

cue-871
01:03:02.212 --> 01:03:04.682
heard questions from Micro-Microsoft's
lawyer where he said,

cue-872
01:03:05.392 --> 01:03:11.932
"Oh, don't people call you all the time
for advice on who should be on a board, or

cue-873
01:03:11.932 --> 01:03:16.082
would this person make a good director?"
That may be true. They may, okay? That's

cue-874
01:03:16.082 --> 01:03:18.532
not what this was. That's not what this
was. This was

cue-875
01:03:19.692 --> 01:03:22.832
exercising, you know, control through

cue-876
01:03:23.572 --> 01:03:27.692
participating in that decision, and that
is not something that they can do.

cue-877
01:03:28.312 --> 01:03:30.232
So that is substantial assistance in the
breach.

cue-878
01:03:30.872 --> 01:03:34.092
Doing it in the first place was a breach
on the part of the OpenAI defendants, and

cue-879
01:03:34.092 --> 01:03:37.692
Microsoft provided that substantial
assistance, and it was a substantial

cue-880
01:03:37.692 --> 01:03:39.752
factor. The substantial factor,

cue-881
01:03:40.392 --> 01:03:40.612
um,

cue-882
01:03:41.472 --> 01:03:42.992
in harming Elon was

cue-883
01:03:44.092 --> 01:03:48.272
thirteen billion dollar investment,
expanding the TRAs, the commercialization,

cue-884
01:03:48.272 --> 01:03:52.152
aggressive action after OpenAI board
fired the CEO for lying,

cue-885
01:03:52.772 --> 01:03:59.632
and then control over OpenAI's IP. So
without Microsoft and, and its enormous

cue-886
01:03:59.632 --> 01:04:05.112
resources that it brought to bear in the
situation, and its active engagement, it

cue-887
01:04:05.112 --> 01:04:09.492
wasn't like it wrote a check and said, you
know, "Go, go take care of things." The

cue-888
01:04:09.492 --> 01:04:13.952
active engagement of people at the very
highest levels of the company, the CEO,

cue-889
01:04:13.952 --> 01:04:16.752
the chief technology officer, the chief
financial officer,

cue-890
01:04:17.572 --> 01:04:17.772
um,

cue-891
01:04:18.732 --> 01:04:21.612
they deployed extensive resources,

cue-892
01:04:22.292 --> 01:04:25.872
and without them, Altman and Bachman could
not have looted

cue-893
01:04:26.592 --> 01:04:28.532
the nonprofit's assets as they have

cue-894
01:04:29.172 --> 01:04:33.472
and, uh, built it on the back of this,
quote, "nonprofit thing," as

cue-895
01:04:34.212 --> 01:04:38.252
it was called in that earlier email. So
the evidence showing Microsoft's

cue-896
01:04:38.252 --> 01:04:40.972
substantial role is absolutely clear, and

cue-897
01:04:41.692 --> 01:04:43.512
that's evidence that proves all three of
our claims.

cue-898
01:04:44.552 --> 01:04:45.372
All three of our claims.

cue-899
01:04:45.972 --> 01:04:47.592
Far beyond a preponderance of the
evidence.

cue-900
01:04:48.372 --> 01:04:51.792
Far beyond a preponderance of the
evidence. All we need to prove is by a

cue-901
01:04:51.792 --> 01:04:56.792
preponderance of the evidence, but that
evidence gets you much, much farther.

cue-902
01:04:57.532 --> 01:04:57.712
Um,

cue-903
01:04:58.452 --> 01:05:03.752
they raise a technical defense called the
statute of limitations, right? They're,

cue-904
01:05:03.752 --> 01:05:04.192
they're claiming,

cue-905
01:05:04.932 --> 01:05:07.932
you know, we brought the suit too late.
That Elon brought the suit too late.

cue-906
01:05:09.272 --> 01:05:14.072
The evidence was very clear that Elon
learned of the twenty twenty-three

cue-907
01:05:14.072 --> 01:05:15.252
Mi-Microsoft deal

cue-908
01:05:16.092 --> 01:05:18.252
in Oc-October twenty-second

cue-909
01:05:18.932 --> 01:05:23.511
of twenty twenty-two. How do I know that
specific date? Because there's a text

cue-910
01:05:23.511 --> 01:05:30.492
message from Elon to Sam Altman, and it
contains an article that talks about the

cue-911
01:05:30.492 --> 01:05:35.192
deal. The deal hadn't closed yet. It was
closing in twenty twenty-three. But it

cue-912
01:05:35.192 --> 01:05:39.162
describes the deal, and Elon is saying to
him in this text message,

cue-913
01:05:39.812 --> 01:05:43.912
uh, this is the article, um, which is
Plaintiff's Exhibit two ninety-five. He

cue-914
01:05:43.912 --> 01:05:49.752
actually sends the article to Altman, and
in it, you know, OpenAI is valued then at,

cue-915
01:05:49.752 --> 01:05:54.092
at twenty billion. It's valued at much
more now. Um, and it says, "In advanced

cue-916
01:05:54.092 --> 01:05:57.092
talks with Microsoft for, for..." And Elon
was upset.

cue-917
01:05:57.712 --> 01:05:58.542
He sent, um,

cue-918
01:05:59.712 --> 01:06:04.042
the text message, and, and you, you saw
that. This is where he said,

cue-919
01:06:04.692 --> 01:06:07.672
um, "Seems like a bait and switch. This is
a bait and switch."

cue-920
01:06:08.292 --> 01:06:10.892
And he expresses concern about, um,

cue-921
01:06:11.572 --> 01:06:12.852
Microsoft. Says, um,

cue-922
01:06:14.092 --> 01:06:16.412
"For sure what rights I can..." And
Altman's response,

cue-923
01:06:17.092 --> 01:06:18.152
Altman's response was,

cue-924
01:06:19.232 --> 01:06:23.692
"I agree this feels bad. We offered you
equity when we established the capped

cue-925
01:06:23.692 --> 01:06:28.492
profit, which you didn't want at the time,
but we are still very happy to do that

cue-926
01:06:28.492 --> 01:06:30.572
a-at any time you want." But what was
Elon's reaction?

cue-927
01:06:31.972 --> 01:06:32.792
He said, "Get out of here.

cue-928
01:06:33.672 --> 01:06:35.912
I'm not gonna do that." I mean, and, and
he didn't.

cue-929
01:06:36.972 --> 01:06:40.652
Again, if he wanted to do this as a
for-profit, he could have done it. He

cue-930
01:06:40.652 --> 01:06:43.672
didn't need to do it this way with Sam
Altman, you know,

cue-931
01:06:44.652 --> 01:06:46.752
basically stealing a charity.

cue-932
01:06:47.412 --> 01:06:51.312
And Altman's response wasn't, "Gee, I'm
sorry about this," or, "I thought you

cue-933
01:06:51.312 --> 01:06:55.272
knew," or anything like that. Nothing like
that. His response was, you know,

cue-934
01:06:56.052 --> 01:06:57.432
"We're happy to give you equity."

cue-935
01:06:58.112 --> 01:07:00.362
Elon called it a bribe. You can call it
whatever you want,

cue-936
01:07:00.992 --> 01:07:02.712
but that's what he called it. So

cue-937
01:07:03.512 --> 01:07:04.852
that defense, um,

cue-938
01:07:05.512 --> 01:07:09.852
goes nowhere. They're gonna try to wave
around, uh, emails or

cue-939
01:07:10.472 --> 01:07:13.752
Chrome sheet or whatever it is that
they're gonna try to put up and try and

cue-940
01:07:13.752 --> 01:07:20.616
convince youUm, that somehow or another
Elon knew that

cue-941
01:07:20.616 --> 01:07:23.736
what happened in twenty twenty-three, he
knew that somehow before.

cue-942
01:07:25.476 --> 01:07:26.576
Th-th-the breach occurs

cue-943
01:07:27.736 --> 01:07:31.255
later. It occurs here in twenty
twenty-three in the ways that I described

cue-944
01:07:31.256 --> 01:07:31.476
to you.

cue-945
01:07:32.096 --> 01:07:33.036
So, um,

cue-946
01:07:34.296 --> 01:07:37.016
the statute of limitations defense is not,
is not

cue-947
01:07:39.056 --> 01:07:42.636
supported by, by the evidence, quite
simply. Don't let them try to trick you

cue-948
01:07:42.636 --> 01:07:43.576
into thinking that it is.

cue-949
01:07:44.416 --> 01:07:49.156
I have an opportunity since I'm the
plaintiff to get to speak with you again,

cue-950
01:07:50.016 --> 01:07:50.356
and

cue-951
01:07:51.136 --> 01:07:55.255
I can tell by the looks on your faces
you're all super excited this could happen

cue-952
01:07:55.255 --> 01:07:55.266
[laughs].

cue-953
01:07:56.836 --> 01:07:57.356
But, um,

cue-954
01:07:58.036 --> 01:07:58.576
uh, so

cue-955
01:07:59.336 --> 01:08:00.016
I'm gonna sit down.

cue-956
01:08:06.516 --> 01:08:11.065
Thank you. Few other things when I get
back,

cue-957
01:08:11.716 --> 01:08:12.046
and,

cue-958
01:08:12.936 --> 01:08:17.076
uh, including some of the things that
they're gonna say. But thanks very much,

cue-959
01:08:17.076 --> 01:08:19.196
really appreciate having this with

cue-960
01:08:20.616 --> 01:08:23.436
-you.
-Okay. We're gonna go ahead and stand and

cue-961
01:08:23.436 --> 01:08:26.636
stretch while, uh, Mr. Savitt, um, gets
ready.

cue-962
01:08:31.656 --> 01:08:32.296
Go on in the...

cue-963
01:08:34.596 --> 01:08:38.836
And if you want, you can

cue-964
01:08:38.836 --> 01:08:43.746
stretch.

cue-965
01:08:52.296 --> 01:08:52.876
Yes, thank you.

cue-966
01:08:55.966 --> 01:08:55.966
Oh,

cue-967
01:09:01.756 --> 01:09:07.076
pardon

cue-968
01:09:15.996 --> 01:09:18.156
-me.
-Ms. Eddy, is this a, is this a different

cue-969
01:09:18.156 --> 01:09:18.396
one?

cue-970
01:09:20.316 --> 01:09:23.096
That you gave, you all gave me one
earlier.

cue-971
01:09:24.156 --> 01:09:29.376
-Uh, something [laughs]
-It's, it's, that's, it, it's your size

cue-972
01:09:29.376 --> 01:09:31.666
[laughs]. It's all right.

cue-973
01:09:32.276 --> 01:09:36.796
-You like that a little bit more.
-Charlie. Charlie. Charlie.

cue-974
01:09:41.316 --> 01:09:43.816
That you all, yeah, you all gave me one
this morning.

cue-975
01:09:43.816 --> 01:09:45.976
-It's another way I needed [laughs].
-You need to think about it. That's how you

cue-976
01:09:45.976 --> 01:09:46.196
tell.

cue-977
01:09:46.965 --> 01:09:47.356
No, go.

cue-978
01:09:48.696 --> 01:09:48.976
Okay.

cue-979
01:09:59.016 --> 01:09:59.716
What if it's so

cue-980
01:09:59.716 --> 01:10:07.716
little?

cue-981
01:10:09.816 --> 01:10:18.826
[background chatter]

cue-982
01:10:18.826 --> 01:10:20.096
I

cue-983
01:10:26.596 --> 01:10:27.796
thought it was Mr. Savitt.

cue-984
01:10:28.696 --> 01:10:30.456
-Are you prepared?
-Ready.

cue-985
01:10:30.456 --> 01:10:30.796
Okay.

cue-986
01:10:33.196 --> 01:10:36.196
-Uh, we are back on the record.
-[laughs]

cue-987
01:10:36.196 --> 01:10:39.736
-Ms. Eddy is at the, uh, podium.
-Thank you. Okay.

cue-988
01:10:40.536 --> 01:10:40.776
[background chatter]

cue-989
01:10:41.636 --> 01:10:51.356
Proceed

cue-990
01:10:51.356 --> 01:10:52.066
as soon as you're

cue-991
01:10:52.696 --> 01:10:53.316
ready.

cue-992
01:10:56.696 --> 01:10:58.076
Can people hear me at the microphone?

cue-993
01:10:58.936 --> 01:10:59.196
What?

cue-994
01:10:59.916 --> 01:11:02.616
-No.
-Again, I would pull it down and then flip

cue-995
01:11:02.616 --> 01:11:05.156
it up a little bit so it catches your
voice.

cue-996
01:11:05.156 --> 01:11:06.476
-Like this?
-Yes.

cue-997
01:11:06.476 --> 01:11:07.536
-Okay.
-Thank you.

cue-998
01:11:07.536 --> 01:11:12.196
Um, good morning, members of the jury, and
it is, uh, it's a pleasure to, uh,

cue-999
01:11:12.196 --> 01:11:15.916
finally address you after these several
weeks.

cue-1000
01:11:16.616 --> 01:11:21.516
Um, Elon Musk has tried to persuade you
that

cue-1001
01:11:22.596 --> 01:11:25.346
-his years-
-Hold on. We're getting some feedback. Let

cue-1002
01:11:25.346 --> 01:11:25.346
us...

cue-1003
01:11:27.516 --> 01:11:28.996
-Try to lower the volume.
-Colin, can you

cue-1004
01:11:29.856 --> 01:11:31.816
push that mic down? I don't know what the-

cue-1005
01:11:32.796 --> 01:11:34.176
-It's off already, yeah.
-Okay.

cue-1006
01:11:36.555 --> 01:11:37.556
Is this still echoing?

cue-1007
01:11:41.336 --> 01:11:42.896
-All right, let's try again. Go ahead.
-Okay.

cue-1008
01:11:43.916 --> 01:11:44.596
Can people hear me?

cue-1009
01:11:45.656 --> 01:11:45.856
All right.

cue-1010
01:11:46.956 --> 01:11:50.166
Um, Mr. Musk, he has tried to persuade you

cue-1011
01:11:50.976 --> 01:11:54.396
that his years-ago donations to OpenAI

cue-1012
01:11:55.836 --> 01:11:57.966
came with specific strings attached,

cue-1013
01:11:59.216 --> 01:12:05.976
that these strings were strong enough to
last forever, to tie OpenAI up in knots

cue-1014
01:12:07.036 --> 01:12:08.456
as it tries to pursue its mission,

cue-1015
01:12:09.976 --> 01:12:14.816
and that these strings gave Mr. Musk
perpetual rights over OpenAI.

cue-1016
01:12:16.716 --> 01:12:18.676
But Mr. Musk has come nowhere close

cue-1017
01:12:19.756 --> 01:12:20.676
to making that case.

cue-1018
01:12:21.896 --> 01:12:22.796
Even the people

cue-1019
01:12:23.576 --> 01:12:24.546
who work for him,

cue-1020
01:12:25.296 --> 01:12:26.496
even the mother of his children,

cue-1021
01:12:27.716 --> 01:12:28.746
can't back his story.

cue-1022
01:12:32.356 --> 01:12:34.156
No one in this case,

cue-1023
01:12:34.876 --> 01:12:36.036
other than Elon Musk,

cue-1024
01:12:36.656 --> 01:12:39.286
has testified to commitments or promises

cue-1025
01:12:40.336 --> 01:12:45.076
that OpenAI or Sam Altman or Greg Brockman
made to Mr. Musk.

cue-1026
01:12:46.296 --> 01:12:49.666
No documents corroborate Mr. Musk's story,

cue-1027
01:12:51.216 --> 01:12:54.996
and that's because no commitments or
promises were made.

cue-1028
01:12:55.696 --> 01:13:01.376
No restrictions were placed on Mr. Musk's
donations. There are no real claims here,

cue-1029
01:13:01.376 --> 01:13:02.156
members of the jury.

cue-1030
01:13:04.516 --> 01:13:06.296
And ask yourself,

cue-1031
01:13:07.316 --> 01:13:07.566
if

cue-1032
01:13:08.336 --> 01:13:15.036
what Elon Musk really believed was that
AGI had to be pursued within a nonprofit,

cue-1033
01:13:15.036 --> 01:13:16.256
open-source company,

cue-1034
01:13:17.576 --> 01:13:23.984
don't you think he'd have launched his
ownCompetitor AI company as a nonprofit

cue-1035
01:13:23.984 --> 01:13:27.284
open source company. That is not what he
did.

cue-1036
01:13:28.944 --> 01:13:30.704
He launched it as a for-profit,

cue-1037
01:13:31.384 --> 01:13:34.584
as he does every, every other company that
he's launched.

cue-1038
01:13:39.024 --> 01:13:39.284
But

cue-1039
01:13:41.644 --> 01:13:45.983
common sense and the law and the evidence,
uh, don't, don't feature all that

cue-1040
01:13:45.984 --> 01:13:47.623
prominently in Mr. Musk's case.

cue-1041
01:13:48.524 --> 01:13:53.364
Because a lot of what you've heard from
plaintiff in opening and throughout this

cue-1042
01:13:53.364 --> 01:13:58.644
trial, and now in closing, is sound bites
and irrelevant false accusations.

cue-1043
01:13:59.604 --> 01:14:02.444
Mr. Musk accuses defendants of stealing a
charity.

cue-1044
01:14:03.544 --> 01:14:09.624
The truth is that through the work of Greg
Brockman, Sam Altman, Ilya Sutskever, the

cue-1045
01:14:09.624 --> 01:14:14.744
many other people who have poured their
hearts and souls into OpenAI since Elon

cue-1046
01:14:14.744 --> 01:14:16.824
Musk abandoned it in twenty eighteen,

cue-1047
01:14:17.984 --> 01:14:23.284
the OpenAI Foundation is now probably the
best-resourced nonprofit in the world.

cue-1048
01:14:23.924 --> 01:14:27.024
It has roughly two hundred billion dollars
in assets.

cue-1049
01:14:28.224 --> 01:14:32.904
It operates under the oversight of not
one, but two state attorneys general in

cue-1050
01:14:32.904 --> 01:14:39.904
deploying these assets to help make AGI
turn out well for humanity.

cue-1051
01:14:39.904 --> 01:14:40.804
And the nonprofit

cue-1052
01:14:41.724 --> 01:14:45.624
still controls the for-profit, which is
now, as you heard, a PBC.

cue-1053
01:14:47.624 --> 01:14:49.194
Mr. Musk says--

cue-1054
01:14:50.104 --> 01:14:52.244
Mr. Mullow says that Sam Altman

cue-1055
01:14:53.104 --> 01:14:53.924
can't be trusted.

cue-1056
01:14:54.584 --> 01:15:00.224
And the truth is, as we will talk about,
Mr. Musk is the one whose testimony is

cue-1057
01:15:00.224 --> 01:15:04.924
contradicted by every other witness and by
all the documents.

cue-1058
01:15:06.204 --> 01:15:11.144
And the evidence shows that ninety
percent, an astounding ninety percent

cue-1059
01:15:11.464 --> 01:15:16.064
of OpenAI employees demanded that Mr.
Altman be reinstated after the people who

cue-1060
01:15:16.064 --> 01:15:21.764
removed him couldn't give a coherent
explanation for why they decided he lacked

cue-1061
01:15:21.764 --> 01:15:22.414
candor. Now,

cue-1062
01:15:23.764 --> 01:15:30.304
Mr. Mullow spent a lot of time attacking
Mr. Altman's credibility in his closing,

cue-1063
01:15:30.304 --> 01:15:34.564
based in part on Mr. Mullow's own
questions that he posed to the witness.

cue-1064
01:15:34.564 --> 01:15:37.654
And questions, as you heard from Judge
Gonzalez Rogers,

cue-1065
01:15:38.364 --> 01:15:39.384
are not evidence.

cue-1066
01:15:40.824 --> 01:15:43.784
Mr. Mullow has his questions. He has his
characterizations.

cue-1067
01:15:44.984 --> 01:15:49.164
You need to exercise your own judgment
about who's telling the truth here.

cue-1068
01:15:50.784 --> 01:15:52.944
Mr. Musk, the witness who

cue-1069
01:15:54.184 --> 01:15:55.184
no document,

cue-1070
01:15:56.004 --> 01:15:57.824
no other witness corroborates,

cue-1071
01:15:58.904 --> 01:15:59.764
or Mr. Altman,

cue-1072
01:16:00.544 --> 01:16:05.703
whom every other important witness in the
case corroborates and whom all the

cue-1073
01:16:05.704 --> 01:16:10.983
documents corroborate. You make that
judgment. Now,

cue-1074
01:16:11.704 --> 01:16:15.884
our project, and this is going to be a
joint project, Mr. Sabot and I are both

cue-1075
01:16:15.884 --> 01:16:21.384
going to address you. Our, our project
will be to focus on the facts and on the

cue-1076
01:16:21.384 --> 01:16:21.604
law,

cue-1077
01:16:22.784 --> 01:16:26.483
and to set aside the false accusations
having nothing to do with the rights that

cue-1078
01:16:26.484 --> 01:16:29.624
Elon Musk now claims he has over OpenAI.

cue-1079
01:16:31.084 --> 01:16:35.344
Um, but because Mr. Musk's counsel has
told you for about two hours what you

cue-1080
01:16:35.344 --> 01:16:35.934
should think,

cue-1081
01:16:36.724 --> 01:16:41.504
we're now going to show you what the
evidence demonstrates. We'll examine

cue-1082
01:16:41.504 --> 01:16:45.904
together how the pieces of, of the jigsaw
puzzle come together.

cue-1083
01:16:47.084 --> 01:16:51.324
We do have a lot of ground to cover, and,
and we'll walk through with you just how

cue-1084
01:16:51.324 --> 01:16:53.344
the evidence that you've seen and heard

cue-1085
01:16:53.984 --> 01:16:55.044
over the past few weeks,

cue-1086
01:16:55.924 --> 01:16:59.584
not just from the defendants, but from Mr.
Musk's own witnesses,

cue-1087
01:17:00.544 --> 01:17:03.004
how all that evidence defeats Mr. Musk's
claims.

cue-1088
01:17:04.084 --> 01:17:07.044
And we will not be able to cover every
piece of evidence.

cue-1089
01:17:07.704 --> 01:17:09.724
Um, there's honestly too much of it,

cue-1090
01:17:10.364 --> 01:17:13.024
but we will try to hit the highlights for
you here.

cue-1091
01:17:14.464 --> 01:17:16.624
Uh, then we will turn to the affirmative
defenses,

cue-1092
01:17:17.264 --> 01:17:19.024
uh, the reasons why Mr. Musk

cue-1093
01:17:19.684 --> 01:17:22.944
couldn't proceed with-- couldn't bring a
claim even if he had one.

cue-1094
01:17:23.664 --> 01:17:27.624
And here we're gonna start with the
statute of limitations,

cue-1095
01:17:28.904 --> 01:17:33.084
which is actually the first issue that
you'll need to address when you go back to

cue-1096
01:17:33.084 --> 01:17:33.744
the jury room.

cue-1097
01:17:34.444 --> 01:17:38.324
Um, so you'll wanna keep the timeline in
mind here as we talk through the evidence.

cue-1098
01:17:38.324 --> 01:17:43.424
Mr. Musk left OpenAI in twenty eighteen.

cue-1099
01:17:44.464 --> 01:17:47.584
His last contribution of any kind was in
twenty twenty.

cue-1100
01:17:48.724 --> 01:17:51.444
This case was filed in August twenty
twenty-four.

cue-1101
01:17:52.584 --> 01:17:56.404
The statute of limitations for Mr. Musk's
charitable trust claim

cue-1102
01:17:57.084 --> 01:17:57.904
is three years.

cue-1103
01:17:59.194 --> 01:18:03.584
The statute of limitations for his unjust
enrichment claim is only two years.

cue-1104
01:18:04.944 --> 01:18:05.764
What that means

cue-1105
01:18:06.564 --> 01:18:09.994
is that if the things Mr. Musk is
complaining about today

cue-1106
01:18:11.364 --> 01:18:15.544
are things he could have complained about
before August fifth, twenty twenty-one,

cue-1107
01:18:16.584 --> 01:18:17.244
then he can't win.

cue-1108
01:18:18.844 --> 01:18:21.224
And as we will show, that is exactly the
case.

cue-1109
01:18:22.044 --> 01:18:26.804
Well before August fifth, twenty
twenty-one, Mr. Musk knew or could, with

cue-1110
01:18:26.804 --> 01:18:31.244
reasonable diligence, have discovered the
facts he now says give him a basis for a

cue-1111
01:18:31.244 --> 01:18:31.744
lawsuit.

cue-1112
01:18:36.724 --> 01:18:37.764
Now, as Mr.

cue-1113
01:18:37.824 --> 01:18:40.224
Sabot told you during his opening,

cue-1114
01:18:41.844 --> 01:18:45.424
and as you've now seen for yourselves, in
this case, the statute of limitations is

cue-1115
01:18:45.424 --> 01:18:46.524
not a technical defense.

cue-1116
01:18:47.484 --> 01:18:53.304
It's a textbook case, this one, for why we
have a statute of limitations defense in

cue-1117
01:18:53.304 --> 01:18:58.044
the first place. So hope people don't run
into court years after the events in

cue-1118
01:18:58.044 --> 01:18:58.524
question,

cue-1119
01:18:59.644 --> 01:19:02.653
manufacturing claims at a time that's
convenient to them,

cue-1120
01:19:03.684 --> 01:19:07.124
when memories are faded and evidence is
lost.

cue-1121
01:19:10.764 --> 01:19:14.044
Finally, after the statute of limitations,
we're going to address the other two

cue-1122
01:19:14.044 --> 01:19:18.144
affirmative defenses that we have here.
Um, and we bear the burden on those.

cue-1123
01:19:19.104 --> 01:19:19.264
So

cue-1124
01:19:19.964 --> 01:19:23.164
they only apply if m-- if you find that
Mr. Musk has a claim,

cue-1125
01:19:23.804 --> 01:19:29.936
but we bear the burden if you found that
he did.And those two defenses are latches

cue-1126
01:19:29.936 --> 01:19:35.116
and unclean hands. We'll explain how M-
how Mr. Musk's calculated delay in

cue-1127
01:19:35.116 --> 01:19:36.016
bringing this case

cue-1128
01:19:36.976 --> 01:19:40.136
and his bad conduct related to his claims

cue-1129
01:19:41.136 --> 01:19:43.716
mean that he could not bring a claim even
if he had one,

cue-1130
01:19:45.336 --> 01:19:50.076
which he doesn't. And that brings me to
the first claim, the charitable trust

cue-1131
01:19:50.076 --> 01:19:50.336
claim.

cue-1132
01:19:52.556 --> 01:19:55.936
The, the theory of this claim, as you've
just heard,

cue-1133
01:19:56.976 --> 01:20:01.236
and have heard all, all through this case,
is that when Mr. Musk donated

cue-1134
01:20:01.236 --> 01:20:06.056
thirty-eight million dollars and four
Teslas to op- the OpenAI nonprofit from

cue-1135
01:20:06.056 --> 01:20:07.776
twenty sixteen to twenty twenty,

cue-1136
01:20:08.896 --> 01:20:14.196
he secured commitments from OpenAI and Sam
Altman and Greg Brockman to him

cue-1137
01:20:14.196 --> 01:20:15.656
personally, Mr. Musk personally,

cue-1138
01:20:16.776 --> 01:20:19.536
that the OpenAI nonprofit would stay a
nonprofit

cue-1139
01:20:20.216 --> 01:20:21.896
and would open source its technology.

cue-1140
01:20:23.416 --> 01:20:28.716
You now know that is not what happened.
Mr. Musk didn't leave OpenAI with

cue-1141
01:20:28.716 --> 01:20:30.916
commitments from anyone to do anything.

cue-1142
01:20:33.116 --> 01:20:36.296
And just before we dive into the elements
of the claim,

cue-1143
01:20:37.096 --> 01:20:39.336
I, I wanna just briefly recap

cue-1144
01:20:40.016 --> 01:20:44.236
what the evidence shows about Mr. Musk's
relationship with OpenAI from the early

cue-1145
01:20:44.236 --> 01:20:45.956
years when he was actually there.

cue-1146
01:20:47.936 --> 01:20:49.336
Before OpenAI launched,

cue-1147
01:20:50.336 --> 01:20:54.596
Elon Musk promised to fund it to the tune
of a billion dollars.

cue-1148
01:20:56.096 --> 01:21:00.696
And he thought at first it should be a
for-profit co- corporation. That is what

cue-1149
01:21:00.696 --> 01:21:03.676
the document shows. But it launched as a
nonprofit.

cue-1150
01:21:05.636 --> 01:21:06.756
By twenty seventeen,

cue-1151
01:21:07.496 --> 01:21:14.036
everyone associated with OpenAI realized
that pursuit of its mission would require

cue-1152
01:21:14.036 --> 01:21:16.696
far more money than could be raised by a
nonprofit.

cue-1153
01:21:17.836 --> 01:21:23.756
You heard this from witness upon witness
in the case. You heard it from Elon Musk,

cue-1154
01:21:23.756 --> 01:21:27.476
Sam Altman, Ilya Sutskever, Greg Brockman,
Tasha McCauley.

cue-1155
01:21:28.936 --> 01:21:30.856
This is essentially uncontroverted.

cue-1156
01:21:32.096 --> 01:21:32.976
And at that time,

cue-1157
01:21:33.816 --> 01:21:38.916
in twenty seventeen, Mr. Musk wanted to
turn OpenAI into a for-profit company that

cue-1158
01:21:38.916 --> 01:21:40.436
he could control.

cue-1159
01:21:44.956 --> 01:21:46.036
But the other founders

cue-1160
01:21:47.236 --> 01:21:52.276
refused to turn the keys of AGI over to
one person, let alone Elon Musk.

cue-1161
01:21:53.236 --> 01:21:56.676
And we're gonna see a lot of this document
here today. You've seen a lot of it

cue-1162
01:21:56.676 --> 01:22:00.546
during the trial. This is the Honest
Thoughts email thread from September

cue-1163
01:22:00.546 --> 01:22:01.856
twenty, twenty seventeen,

cue-1164
01:22:02.556 --> 01:22:06.236
when the four initial intense for-profit
negotiations fell apart.

cue-1165
01:22:08.766 --> 01:22:12.216
This is where Greg Brockman and Ilya
Sutskever stood up to Elon Musk.

cue-1166
01:22:13.496 --> 01:22:14.826
And shortly after they did that,

cue-1167
01:22:15.716 --> 01:22:18.816
Mr. Musk sought to merge OpenAI into
Tesla.

cue-1168
01:22:20.716 --> 01:22:25.616
But the other founders rejected that idea
too. They didn't want to be absorbed into

cue-1169
01:22:25.616 --> 01:22:26.706
Mr. Musk's empire.

cue-1170
01:22:29.156 --> 01:22:31.636
So Mr. Musk picked up his marbles.

cue-1171
01:22:32.436 --> 01:22:36.816
He went home, declaring that OpenAI would
have zero chance of success.

cue-1172
01:22:37.596 --> 01:22:38.916
Those were his parting words.

cue-1173
01:22:39.976 --> 01:22:44.196
And w- we're gonna get into much more
detail about these early years, this, this

cue-1174
01:22:44.196 --> 01:22:45.296
early part,

cue-1175
01:22:46.396 --> 01:22:47.836
but that's the basic framework.

cue-1176
01:22:48.776 --> 01:22:52.296
It's what we told you the evidence would
show at the beginning of this case, and

cue-1177
01:22:52.296 --> 01:22:53.916
it's what the evidence has shown.

cue-1178
01:22:55.736 --> 01:22:57.276
What the evidence has not shown

cue-1179
01:22:58.116 --> 01:23:00.076
is any breach of charitable trust.

cue-1180
01:23:02.136 --> 01:23:02.276
Now,

cue-1181
01:23:03.056 --> 01:23:04.036
to prove this claim,

cue-1182
01:23:04.796 --> 01:23:07.935
there were three things that Mr. Musk
needed to establish.

cue-1183
01:23:08.536 --> 01:23:11.356
First, he needed to prove the existence of
a charitable trust.

cue-1184
01:23:12.636 --> 01:23:16.935
Second, he had to prove that any supposed
trust still existed

cue-1185
01:23:17.836 --> 01:23:20.556
by the time of the breach, twenty
twenty-three.

cue-1186
01:23:22.696 --> 01:23:25.616
Third, he had to prove that his donations
were misused.

cue-1187
01:23:27.156 --> 01:23:30.306
He had to prove all three of those things,
and he failed on each one of them.

cue-1188
01:23:32.466 --> 01:23:35.116
So first, let's start with the existence
of the trust.

cue-1189
01:23:35.836 --> 01:23:39.436
Remember Jud- Judge Gonzales Rogers's
instructions.

cue-1190
01:23:40.576 --> 01:23:41.896
To have a charitable trust,

cue-1191
01:23:42.576 --> 01:23:46.996
it's not enough to make an outright
donation to a nonprofit to be used for

cue-1192
01:23:46.996 --> 01:23:49.665
general charitable-- for general purposes.

cue-1193
01:23:51.516 --> 01:23:53.396
Instead, you need to have a donation

cue-1194
01:23:54.176 --> 01:23:56.196
for a purpose that's both charitable

cue-1195
01:23:56.856 --> 01:23:57.816
and specific.

cue-1196
01:23:58.876 --> 01:24:02.396
If you donate money to a hospital, that's
charitable, but it doesn't give you a

cue-1197
01:24:02.396 --> 01:24:03.276
charitable trust.

cue-1198
01:24:04.416 --> 01:24:08.416
If you donate to a hospital specifically
to support medical research on a

cue-1199
01:24:08.416 --> 01:24:09.476
particular disease,

cue-1200
01:24:10.376 --> 01:24:12.796
that might give you a charitable trust. It
might,

cue-1201
01:24:14.196 --> 01:24:14.646
but only

cue-1202
01:24:15.396 --> 01:24:19.516
if you also properly manifested an intent
to create the trust,

cue-1203
01:24:20.196 --> 01:24:22.316
which means you need to have communicated
it,

cue-1204
01:24:23.556 --> 01:24:25.236
communicated an intention

cue-1205
01:24:25.856 --> 01:24:28.056
to impose an enforceable obligation.

cue-1206
01:24:28.736 --> 01:24:30.666
So that's what Mr. Musk needed to prove

cue-1207
01:24:31.416 --> 01:24:34.796
to prove the existence of a charitable
trust. He didn't meet that test.

cue-1208
01:24:36.736 --> 01:24:40.186
First of all, all the documents and all
the testimony

cue-1209
01:24:40.976 --> 01:24:46.156
show that he gave money to OpenAI for
general purposes, general charitable

cue-1210
01:24:46.156 --> 01:24:48.126
purposes, to advance OpenAI's mission,

cue-1211
01:24:49.036 --> 01:24:54.636
not for some specific charitable purpose.
So let's just review his contributions.

cue-1212
01:24:55.556 --> 01:25:00.996
One thing Mr. Musk gave to the OpenAI
nonprofit was about twelve million dollars

cue-1213
01:25:00.996 --> 01:25:06.656
from twenty sixteen to twenty twenty to
help cover rent and building expenses.

cue-1214
01:25:06.656 --> 01:25:12.096
This was the money that Mr. Musk, you
heard, contributed so that OpenAI could,

cue-1215
01:25:12.096 --> 01:25:16.536
uh, turn around and cover Mr. Musk's lease
payments on the building that housed

cue-1216
01:25:16.536 --> 01:25:17.696
OpenAI and Neuralink.

cue-1217
01:25:19.116 --> 01:25:21.456
And these rent and building payments,

cue-1218
01:25:22.616 --> 01:25:27.556
they do not meet the test for a charitable
trust. Some of them are specific

cue-1219
01:25:28.156 --> 01:25:33.628
insofar as they are earmarked for rent.But
that's not a specific charitable purpose,

cue-1220
01:25:33.628 --> 01:25:38.628
right? That's, that's... A specific
charitable purpose is something like

cue-1221
01:25:38.628 --> 01:25:42.848
donating to a hospital to support medical
research on a particular disease.

cue-1222
01:25:43.828 --> 01:25:46.788
It's not rent or salaries or office
supplies.

cue-1223
01:25:47.728 --> 01:25:48.608
Those are for general,

cue-1224
01:25:49.368 --> 01:25:51.568
general use by, by the nonprofit.

cue-1225
01:25:54.108 --> 01:25:54.988
Next, um,

cue-1226
01:25:55.608 --> 01:25:56.228
Mr. Musk

cue-1227
01:25:58.788 --> 01:26:04.588
gave the Teslas, the four Teslas in 2017,
um, and he took tax deductions for those

cue-1228
01:26:05.648 --> 01:26:07.648
as charitable donations to the nonprofit.

cue-1229
01:26:08.308 --> 01:26:12.488
You heard that what they were really for,
um, Musk gave them as bonuses to four

cue-1230
01:26:12.488 --> 01:26:14.488
employees, including Greg Brockman

cue-1231
01:26:15.108 --> 01:26:15.948
and Ilya Sutskever,

cue-1232
01:26:16.708 --> 01:26:21.928
um, right as the conversation about
pivoting to a for-profit OpenAI began.

cue-1233
01:26:23.268 --> 01:26:24.708
And as the negotiations

cue-1234
01:26:25.528 --> 01:26:27.768
over the new OpenAI for-profit

cue-1235
01:26:28.468 --> 01:26:29.248
heated up a bit,

cue-1236
01:26:30.368 --> 01:26:34.408
Mr. Musk rushed the production of those
cars to soften up Bro- Brockman and

cue-1237
01:26:34.408 --> 01:26:38.848
Sutskever, so that they would accept the
terms that he wanted. And you, you

cue-1238
01:26:38.848 --> 01:26:40.128
remember this text message,

cue-1239
01:26:41.068 --> 01:26:42.208
this exchange, uh,

cue-1240
01:26:43.228 --> 01:26:48.428
from, from Dr. Sutskever, "At least we're
getting our Teslas. Will a Model 3 make

cue-1241
01:26:48.428 --> 01:26:52.368
you willing to accept massively
unfavorable terms?" And you'll recall Greg

cue-1242
01:26:52.368 --> 01:26:59.208
Brockman testified that's the day that
Elon Musk proposed 62.5% of the

cue-1243
01:26:59.208 --> 01:27:00.828
new for-profit would belong to him.

cue-1244
01:27:03.648 --> 01:27:08.348
But let, let's say for a moment that the
Teslas count as charitable donations to

cue-1245
01:27:08.348 --> 01:27:08.818
OpenAI.

cue-1246
01:27:10.008 --> 01:27:13.108
Um, they still don't meet the charitable
trust test

cue-1247
01:27:13.728 --> 01:27:18.128
because as with the rent payments, there
was no specific charitable purpose here.

cue-1248
01:27:18.128 --> 01:27:18.648
There was no...

cue-1249
01:27:19.428 --> 01:27:24.008
They, they were to specific employees, not
specific charitable purposes.

cue-1250
01:27:25.448 --> 01:27:30.728
So to the extent the Tesla gifts were
charitable at all, they were for general

cue-1251
01:27:30.728 --> 01:27:31.248
purposes.

cue-1252
01:27:32.548 --> 01:27:34.468
So that leaves the $25 million

cue-1253
01:27:35.128 --> 01:27:39.648
in quarterly donations that Mr. Musk gave
the OpenAI nonprofit starting in

cue-1254
01:27:42.588 --> 01:27:48.128
final payment gets actually gone through
in June 2017. You saw that from Mr. Dudni.

cue-1255
01:27:48.128 --> 01:27:55.128
These don't meet the charitable trust
test either. They were all given

cue-1256
01:27:55.128 --> 01:27:58.768
for OpenAI's general purposes and not for
restricted purposes.

cue-1257
01:27:59.648 --> 01:28:00.507
The documents show it.

cue-1258
01:28:01.768 --> 01:28:07.588
Here are the Form 990s, the tax reports
that OpenAI filed every year, in which it

cue-1259
01:28:07.588 --> 01:28:10.268
had to disclose whether it had any
restricted assets.

cue-1260
01:28:10.948 --> 01:28:12.948
You see those boxes are all blank

cue-1261
01:28:13.568 --> 01:28:16.308
because OpenAI had no restricted assets to
report.

cue-1262
01:28:18.388 --> 01:28:19.388
Here are the letters

cue-1263
01:28:20.028 --> 01:28:24.368
from Mr. Musk's donor-advised funds, the
entities that actually paid the gifts to

cue-1264
01:28:24.368 --> 01:28:24.928
OpenAI.

cue-1265
01:28:25.668 --> 01:28:31.028
These make clear the gifts are for general
and not specific charitable purposes.

cue-1266
01:28:33.468 --> 01:28:35.728
They're to be used for OpenAI.

cue-1267
01:28:38.028 --> 01:28:42.608
Not a single document shows any specific
purpose for any of these quarterly

cue-1268
01:28:42.608 --> 01:28:43.228
donations,

cue-1269
01:28:43.868 --> 01:28:49.448
and Mr. Musk's own witnesses confirm that
there were none. Jared Birchall, the head

cue-1270
01:28:49.448 --> 01:28:52.868
of Mr. Musk's family office and the man in
charge of his charitable giving,

cue-1271
01:28:53.648 --> 01:28:58.228
testified Mr. Musk never told him of any
specific purposes.

cue-1272
01:29:00.508 --> 01:29:04.108
If not the man in charge of his charitable
contributions, then who?

cue-1273
01:29:05.688 --> 01:29:06.468
The answer is no one.

cue-1274
01:29:08.208 --> 01:29:12.808
There were no specific charitable purposes
here, no special commitments to Mr. Musk

cue-1275
01:29:12.808 --> 01:29:18.048
about how his money would be used. Every
one of Mr. Musk's cofounders told you

cue-1276
01:29:18.048 --> 01:29:18.248
that.

cue-1277
01:29:19.008 --> 01:29:23.028
Greg Brockman said it. Sam Altman said it.
Dr. Sutskever said it too.

cue-1278
01:29:24.128 --> 01:29:25.108
Here's his testimony.

cue-1279
01:29:26.728 --> 01:29:33.108
It was here a moment ago. Um, [laughs] he
said, "Do you recall ever conditioning

cue-1280
01:29:33.108 --> 01:29:37.448
donations to OpenAI on the organization
remaining a nonprofit?"

cue-1281
01:29:38.128 --> 01:29:39.888
Answer, "I have no such knowledge."

cue-1282
01:29:41.628 --> 01:29:45.668
Even Musk's own witnesses said it, not
just Jared Birchall,

cue-1283
01:29:46.328 --> 01:29:47.768
but his own senior advisors.

cue-1284
01:29:48.808 --> 01:29:49.848
Here's Siobhan Sylas.

cue-1285
01:29:51.808 --> 01:29:55.488
So you don't recall any restriction placed
on Mr. Musk's donations that required

cue-1286
01:29:55.488 --> 01:29:59.908
OpenAI to remain nonprofit and open
source, do you? Answer, "I don't."

cue-1287
01:30:02.548 --> 01:30:06.847
And S- Sam Teller too, the witness who,
who appeared by deposition, I think

cue-1288
01:30:06.848 --> 01:30:07.368
yesterday.

cue-1289
01:30:08.688 --> 01:30:12.708
As far as you know, did Mr. Musk impose
any conditions on the use of his donations

cue-1290
01:30:12.708 --> 01:30:14.288
to OpenAI? "I don't know."

cue-1291
01:30:15.348 --> 01:30:19.528
Did Mr. Musk ever tell you there were any
conditions on his donations to OpenAI? "I

cue-1292
01:30:19.528 --> 01:30:20.148
don't remember."

cue-1293
01:30:22.448 --> 01:30:23.948
All these witnesses say the same thing.

cue-1294
01:30:24.948 --> 01:30:28.248
No specific commitments were made to Mr.
Musk.

cue-1295
01:30:30.028 --> 01:30:30.188
And

cue-1296
01:30:31.288 --> 01:30:32.808
you know Mr. Musk knows very well,

cue-1297
01:30:33.868 --> 01:30:35.808
by the way, what specific

cue-1298
01:30:36.628 --> 01:30:38.748
terms and conditions on donations look
like

cue-1299
01:30:39.508 --> 01:30:44.068
because you remember this email that we
saw between him and Sam Altman back in

cue-1300
01:30:44.108 --> 01:30:44.868
2016

cue-1301
01:30:47.108 --> 01:30:51.488
asking to see the exact terms and
conditions because gifts are only as good

cue-1302
01:30:51.488 --> 01:30:54.208
as the T and C, the terms and conditions.

cue-1303
01:30:56.818 --> 01:31:01.488
Mr. Musk did not get terms and conditions
on his own donations. He gave them for

cue-1304
01:31:01.488 --> 01:31:04.488
OpenAI's general purposes without
restriction.

cue-1305
01:31:05.868 --> 01:31:06.588
And really,

cue-1306
01:31:07.268 --> 01:31:09.068
this is where the claim should end.

cue-1307
01:31:10.508 --> 01:31:14.568
If the donations came with no
restrictions, with no specific charitable

cue-1308
01:31:14.568 --> 01:31:19.808
purpose, with no strings attached, then
Mr. Musk does not have a charitable trust

cue-1309
01:31:19.808 --> 01:31:20.587
to enforce.

cue-1310
01:31:22.808 --> 01:31:23.608
Now, Mr. Mullowe

cue-1311
01:31:24.288 --> 01:31:27.748
spent a lot of this trial, and a lot of
his closing argument

cue-1312
01:31:29.108 --> 01:31:34.128
focused not on specific charitable
purposes, but on OpenAI's general purposes

cue-1313
01:31:34.128 --> 01:31:40.208
and missionHe asked witness after witness
about the original certificate of

cue-1314
01:31:40.208 --> 01:31:41.948
incorporation for the nonprofit.

cue-1315
01:31:43.388 --> 01:31:48.168
He asked about that blog post introducing
OpenAI, which launched the company.

cue-1316
01:31:49.268 --> 01:31:53.968
He asked about the charter that OpenAI
published in twenty eighteen on its

cue-1317
01:31:53.968 --> 01:31:55.968
website after Mr. Musk left

cue-1318
01:31:56.568 --> 01:31:57.088
the company.

cue-1319
01:31:58.428 --> 01:32:00.728
Mr. Musk-- or Mr. Mullo also showed you

cue-1320
01:32:01.428 --> 01:32:03.948
emails from before OpenAI was founded

cue-1321
01:32:04.988 --> 01:32:07.847
that are talking about what the mission
will be

cue-1322
01:32:08.568 --> 01:32:09.948
and how it'll be structured.

cue-1323
01:32:11.598 --> 01:32:11.738
And

cue-1324
01:32:12.748 --> 01:32:18.588
remember Mr. Mullo asked you in opening as
the very first question that you should

cue-1325
01:32:18.588 --> 01:32:20.088
ask yourselves and answer,

cue-1326
01:32:21.228 --> 01:32:23.468
did OpenAI have a charitable mission

cue-1327
01:32:24.188 --> 01:32:29.768
to operate as a nonprofit to develop safe
AI open source for the good of humanity?

cue-1328
01:32:30.708 --> 01:32:31.528
That was his question.

cue-1329
01:32:32.828 --> 01:32:33.908
That is the wrong question,

cue-1330
01:32:34.608 --> 01:32:37.128
and these are the wrong documents that
he's focused on.

cue-1331
01:32:37.948 --> 01:32:43.408
Because even if OpenAI had somehow
violated its general charitable purposes,

cue-1332
01:32:43.408 --> 01:32:44.328
and we will get to that,

cue-1333
01:32:45.168 --> 01:32:50.028
it did not, it has not violated its
general charitable purposes. But even if

cue-1334
01:32:50.028 --> 01:32:50.538
it had,

cue-1335
01:32:52.468 --> 01:32:54.968
that wouldn't give Elon Musk a right to
sue.

cue-1336
01:32:57.708 --> 01:33:03.228
If a private person wants to sue a
nonprofit, he has to show that he properly

cue-1337
01:33:03.228 --> 01:33:07.408
manifested an intent to further a
specific charitable purpose,

cue-1338
01:33:08.268 --> 01:33:09.348
and there's nothing like that here.

cue-1339
01:33:10.488 --> 01:33:14.608
It is not enough to point to general
mission statements.

cue-1340
01:33:15.908 --> 01:33:17.468
So where does that leave Mr. Musk?

cue-1341
01:33:18.908 --> 01:33:19.048
Um,

cue-1342
01:33:19.728 --> 01:33:19.928
well,

cue-1343
01:33:23.068 --> 01:33:23.288
he

cue-1344
01:33:23.988 --> 01:33:26.688
has to try to argue that you can infer

cue-1345
01:33:27.928 --> 01:33:33.088
an intention to have the donations applied
only to specific purposes rather than for

cue-1346
01:33:33.088 --> 01:33:36.168
OpenAI's general purposes. That has to be
the argument.

cue-1347
01:33:38.308 --> 01:33:38.468
And

cue-1348
01:33:40.688 --> 01:33:42.738
the restrictions, it seems, can be,

cue-1349
01:33:43.368 --> 01:33:45.348
the argument seems to be, can be implied

cue-1350
01:33:47.048 --> 01:33:48.288
not from any contract,

cue-1351
01:33:49.088 --> 01:33:49.568
not from

cue-1352
01:33:50.248 --> 01:33:51.428
specific things written down,

cue-1353
01:33:52.068 --> 01:33:54.848
but from discussions and interactions
generally.

cue-1354
01:33:56.388 --> 01:34:00.788
And to be clear, this is coming from a man
who, who told you

cue-1355
01:34:02.148 --> 01:34:03.088
he's so literal

cue-1356
01:34:04.088 --> 01:34:08.148
you never need to read between the lines
with him. This is what Mr. Musk says about

cue-1357
01:34:08.148 --> 01:34:08.668
himself,

cue-1358
01:34:10.268 --> 01:34:13.368
that if he wants to say something, he says
it clearly.

cue-1359
01:34:14.348 --> 01:34:15.568
You don't have that in this case.

cue-1360
01:34:17.648 --> 01:34:21.648
You have an argument about what can be
implied from course of dealings.

cue-1361
01:34:24.988 --> 01:34:25.138
And

cue-1362
01:34:27.088 --> 01:34:31.408
you have an argument that everything that
happened was between the lines.

cue-1363
01:34:32.588 --> 01:34:36.468
But what's between the lines is really, as
the evidence has shown, it's just blank.

cue-1364
01:34:39.308 --> 01:34:42.908
The, the, the restrictions that are at
issue here

cue-1365
01:34:44.148 --> 01:34:44.928
that seem to be

cue-1366
01:34:47.388 --> 01:34:48.328
what are being advanced

cue-1367
01:34:49.368 --> 01:34:49.888
is first

cue-1368
01:34:51.388 --> 01:34:57.568
that any OpenAI for-profit would only ever
fund the nonprofit, right? Like a museum

cue-1369
01:34:57.568 --> 01:35:02.428
store. Uh, that seems to be the argument,
that that restriction was imposed

cue-1370
01:35:02.428 --> 01:35:07.408
implicitly through the dealings between
Mr. Musk, Mr. Brockman, Mr. Altman, and

cue-1371
01:35:07.408 --> 01:35:07.837
OpenAI.

cue-1372
01:35:08.938 --> 01:35:09.418
And second,

cue-1373
01:35:10.048 --> 01:35:14.528
that OpenAI would open source all of its
technology absent safety concerns. Or as

cue-1374
01:35:14.528 --> 01:35:20.208
we heard today, I think for the first
time, uh, if it's practical and reasonable

cue-1375
01:35:20.208 --> 01:35:22.148
to open source. So that's the new
restriction.

cue-1376
01:35:24.748 --> 01:35:25.688
As we will review,

cue-1377
01:35:27.088 --> 01:35:31.688
the evidence does not remotely show either
of these things. And I'm not just

cue-1378
01:35:31.688 --> 01:35:35.048
talking, by the way, about the direct
evidence, the circumstantial evidence.

cue-1379
01:35:35.048 --> 01:35:36.628
There's nothing here.

cue-1380
01:35:37.728 --> 01:35:39.508
And keep in mind, as we're talking about

cue-1381
01:35:40.608 --> 01:35:42.988
this, these, these implied restrictions,

cue-1382
01:35:44.428 --> 01:35:48.708
Mr. Musk has to show again that he
communicated, that he intended to impose

cue-1383
01:35:48.708 --> 01:35:53.568
an enforceable obligation to devote the
trust property to the specific purposes

cue-1384
01:35:53.568 --> 01:35:54.358
he's now claiming.

cue-1385
01:35:55.458 --> 01:35:57.747
The specific purposes cannot just be in
his head.

cue-1386
01:35:59.448 --> 01:36:01.728
And if the purposes aren't spelled out in
writing,

cue-1387
01:36:02.808 --> 01:36:06.528
then Musk has to prove the existence of
the terms of an oral trust

cue-1388
01:36:07.228 --> 01:36:09.088
by clear and convincing evidence.

cue-1389
01:36:10.488 --> 01:36:14.588
So keep those instructions in mind as
we're reviewing the documents and the

cue-1390
01:36:14.588 --> 01:36:17.288
witness testimony from the early years of
OpenAI.

cue-1391
01:36:18.028 --> 01:36:19.308
But under any standard,

cue-1392
01:36:20.808 --> 01:36:25.248
Mr. Musk comes nowhere near proving these
restrictions he says were somehow mapped

cue-1393
01:36:25.248 --> 01:36:26.228
onto his donations.

cue-1394
01:36:28.588 --> 01:36:29.708
Let's start with the first

cue-1395
01:36:30.368 --> 01:36:32.808
alleged implied restriction, which

cue-1396
01:36:33.788 --> 01:36:38.768
again supposedly limits the size and kind
of for-profit that OpenAI can have.

cue-1397
01:36:39.688 --> 01:36:41.068
And y- you may have noticed

cue-1398
01:36:41.688 --> 01:36:45.028
that I'm str- struggling a bit to
articulate quite what this restriction is

cue-1399
01:36:45.028 --> 01:36:48.768
supposed to be because it has been a
moving target throughout this trial.

cue-1400
01:36:50.388 --> 01:36:50.528
Um,

cue-1401
01:36:51.268 --> 01:36:56.708
early in the trial, Mr. Musk suggested the
only for-profit that OpenAI could create

cue-1402
01:36:56.708 --> 01:37:01.328
was a small adjunct to the nonprofit. In
fact, Mr. Musk said it twice in the space

cue-1403
01:37:01.328 --> 01:37:01.828
of a minute.

cue-1404
01:37:03.348 --> 01:37:05.748
But no document anywhere uses that term.

cue-1405
01:37:06.378 --> 01:37:10.868
No one had ever heard about this small
adjunct before this trial.

cue-1406
01:37:11.788 --> 01:37:13.458
Greg Brockman testified

cue-1407
01:37:14.608 --> 01:37:18.327
about that. I literally never heard those
words really until the beginning of this

cue-1408
01:37:18.328 --> 01:37:18.668
trial.

cue-1409
01:37:19.968 --> 01:37:20.808
Ilya Sutskever

cue-1410
01:37:21.408 --> 01:37:22.008
said the same.

cue-1411
01:37:23.328 --> 01:37:26.528
Do you ever hear him say that any
for-profit must be a small adjunct to the

cue-1412
01:37:26.528 --> 01:37:27.848
nonprofit? No.

cue-1413
01:37:30.228 --> 01:37:34.568
And the phrase, in fact, is such a novelty
that when we used it in questioning Mr.

cue-1414
01:37:34.568 --> 01:37:38.908
Brockman, you'll remember Mr. Mullo stood
up and said, accused us of making up the

cue-1415
01:37:38.908 --> 01:37:41.448
term. The judge had to instruct you

cue-1416
01:37:42.168 --> 01:37:49.136
that, in fact, that wasn't the case.That
in

cue-1417
01:37:49.136 --> 01:37:50.696
fact, this phrase came

cue-1418
01:37:51.456 --> 01:37:53.516
from Mr. Musk. It's the theory of his
case.

cue-1419
01:37:54.156 --> 01:37:56.856
But it's been phrased differently
throughout this trial.

cue-1420
01:37:57.756 --> 01:38:01.876
Um, now the for-profit has to be
subservient to the nonprofit,

cue-1421
01:38:02.556 --> 01:38:04.836
or it can only fund the nonprofit,

cue-1422
01:38:05.616 --> 01:38:09.396
or sometimes it has to have a profit cap.
Those are the things we've heard.

cue-1423
01:38:11.516 --> 01:38:11.676
And

cue-1424
01:38:13.056 --> 01:38:14.856
it doesn't really matter how you frame it
though,

cue-1425
01:38:15.556 --> 01:38:17.656
members of the jury. It is all made up.

cue-1426
01:38:19.376 --> 01:38:20.336
And how do you know that?

cue-1427
01:38:21.636 --> 01:38:24.356
You know it the same way that a diligent
accountant

cue-1428
01:38:25.096 --> 01:38:27.776
decides what taxes are owed. You check the
records.

cue-1429
01:38:28.536 --> 01:38:32.556
The same way that a healthcare provider
decides what medication to prescribe,

cue-1430
01:38:33.296 --> 01:38:35.946
what the history of the patient is, you
check the records.

cue-1431
01:38:37.086 --> 01:38:38.276
So let's look at the records.

cue-1432
01:38:39.076 --> 01:38:43.716
And let's start with Mr. Musk's own
statements around the time that OpenAI was

cue-1433
01:38:43.716 --> 01:38:48.436
formed and in its early years, even
before the for-profit negotiations began.

cue-1434
01:38:49.955 --> 01:38:50.815
You will recall

cue-1435
01:38:51.636 --> 01:38:57.356
that Sam Altman and Greg Brockman had the
idea to set up the new AI lab as a

cue-1436
01:38:57.356 --> 01:39:03.636
nonprofit. It, it was going to be a
nonprofit arm of Y Combinator. That was

cue-1437
01:39:03.636 --> 01:39:04.876
Mr. Brockman's testimony.

cue-1438
01:39:05.656 --> 01:39:05.736
And

cue-1439
01:39:06.976 --> 01:39:07.516
that's where

cue-1440
01:39:08.196 --> 01:39:10.096
Sam Altman was president at the time.

cue-1441
01:39:11.236 --> 01:39:14.796
And when he heard about this structure,
Mr. Musk thought it was not optimal,

cue-1442
01:39:15.916 --> 01:39:17.096
that a better structure

cue-1443
01:39:17.776 --> 01:39:21.316
would be a C corp with a parallel
nonprofit.

cue-1444
01:39:21.916 --> 01:39:23.296
Not a small, small adjunct

cue-1445
01:39:24.016 --> 01:39:29.216
for-profit, not an engine to fund the
nonprofit, but a standard for-profit with

cue-1446
01:39:29.216 --> 01:39:30.716
a parallel nonprofit.

cue-1447
01:39:32.136 --> 01:39:37.646
And now Mr. Musk did ultimately agree with
Sam Altman that a pure play

cue-1448
01:39:38.296 --> 01:39:44.036
nonprofit structure made sense to start.
And we've talked about this email.

cue-1449
01:39:45.076 --> 01:39:47.636
Mr. Molla showed it to you again earlier
today.

cue-1450
01:39:48.676 --> 01:39:49.776
He didn't show you the one

cue-1451
01:39:50.556 --> 01:39:56.076
a year later when Mr. Musk doubted whether
a nonprofit had been the right call.

cue-1452
01:39:58.436 --> 01:40:03.116
"I'm concerned," he writes, "that OpenAI
is not on a path to catch up. Setting it

cue-1453
01:40:03.116 --> 01:40:06.216
up as a nonprofit might, in hindsight,
have been the wrong move."

cue-1454
01:40:10.156 --> 01:40:11.936
And he didn't show you the texts

cue-1455
01:40:12.656 --> 01:40:13.756
from 2017

cue-1456
01:40:14.656 --> 01:40:15.396
where Mr. Musk,

cue-1457
01:40:16.236 --> 01:40:20.256
um, Mr. Brockman's reporting to Siobhan
Zillis a conversation he had with Mr.

cue-1458
01:40:20.256 --> 01:40:20.616
Musk,

cue-1459
01:40:21.396 --> 01:40:24.576
where Mr. Musk said he thought a change
from the nonprofit structure

cue-1460
01:40:25.396 --> 01:40:25.856
was needed.

cue-1461
01:40:28.036 --> 01:40:30.736
Now, there is not a word in any of these
documents

cue-1462
01:40:31.856 --> 01:40:37.455
or any others from Mr. Musk's time with
OpenAI about any restrictions on the kind

cue-1463
01:40:37.456 --> 01:40:39.485
of for-profit that OpenAI could consider.

cue-1464
01:40:40.816 --> 01:40:44.736
Just a repeated thought from Mr. Musk that
maybe a for-profit

cue-1465
01:40:44.796 --> 01:40:45.536
would work better.

cue-1466
01:40:47.876 --> 01:40:52.496
Then let's look at what the records do
show about a potential for-profit.

cue-1467
01:40:53.536 --> 01:40:57.056
And that brings us to 2017. In mid-2017,

cue-1468
01:40:58.216 --> 01:41:02.596
as they were working on the Dota 1v1
project that you heard about, that video

cue-1469
01:41:02.596 --> 01:41:02.836
game,

cue-1470
01:41:03.716 --> 01:41:08.876
Greg Brockman and Ilya Sutskever started
realizing that way, way more money was

cue-1471
01:41:08.876 --> 01:41:15.816
gonna be needed to further the mission of
building beneficial AGI because way,

cue-1472
01:41:15.816 --> 01:41:18.916
way more compute, this computing power,

cue-1473
01:41:19.576 --> 01:41:22.876
was going to be needed than was initially
contemplated.

cue-1474
01:41:24.056 --> 01:41:24.936
And everyone,

cue-1475
01:41:25.896 --> 01:41:31.236
everyone including Mr. Musk, agreed that
donations to a nonprofit would not suffice

cue-1476
01:41:31.236 --> 01:41:32.316
to fund the mission.

cue-1477
01:41:33.316 --> 01:41:35.376
More was needed, billions were needed,

cue-1478
01:41:36.216 --> 01:41:38.536
and it would require a pivot to a
for-profit,

cue-1479
01:41:40.136 --> 01:41:41.176
at least in part.

cue-1480
01:41:42.436 --> 01:41:45.636
And here's Mr. Musk in July 2017
suggesting

cue-1481
01:41:46.776 --> 01:41:51.336
it's time to change course. He says,
"Maybe another reason to change course."

cue-1482
01:41:51.336 --> 01:41:53.396
You saw this email after, um,

cue-1483
01:41:54.036 --> 01:41:56.056
i- in connection with him circulating
that,

cue-1484
01:41:56.696 --> 01:42:01.576
uh, article about China's advances in AI,
and he writes to his co-founders.

cue-1485
01:42:02.276 --> 01:42:04.496
He also writes, "I have a tentative game
plan."

cue-1486
01:42:06.256 --> 01:42:08.466
And you heard testimony from Mr. Brockman
about

cue-1487
01:42:09.176 --> 01:42:11.356
what that tentative game plan turned out
to be.

cue-1488
01:42:12.416 --> 01:42:14.416
He met up with Mr. Musk after this email

cue-1489
01:42:15.176 --> 01:42:16.236
to discuss the plan,

cue-1490
01:42:16.956 --> 01:42:22.436
and the idea that Mr. Musk shared was they
would change to a for-profit once OpenAI

cue-1491
01:42:22.436 --> 01:42:23.196
had a big win.

cue-1492
01:42:25.076 --> 01:42:26.576
Then they did have that big win.

cue-1493
01:42:27.356 --> 01:42:28.256
About three weeks later,

cue-1494
01:42:29.356 --> 01:42:34.336
OpenAI's Dota 2 technology beat the
world's best human at the 1v1 version of

cue-1495
01:42:34.336 --> 01:42:34.706
the game,

cue-1496
01:42:35.556 --> 01:42:38.576
and the win was a huge milestone for
OpenAI.

cue-1497
01:42:40.756 --> 01:42:41.416
The next day,

cue-1498
01:42:42.056 --> 01:42:42.556
Brockman,

cue-1499
01:42:43.456 --> 01:42:43.956
Sutskever,

cue-1500
01:42:44.656 --> 01:42:45.536
Siobhan Zillis,

cue-1501
01:42:46.736 --> 01:42:50.475
they were all at the tournament, the Dota
tournament in Seattle. They flew back to

cue-1502
01:42:50.476 --> 01:42:54.456
the Bay Area for a meeting at the haunted
mansion that Mr. Musk had just bought.

cue-1503
01:42:55.796 --> 01:42:56.836
And for Mr. Musk,

cue-1504
01:42:57.856 --> 01:43:01.516
the Dota win was the triggering event.
This was the triggering event.

cue-1505
01:43:03.616 --> 01:43:07.836
You heard what was discussed at the
meeting on August 12, 2017. Now, Mr. Musk

cue-1506
01:43:07.836 --> 01:43:09.286
claimed he couldn't remember,

cue-1507
01:43:11.356 --> 01:43:15.766
but as he apparently did not realize until
this trial, Siobhan Zillis was taking

cue-1508
01:43:15.766 --> 01:43:16.176
notes,

cue-1509
01:43:17.436 --> 01:43:19.426
and those notes tell you what you need to
know.

cue-1510
01:43:20.986 --> 01:43:25.556
"Switch to for-profit in next couple of
weeks. Whoa, fast."

cue-1511
01:43:26.976 --> 01:43:31.216
A little further down, "Keep a subset of
people at OpenAI nonprofit."

cue-1512
01:43:33.416 --> 01:43:35.716
The bulk of the people were gonna go to
the for-profit.

cue-1513
01:43:37.516 --> 01:43:42.236
Now, from August 12th, 2017, and in, uh,
over the next six weeks or so,

cue-1514
01:43:43.096 --> 01:43:47.376
Mr. Musk and his co-founders engaged in
what Greg Brockman and Ilya Sutskever

cue-1515
01:43:47.376 --> 01:43:50.516
described as the most high-stakes

cue-1516
01:43:51.476 --> 01:43:57.628
negotiation they'd ever had.And you recall
Mr. Brockman walked you through the ups

cue-1517
01:43:57.628 --> 01:44:03.007
and downs of those six intense weeks, the
negotiations with Mr. Musk over equity

cue-1518
01:44:03.008 --> 01:44:09.288
splits and control and who was going to be
CEO, and the go-forward corporate

cue-1519
01:44:09.288 --> 01:44:09.808
structure.

cue-1520
01:44:12.228 --> 01:44:15.828
These were not discussions about a small
for-profit

cue-1521
01:44:16.588 --> 01:44:21.988
or a for-profit that would exist simply to
fund the nonprofit, like a museum store.

cue-1522
01:44:21.988 --> 01:44:24.778
They were about a switch to a for-profit.

cue-1523
01:44:27.007 --> 01:44:27.437
You saw

cue-1524
01:44:28.088 --> 01:44:28.788
the documents,

cue-1525
01:44:29.928 --> 01:44:30.657
we can move to them,

cue-1526
01:44:32.108 --> 01:44:34.728
uh, between Mr. Musk's own staff

cue-1527
01:44:35.628 --> 01:44:37.708
discussing the idea of a complete
conversion

cue-1528
01:44:38.328 --> 01:44:39.948
in which the nonprofit would be shut down.

cue-1529
01:44:42.528 --> 01:44:43.288
No, no matter

cue-1530
01:44:44.608 --> 01:44:49.928
what Mr. M-Musk and Mr. Mollo tell you,
this was one of the options that was under

cue-1531
01:44:49.928 --> 01:44:51.337
serious consideration.

cue-1532
01:44:52.018 --> 01:44:53.628
The documents tell the truth here.

cue-1533
01:44:55.728 --> 01:44:58.288
Here's, here's one of the documents we
were looking at.

cue-1534
01:44:59.188 --> 01:45:01.508
Uh, this is DX664.

cue-1535
01:45:02.388 --> 01:45:06.828
The two options on the table for the
operating structure. The first one, roll

cue-1536
01:45:06.828 --> 01:45:11.568
everything into a B corp, would include
all of OpenAI. B corp is a for-profit

cue-1537
01:45:11.568 --> 01:45:12.668
corporation, as you heard.

cue-1538
01:45:15.488 --> 01:45:20.788
Ms. Zilis acknowledged under questioning
that converting from a nor- a nonprofit

cue-1539
01:45:20.788 --> 01:45:22.468
was one of the options being considered.

cue-1540
01:45:24.388 --> 01:45:26.287
Sam Teller told you the same thing.

cue-1541
01:45:29.168 --> 01:45:29.328
And

cue-1542
01:45:31.088 --> 01:45:33.208
there was another option being considered.

cue-1543
01:45:34.408 --> 01:45:38.268
This was the parallel C corp while keeping
the nonprofit in place.

cue-1544
01:45:39.308 --> 01:45:40.168
But in this structure,

cue-1545
01:45:40.888 --> 01:45:46.308
there's no hint that the for-profit would
be subservient in these internal documents

cue-1546
01:45:46.308 --> 01:45:47.908
among Mr. Musk and his staff.

cue-1547
01:45:49.708 --> 01:45:53.628
These are more of Siobhan Zilis'
contemporaneous notes. They show the

cue-1548
01:45:53.628 --> 01:45:57.448
alternative of keeping the nonprofit would
have had the main operations,

cue-1549
01:45:58.088 --> 01:46:03.807
see at the bottom, cor-- .com, core
research and applications. .com is the

cue-1550
01:46:03.808 --> 01:46:04.587
for-profit.

cue-1551
01:46:05.788 --> 01:46:10.128
So the research, the core research and
applications would go to the for-profit.

cue-1552
01:46:11.068 --> 01:46:13.448
Most of the employees would go to the
for-profit.

cue-1553
01:46:15.868 --> 01:46:18.988
And see the note at the top of this
summary from Ms. Zilis,

cue-1554
01:46:19.708 --> 01:46:21.408
file for C corp immediately.

cue-1555
01:46:22.708 --> 01:46:25.988
You heard Mr. Birchall testify that he
actually went ahead and did that.

cue-1556
01:46:29.248 --> 01:46:33.508
And you recall what Greg Brockman told you
about his discussions with Mr. Musk in

cue-1557
01:46:33.508 --> 01:46:34.268
this time period.

cue-1558
01:46:34.908 --> 01:46:38.268
They were planning to have the
intellectual property, the, the assets of

cue-1559
01:46:38.268 --> 01:46:40.588
the operation, move to the for-profit,

cue-1560
01:46:42.888 --> 01:46:43.898
and the employees as well.

cue-1561
01:46:47.288 --> 01:46:52.148
These founders were negotiating equity
splits and control provisions, who would

cue-1562
01:46:52.148 --> 01:46:53.008
be CEO.

cue-1563
01:46:53.908 --> 01:46:58.688
The negotiations were intense. This was
not about a subservient for-profit arm. It

cue-1564
01:46:58.688 --> 01:47:02.908
was about the core structure in which
OpenAI would pursue its mission going

cue-1565
01:47:02.908 --> 01:47:03.388
forward.

cue-1566
01:47:07.208 --> 01:47:08.708
And as witness after witness told you,

cue-1567
01:47:09.308 --> 01:47:10.248
OpenAI's mission

cue-1568
01:47:11.248 --> 01:47:17.568
wasn't necessarily tied to its structure
as a nonprofit. This is Ilya

cue-1569
01:47:17.568 --> 01:47:22.688
Sutskever. The mission, he says, the
mission of OpenAI is larger than a

cue-1570
01:47:22.688 --> 01:47:28.588
structure. The mission was and remains to
ensure that AGI benefits all of humanity.

cue-1571
01:47:31.548 --> 01:47:32.948
You can't build AGI

cue-1572
01:47:33.788 --> 01:47:35.488
or distribute its benefits

cue-1573
01:47:36.848 --> 01:47:38.708
unless you have the resources to do so.

cue-1574
01:47:39.508 --> 01:47:42.588
You've heard that from every witness,
virtually, in this case.

cue-1575
01:47:43.388 --> 01:47:47.468
And if you can't get the research or the
resources within a pure nonprofit

cue-1576
01:47:47.468 --> 01:47:47.948
structure,

cue-1577
01:47:48.648 --> 01:47:50.208
then the mission collapses.

cue-1578
01:47:51.168 --> 01:47:52.408
You have to make that change.

cue-1579
01:47:53.628 --> 01:47:58.888
That's what all the founders, including
Mr. Musk, understood as early as 2017.

cue-1580
01:47:59.888 --> 01:48:00.028
Um,

cue-1581
01:48:00.708 --> 01:48:01.768
now Mr. Mollo

cue-1582
01:48:02.428 --> 01:48:05.968
showed you-- I'm gonna show how
un-tech-savvy I am right now, but Mr.

cue-1583
01:48:05.968 --> 01:48:06.928
Mollo showed you that--

cue-1584
01:48:07.568 --> 01:48:10.288
the snippet from Mr. Brockman's journal

cue-1585
01:48:11.148 --> 01:48:15.308
where he refers to the philanthropic
endeavor, and I'm gonna hold up the slide

cue-1586
01:48:16.028 --> 01:48:16.588
that

cue-1587
01:48:17.648 --> 01:48:19.268
Mr. uh, Mollo showed you,

cue-1588
01:48:19.928 --> 01:48:24.528
the snippet that says, "Elon wanted OpenAI
to remain essentially philanthropic

cue-1589
01:48:24.528 --> 01:48:26.867
endeavor."

cue-1590
01:48:27.368 --> 01:48:29.748
Um, you heard Mr. Brockman--

cue-1591
01:48:30.528 --> 01:48:34.068
This is a snippet from, from one of Mr.
Brockman's journal entries. And you heard

cue-1592
01:48:34.068 --> 01:48:37.688
Mr. Brockman des-describe that
conversation with Mr. Musk.

cue-1593
01:48:39.828 --> 01:48:43.448
He, he was taking notes, right, of what
Mr., Mr. Musk said.

cue-1594
01:48:44.308 --> 01:48:45.428
And Mr. Brockman explained

cue-1595
01:48:46.288 --> 01:48:46.828
that

cue-1596
01:48:47.568 --> 01:48:52.038
what Mr. Mollo's selective quoting the
language from the, from the journal

cue-1597
01:48:52.708 --> 01:48:56.868
was distorting, which was the
philanthropic endeavor

cue-1598
01:48:56.928 --> 01:49:01.107
was about the re- the remaining nonprofit.
So it was in this parallel structure. He

cue-1599
01:49:01.108 --> 01:49:02.728
was talking about philanthropic endeavor,

cue-1600
01:49:03.628 --> 01:49:09.008
not the for-profit that they were
creating. And if you read back in the jury

cue-1601
01:49:09.008 --> 01:49:15.478
room, this is PX one fifty-four, the
actual entry. You'll see it reads,

cue-1602
01:49:15.478 --> 01:49:20.928
"To something which is essentially
philanthropic endeavor and is B corp or C

cue-1603
01:49:20.928 --> 01:49:24.268
corp or something." That's the part that
was left off this slide.

cue-1604
01:49:30.368 --> 01:49:35.768
Perhaps the clearest evidence that the
2017 discussions were not about a, a

cue-1605
01:49:35.768 --> 01:49:38.008
vehicle just to fund the nonprofit

cue-1606
01:49:39.008 --> 01:49:43.928
is the document that we've seen a lot in
this trial and already in my remarks to

cue-1607
01:49:43.928 --> 01:49:46.068
you, the honest thoughts email chain.

cue-1608
01:49:49.028 --> 01:49:53.348
Um, and here again, this is the document
that essentially ended the six weeks of

cue-1609
01:49:53.348 --> 01:49:59.708
intense negotiations over the for-profit
structure.Greg Brockman and Ilya Sutskever

cue-1610
01:49:59.708 --> 01:50:00.748
dra- drafted it up.

cue-1611
01:50:01.428 --> 01:50:05.128
They previewed it for Mr. Altman, then
sent it to Mr. Musk and Mr. Altman as

cue-1612
01:50:05.128 --> 01:50:07.488
well. And you see

cue-1613
01:50:08.488 --> 01:50:08.828
this

cue-1614
01:50:09.548 --> 01:50:14.347
in the, in the slide and in the document.
The current structure provides you with a

cue-1615
01:50:14.348 --> 01:50:18.108
path where you end up with unilateral
absolute control over the AGI.

cue-1616
01:50:18.838 --> 01:50:21.568
You stated that you don't want to control
the final AGI,

cue-1617
01:50:22.348 --> 01:50:23.678
but during this negotiation,

cue-1618
01:50:24.508 --> 01:50:28.228
you've shown us that absolute control is
extremely important to you.

cue-1619
01:50:29.488 --> 01:50:32.908
Now, a-as you're considering this
document, ask yourselves,

cue-1620
01:50:33.748 --> 01:50:38.008
are these guys talking about the risk of
this structure creating

cue-1621
01:50:39.048 --> 01:50:43.088
an AGI dictata- dictatorship because
they're thinking about a small structure

cue-1622
01:50:43.088 --> 01:50:47.318
just to fund an existing nonprofit,
something like a museum store?

cue-1623
01:50:48.248 --> 01:50:53.428
No. Why, why would Mr. Musk care about who
controls the for-profit

cue-1624
01:50:54.108 --> 01:50:56.688
if the nonprofit was the one that was
gonna be in control?

cue-1625
01:50:58.128 --> 01:51:02.888
Why would the configuration of the
for-profit matter if the for-profit itself

cue-1626
01:51:02.888 --> 01:51:07.508
was just going to be a small adjunct to
an existing nonprofit?

cue-1627
01:51:08.588 --> 01:51:12.188
Why the fear about unilateral control by
Elon Musk?

cue-1628
01:51:13.448 --> 01:51:18.048
And why would any of this matter if what
we were talking about was a museum store?

cue-1629
01:51:19.728 --> 01:51:21.008
And the answer is that it wouldn't.

cue-1630
01:51:21.738 --> 01:51:23.478
That's, that's not what they were
discussing.

cue-1631
01:51:24.688 --> 01:51:28.468
What they were discussing was a structure
that would house the operations for

cue-1632
01:51:28.468 --> 01:51:32.188
building beneficial AGI and deciding how
that AGI was used.

cue-1633
01:51:33.828 --> 01:51:39.348
And the reason the discussions fell apart
was that Elon Musk was demanding

cue-1634
01:51:39.348 --> 01:51:40.708
unilateral control.

cue-1635
01:51:43.208 --> 01:51:45.988
Now, Mr. Musk says that that's not what
happened.

cue-1636
01:51:46.988 --> 01:51:49.748
He says he just wanted short-term control

cue-1637
01:51:51.148 --> 01:51:54.868
because he was starting to suspect that
his co-founders were up to no good.

cue-1638
01:51:56.148 --> 01:51:58.928
But the evidence here again crushes that
argument

cue-1639
01:51:59.748 --> 01:52:00.648
and that testimony.

cue-1640
01:52:02.608 --> 01:52:05.168
Uh, Mr. Mulloway kept showing you this
document during the trial

cue-1641
01:52:05.908 --> 01:52:09.588
in which Mr. Musk says he wants four out
of seven board seats,

cue-1642
01:52:10.268 --> 01:52:12.888
so he'll unequivocally have initial
control,

cue-1643
01:52:13.628 --> 01:52:15.468
but this will change quickly.

cue-1644
01:52:16.308 --> 01:52:19.588
And Mr. Musk says here the rough target
would be to grow the board to twelve

cue-1645
01:52:19.588 --> 01:52:20.768
people or even sixteen.

cue-1646
01:52:22.008 --> 01:52:26.588
But let's pause for a moment on this
document and consider it in the context of

cue-1647
01:52:26.588 --> 01:52:30.048
the other evidence, including the honest
thoughts email chain that we were just

cue-1648
01:52:30.048 --> 01:52:30.548
looking at.

cue-1649
01:52:32.028 --> 01:52:33.548
You can see from the face of this

cue-1650
01:52:34.508 --> 01:52:35.928
that Mr. Musk is offering

cue-1651
01:52:36.628 --> 01:52:38.767
nothing other than his word

cue-1652
01:52:38.768 --> 01:52:41.268
about the eventual dilution of his
control.

cue-1653
01:52:42.348 --> 01:52:46.948
He'll have a majority of the board and
unequivocal initial control, which means

cue-1654
01:52:46.948 --> 01:52:53.068
he'll be the one deciding whether and when
new board members come on.

cue-1655
01:52:54.288 --> 01:52:59.608
And you heard from all three of Mr. Musk's
co-founders that Mr. Musk provided no

cue-1656
01:52:59.608 --> 01:53:03.408
guarantee, no certainty that this control
would diminish

cue-1657
01:53:04.368 --> 01:53:05.368
as he had suggested.

cue-1658
01:53:06.658 --> 01:53:11.368
You heard the testimony that Mr. Musk
discussed his children inheriting the

cue-1659
01:53:11.368 --> 01:53:13.878
company, inheriting control of the
company.

cue-1660
01:53:15.168 --> 01:53:17.408
He wanted dominion over AGI.

cue-1661
01:53:18.648 --> 01:53:23.048
That's why this was such a high-stakes
conversation. Mr. Musk wanted total

cue-1662
01:53:23.048 --> 01:53:26.148
control. Maybe, maybe he'd give it up over
time,

cue-1663
01:53:26.828 --> 01:53:27.488
or maybe not,

cue-1664
01:53:28.988 --> 01:53:31.048
but it was up to him, and that was the
problem.

cue-1665
01:53:32.988 --> 01:53:38.228
There is not a scrap of evidence,
moreover, to support Mr. Musk's testimony

cue-1666
01:53:38.228 --> 01:53:42.408
that the reason he wanted control was
because he doubted his co-founders' good

cue-1667
01:53:42.408 --> 01:53:43.068
intentions.

cue-1668
01:53:44.288 --> 01:53:48.648
Just look at Mr. Musk's own words at the
time before it became convenient to spin

cue-1669
01:53:48.648 --> 01:53:49.548
this different story.

cue-1670
01:53:50.268 --> 01:53:53.908
They're at the end of this email, the
same, same email that we were looking at,

cue-1671
01:53:53.908 --> 01:53:55.188
PX-156.

cue-1672
01:53:56.708 --> 01:53:58.648
As a closing note, he writes,

cue-1673
01:53:59.388 --> 01:54:03.177
"I've been really impressed with the
quality of discussion with you guys on the

cue-1674
01:54:03.177 --> 01:54:04.628
equity and board stuff.

cue-1675
01:54:05.278 --> 01:54:07.188
I have a really good feeling about this."

cue-1676
01:54:12.568 --> 01:54:16.927
The truth, um, as the evidence shows,

cue-1677
01:54:17.668 --> 01:54:18.948
is that in 2017,

cue-1678
01:54:20.088 --> 01:54:21.918
Mr. Musk wanted a for-profit OpenAI,

cue-1679
01:54:23.008 --> 01:54:29.528
and he wanted to dominate it. The idea
that he was pushing for a nonprofit in

cue-1680
01:54:29.528 --> 01:54:34.128
control or for a small adjunct, a
subservient for-profit, that is all made

cue-1681
01:54:34.128 --> 01:54:34.268
up.

cue-1682
01:54:36.208 --> 01:54:37.888
Uh, here is another piece that is made up,

cue-1683
01:54:39.138 --> 01:54:41.848
the notion that anyone, much less Mr.
Musk,

cue-1684
01:54:42.548 --> 01:54:44.868
even discussed profit caps in 2017.

cue-1685
01:54:46.068 --> 01:54:46.548
No one did.

cue-1686
01:54:47.978 --> 01:54:52.788
That was not part of the discussion with
Mr. Musk about the new for-profit, and it

cue-1687
01:54:52.788 --> 01:54:57.108
certainly was never something that Elon
Musk said should be required.

cue-1688
01:54:58.068 --> 01:54:59.988
Uh, Greg Brockman, you heard testify,

cue-1689
01:55:00.968 --> 01:55:05.668
the first time he discussed the idea of
profit cap was with Josh Akiam

cue-1690
01:55:06.548 --> 01:55:07.908
in January 2018.

cue-1691
01:55:09.428 --> 01:55:10.048
And you remember,

cue-1692
01:55:10.768 --> 01:55:12.968
uh, Mr. Akiam recalled that as well.

cue-1693
01:55:14.508 --> 01:55:19.728
January 2018 is after the intense
negotiations with Musk over creating a new

cue-1694
01:55:19.728 --> 01:55:21.588
for-profit structure had already ended.

cue-1695
01:55:23.728 --> 01:55:29.377
No document anywhere in the record from
2017 shows a discussion of for-profit, uh,

cue-1696
01:55:29.377 --> 01:55:31.808
uh, sorry, of profit caps with, with Mr.
Musk.

cue-1697
01:55:32.408 --> 01:55:32.948
It didn't happen.

cue-1698
01:55:35.868 --> 01:55:36.468
Um,

cue-1699
01:55:37.268 --> 01:55:38.088
and let me address

cue-1700
01:55:39.008 --> 01:55:44.248
another thing made up for this litigation.
This is the notion that the only

cue-1701
01:55:44.248 --> 01:55:48.208
for-profit that Mr. Musk would support and
accept was one, again, where the

cue-1702
01:55:48.208 --> 01:55:53.968
nonprofit, or sorry, the, the profits from
the for-profit would all flow to the

cue-1703
01:55:53.968 --> 01:55:54.728
nonprofit.

cue-1704
01:55:56.428 --> 01:55:57.448
That makes no sense.

cue-1705
01:55:58.628 --> 01:56:04.268
If all the company's profits are flowing
to a nonprofit, then the company is not a

cue-1706
01:56:04.268 --> 01:56:10.918
for-profit, it's a nonprofit.That's what
a, a nonprofit is. And, and you heard

cue-1707
01:56:10.918 --> 01:56:17.658
Professor Hemel talk about this museum
store example. That's how a museum store

cue-1708
01:56:17.658 --> 01:56:22.578
typically operates. It sells souvenirs,
proceeds from those sales go to fund the

cue-1709
01:56:22.578 --> 01:56:24.018
nonprofit's operations.

cue-1710
01:56:24.658 --> 01:56:28.518
The store is not a for-profit. It doesn't
have investors.

cue-1711
01:56:29.118 --> 01:56:33.008
It doesn't have them because investors, by
definition, use their money

cue-1712
01:56:33.758 --> 01:56:35.368
to make returns, to make profits.

cue-1713
01:56:37.558 --> 01:56:40.948
If all the profits are going to the
nonprofit entity and not to the investors,

cue-1714
01:56:40.948 --> 01:56:43.538
then an investor can't make a return,

cue-1715
01:56:44.438 --> 01:56:45.598
and they won't participate,

cue-1716
01:56:46.408 --> 01:56:46.838
period.

cue-1717
01:56:49.358 --> 01:56:49.498
And

cue-1718
01:56:50.218 --> 01:56:50.898
in all events,

cue-1719
01:56:51.838 --> 01:56:56.698
a company that would flow its profits to
the nonprofit wasn't what was being

cue-1720
01:56:56.698 --> 01:57:01.318
contemplated by Mr. Musk in his
discussions with his co-founders in 2017.

cue-1721
01:57:02.858 --> 01:57:07.558
Musk was looking to divide up the
ownership of the new for-profit into a pie

cue-1722
01:57:07.558 --> 01:57:09.098
that gave the biggest slice to him,

cue-1723
01:57:10.338 --> 01:57:12.898
smaller slices to the other co-founders,

cue-1724
01:57:13.658 --> 01:57:18.568
and a last slice to the existing
employees. And look at how they discussed

cue-1725
01:57:18.568 --> 01:57:20.518
the equity in this for-profit.

cue-1726
01:57:23.178 --> 01:57:25.248
Where in, in this document,

cue-1727
01:57:25.868 --> 01:57:28.718
these documents with Mr. Musk's internal
team

cue-1728
01:57:29.898 --> 01:57:31.798
is the nonprofit's piece supposed to

cue-1729
01:57:32.498 --> 01:57:32.878
fit in?

cue-1730
01:57:34.158 --> 01:57:38.798
Where do you see any profits flowing to
the nonprofit, much less all or most of

cue-1731
01:57:38.798 --> 01:57:38.958
them?

cue-1732
01:57:40.008 --> 01:57:40.478
Nowhere,

cue-1733
01:57:41.478 --> 01:57:45.157
because there was no plan to fund the
nonprofit through this for-profit

cue-1734
01:57:45.158 --> 01:57:45.598
structure.

cue-1735
01:57:46.518 --> 01:57:48.868
You heard Mr. Brockman's testimony on
this,

cue-1736
01:57:49.678 --> 01:57:55.418
that the plan, as Elon Musk presented it,
was if the nonprofit was going to stay,

cue-1737
01:57:56.538 --> 01:58:00.178
it would continue to be funded flat at the
level that it was, supported by

cue-1738
01:58:00.178 --> 01:58:05.238
donations. That's how the nonprofit was
going to stay. It wasn't gonna be funded

cue-1739
01:58:05.238 --> 01:58:06.178
by the for-profit.

cue-1740
01:58:09.498 --> 01:58:11.458
Um, Mr. Mollo a few times has

cue-1741
01:58:12.438 --> 01:58:17.598
pointed you to this document during the
trial about revenues with excess c-cash

cue-1742
01:58:17.598 --> 01:58:18.688
flowing to reserves,

cue-1743
01:58:19.498 --> 01:58:21.988
and I mention this because

cue-1744
01:58:22.958 --> 01:58:24.158
it's important to look at the date.

cue-1745
01:58:26.518 --> 01:58:30.708
This is from November 2015, before OpenAI
is even founded,

cue-1746
01:58:31.518 --> 01:58:36.298
and way before it became clear that much
more capital was needed and that a

cue-1747
01:58:36.298 --> 01:58:38.378
for-profit pivot was required.

cue-1748
01:58:38.978 --> 01:58:42.318
Some sort of for-profit element would be
needed.

cue-1749
01:58:43.378 --> 01:58:47.138
That document was not about a for-profit
enterprise. It's about the nonprofit

cue-1750
01:58:47.138 --> 01:58:47.998
generating revenue,

cue-1751
01:58:48.758 --> 01:58:53.938
and it's roughly two years before the 2017
for-profit negotiations.

cue-1752
01:58:57.558 --> 01:58:59.058
And by 2017,

cue-1753
01:58:59.658 --> 01:59:02.768
Mr. Musk himself was seriously considered,
considering,

cue-1754
01:59:03.378 --> 01:59:04.948
and even actually formed

cue-1755
01:59:05.938 --> 01:59:11.278
a for-profit that was going to effectively
replace the nonprofit as the main event.

cue-1756
01:59:12.018 --> 01:59:14.078
This is back in 2017. And

cue-1757
01:59:14.858 --> 01:59:17.858
you saw the certificate. The B Corp got
formed

cue-1758
01:59:18.938 --> 01:59:20.527
at Mr. Musk's instruction.

cue-1759
01:59:24.298 --> 01:59:28.018
At no point during any of this did Mr.
Musk say, "Hey, wait,

cue-1760
01:59:29.138 --> 01:59:29.858
we can't do this.

cue-1761
01:59:30.698 --> 01:59:32.798
It would violate a commitment you made to
me."

cue-1762
01:59:37.058 --> 01:59:41.618
So now we've looked at the records from
the 2017 for-profit negotiations.

cue-1763
01:59:42.278 --> 01:59:46.678
No discussion of a small adjunct
for-profit or subservient for-profit, a

cue-1764
01:59:46.678 --> 01:59:48.738
for-profit to fund the nonprofit,

cue-1765
01:59:50.018 --> 01:59:51.638
or of any caps of any kind.

cue-1766
01:59:53.238 --> 01:59:58.358
No possible basis from which to infer a
restriction about what an OpenAI

cue-1767
01:59:58.358 --> 02:00:01.678
for-profit should or could be in the
future.

cue-1768
02:00:03.238 --> 02:00:05.318
The record gets even worse for Mr. Musk

cue-1769
02:00:06.138 --> 02:00:09.538
when you move later into 2017 and into
early 2018

cue-1770
02:00:10.358 --> 02:00:13.458
because as the nonprofit is running out of
cash,

cue-1771
02:00:14.658 --> 02:00:15.318
what is he doing?

cue-1772
02:00:16.578 --> 02:00:16.878
He's

cue-1773
02:00:17.838 --> 02:00:22.098
pushing to absorb the whole operation into
a for-profit company.

cue-1774
02:00:24.178 --> 02:00:28.738
Remember, Mr. Musk froze his $5 million
quarterly funding in August 2017,

cue-1775
02:00:31.198 --> 02:00:32.878
right in the middle of the negotiations.

cue-1776
02:00:34.418 --> 02:00:40.018
He did so even though he pledged at the
outset to pay, to, to give a billion

cue-1777
02:00:40.018 --> 02:00:40.478
dollars.

cue-1778
02:00:42.378 --> 02:00:43.778
And by late 2017,

cue-1779
02:00:44.558 --> 02:00:49.398
he saw his opportunity, which was to try
to pull the talent from OpenAI into Tesla

cue-1780
02:00:49.398 --> 02:00:50.738
and build AGI there.

cue-1781
02:00:52.458 --> 02:00:54.618
The documents tell the truth on this.

cue-1782
02:00:55.338 --> 02:00:59.298
What they show is that Mr. Musk was
planning to build a direct competitor to

cue-1783
02:00:59.298 --> 02:01:03.238
OpenAI, all while he was still the
co-chair of, of OpenAI.

cue-1784
02:01:07.298 --> 02:01:12.678
He knew that OpenAI couldn't raise what it
needed to build AGI from donations alone,

cue-1785
02:01:13.458 --> 02:01:15.198
and his co-founders had given him

cue-1786
02:01:16.158 --> 02:01:20.878
clear signal they were not going to grant
control to him over a new for-profit.

cue-1787
02:01:22.498 --> 02:01:27.358
So he started laying plan-- He-- And they
also gave him, by the way, clear signals,

cue-1788
02:01:27.358 --> 02:01:31.118
as you heard from Greg Brockman, they were
not gonna give him control, new control

cue-1789
02:01:31.118 --> 02:01:35.098
over the nonprofit either. So he was
losing his bid for control.

cue-1790
02:01:37.438 --> 02:01:42.838
He started laying plans for Tesla AI, an
AGI lab within a for-profit car company

cue-1791
02:01:42.838 --> 02:01:43.538
that he controlled.

cue-1792
02:01:47.078 --> 02:01:47.658
Mr. Musk,

cue-1793
02:01:48.418 --> 02:01:50.078
you'll recall, denied

cue-1794
02:01:50.718 --> 02:01:52.478
trying to pursue AGI at Tesla.

cue-1795
02:01:53.858 --> 02:01:54.998
And why did he deny that?

cue-1796
02:01:56.318 --> 02:01:59.118
He denied it because it shows two things,

cue-1797
02:02:00.538 --> 02:02:04.818
one that hurts his case and one that hurts
his ego and his case.

cue-1798
02:02:05.718 --> 02:02:11.498
The thing that hurts Mr. Musk's case is
that his push to merge OpenAI into Tesla

cue-1799
02:02:11.498 --> 02:02:17.434
shows he never cared about the nonprofit
structure or about open sourcing.What he

cue-1800
02:02:17.434 --> 02:02:18.594
cared about was winning.

cue-1801
02:02:20.334 --> 02:02:23.824
Musk thought the best chance of winning,
of beating Google DeepMind,

cue-1802
02:02:24.494 --> 02:02:26.654
was to go build AGI within Tesla,

cue-1803
02:02:27.634 --> 02:02:29.034
where he could do it in secret,

cue-1804
02:02:30.074 --> 02:02:33.114
funded by shareholders who wouldn't be
clued into what he was doing.

cue-1805
02:02:34.414 --> 02:02:37.094
Remember, Mr. Brockman testified about
those conversations.

cue-1806
02:02:39.214 --> 02:02:40.614
You heard Josh Ackham

cue-1807
02:02:41.434 --> 02:02:45.954
tell you about that last address that Musk
gave to all OpenAI employees in February

cue-1808
02:02:45.954 --> 02:02:50.034
twenty eighteen as he's leaving the
company to go pursue AGI at Tesla.

cue-1809
02:02:51.694 --> 02:02:52.874
It was all about the race.

cue-1810
02:02:54.274 --> 02:02:57.714
It was the opposite of open source and
nonprofit.

cue-1811
02:02:59.874 --> 02:03:02.854
The nonprofit would disappear at most.

cue-1812
02:03:03.594 --> 02:03:07.294
OpenAI would become a B corp subsidiary of
Tesla.

cue-1813
02:03:08.154 --> 02:03:09.994
You saw those text messages from

cue-1814
02:03:10.974 --> 02:03:12.854
Siobhan Zillis to, to Sam Altman.

cue-1815
02:03:14.934 --> 02:03:19.614
"You think through a B corp subsidiary of
Tesla," she says. That's a for-profit

cue-1816
02:03:19.614 --> 02:03:21.494
subsidiary of another for-profit.

cue-1817
02:03:24.754 --> 02:03:30.234
Now, the second reason that, that Mr. Musk
can't admit the existence of Tesla AI,

cue-1818
02:03:32.374 --> 02:03:38.974
this plan to go pursue AGI at Tesla to
recruit the top talent at OpenAI to do

cue-1819
02:03:38.974 --> 02:03:42.154
what OpenAI was doing but in secret and
well-funded,

cue-1820
02:03:43.434 --> 02:03:45.334
is that it's evidence of Mr. Musk's
failure.

cue-1821
02:03:46.434 --> 02:03:48.354
He tried to compete with OpenAI.

cue-1822
02:03:49.034 --> 02:03:51.454
He recruited Andrej Karpathy from OpenAI,

cue-1823
02:03:52.394 --> 02:03:55.094
then he tasked him with heading up the new
AGI lab.

cue-1824
02:03:56.474 --> 02:03:59.734
Mr. Musk then left OpenAI to pursue Tesla
AI.

cue-1825
02:04:00.894 --> 02:04:03.434
He tried to recruit more people from
OpenAI.

cue-1826
02:04:04.914 --> 02:04:08.054
Thought even, as you saw, about recruiting
Ilya Sutskever.

cue-1827
02:04:08.794 --> 02:04:10.294
He and Siobhan Zillis discussed that.

cue-1828
02:04:12.474 --> 02:04:15.974
He tried to build the AGI lab within
Tesla, and it failed.

cue-1829
02:04:17.514 --> 02:04:19.574
Mr. Musk does not want to admit that.

cue-1830
02:04:21.134 --> 02:04:25.713
But the documents here do tell the true
story, and so does Siobhan Zillis now.

cue-1831
02:04:26.474 --> 02:04:27.394
You saw her on the stand.

cue-1832
02:04:28.834 --> 02:04:34.354
Musk, Tesla AI plan would have seen the
end of the OpenAI nonprofit as a

cue-1833
02:04:34.354 --> 02:04:40.654
significant organization. It would have
seen the end of any openness in OpenAI.

cue-1834
02:04:41.654 --> 02:04:42.234
Everything

cue-1835
02:04:42.994 --> 02:04:43.854
would have been buried.

cue-1836
02:04:45.434 --> 02:04:49.874
See the advantages of burying this in
Tesla for stealth advantage.

cue-1837
02:04:53.774 --> 02:04:55.134
And Ms. Zillis is saying

cue-1838
02:04:55.834 --> 02:04:58.574
that Greg Brockman and Ilya Sutskever,

cue-1839
02:04:59.254 --> 02:05:00.114
Sam Altman, they're not

cue-1840
02:05:00.974 --> 02:05:02.454
naturally hardwired

cue-1841
02:05:03.054 --> 02:05:04.094
as maneuverers.

cue-1842
02:05:05.314 --> 02:05:08.253
That's why they're resisting the Tesla AI
plan.

cue-1843
02:05:13.074 --> 02:05:19.274
So let me turn now to one last category of
records of, of evidence that Elon Musk

cue-1844
02:05:19.274 --> 02:05:19.934
never thought,

cue-1845
02:05:21.234 --> 02:05:26.174
and no one ever gave him reason to think,
that the only for-profit that OpenAI could

cue-1846
02:05:26.174 --> 02:05:28.514
create was one that merely funded the
nonprofit.

cue-1847
02:05:29.354 --> 02:05:34.414
That evidence is about the creation of
OpenAI's for-profit subsidiary in twenty

cue-1848
02:05:34.414 --> 02:05:34.874
eighteen.

cue-1849
02:05:36.914 --> 02:05:41.794
You saw Mr. Musk knew there was a plan to
do this even while he was getting ready to

cue-1850
02:05:41.794 --> 02:05:44.313
leave OpenAI. So this is February twenty
eighteen.

cue-1851
02:05:44.914 --> 02:05:47.174
This is Ms. Zillis emailing Sam Teller,

cue-1852
02:05:47.834 --> 02:05:52.894
Mr. Musk's chief of staff, in February
twenty eighteen, reporting Sam Altman's

cue-1853
02:05:52.894 --> 02:05:58.214
plans to possibly have a traditional
equity offering as a for-profit or B corp.

cue-1854
02:05:58.214 --> 02:06:01.674
Here's some more

cue-1855
02:06:02.634 --> 02:06:05.874
emails from March and April of twenty
eighteen.

cue-1856
02:06:07.054 --> 02:06:07.814
In the first one,

cue-1857
02:06:08.494 --> 02:06:11.854
Ms. Zillis says Altman is thinking through
an instrument where f-- the four to five

cue-1858
02:06:11.854 --> 02:06:16.374
large corporates who are interested can
invest with a return capped at fifty x

cue-1859
02:06:17.054 --> 02:06:20.794
if OpenAI does get some semblance of
money-making AGI.

cue-1860
02:06:22.594 --> 02:06:26.354
And further update on, in April, Mr.
Musk's response,

cue-1861
02:06:27.224 --> 02:06:27.914
"Okay by me."

cue-1862
02:06:31.554 --> 02:06:33.574
You also saw the four-page term sheet

cue-1863
02:06:35.074 --> 02:06:37.014
that has featured so prominently in this
case

cue-1864
02:06:38.033 --> 02:06:41.634
that Mr. Altman shared with Mr. Musk in
August of twenty eighteen.

cue-1865
02:06:42.774 --> 02:06:48.054
Mr. Altman met with Mr. Musk at the Tesla
building on Deer Creek Road to discuss the

cue-1866
02:06:48.054 --> 02:06:48.584
term sheet,

cue-1867
02:06:49.314 --> 02:06:50.394
an earlier draft of it.

cue-1868
02:06:51.214 --> 02:06:52.854
Mr. Altman testified to this.

cue-1869
02:06:53.954 --> 02:06:58.154
Ms. Zillis confirmed that meeting
happened, you'll recall.

cue-1870
02:06:59.814 --> 02:07:05.354
And then when Mr. Altman sent the term
sheet to Mr. Musk, Musk's staff analyzed

cue-1871
02:07:05.354 --> 02:07:09.774
it in detail. Both Siobhan Zillis and
Jared Birchall examined the term sheet.

cue-1872
02:07:12.554 --> 02:07:14.334
So what does the term sheet show? It--

cue-1873
02:07:16.134 --> 02:07:19.014
This is a, this is a capped subsidiary i-
idea that was

cue-1874
02:07:19.754 --> 02:07:21.954
generated by OpenAI, not by Mr. Musk.

cue-1875
02:07:24.314 --> 02:07:29.194
It, it wasn't his idea. So it had-- it was
different from the for-profit structures

cue-1876
02:07:29.194 --> 02:07:31.974
that were being discussed in twenty
seventeen with Mr. Musk.

cue-1877
02:07:33.674 --> 02:07:33.854
And

cue-1878
02:07:34.814 --> 02:07:39.334
here, instead of splitting all the profits
among the founders and employees, the

cue-1879
02:07:39.334 --> 02:07:44.053
nonprofit itself would get a slice. It
would get a capped profit interest in

cue-1880
02:07:44.054 --> 02:07:46.374
return for transferring the value of its
IP,

cue-1881
02:07:47.174 --> 02:07:50.974
and it would get what's called what you
heard discussed as the residual interest,

cue-1882
02:07:50.974 --> 02:07:55.054
any profits exceeding what the investors
would make once they met their caps.

cue-1883
02:07:56.854 --> 02:07:59.374
But even though there were these caps,

cue-1884
02:08:02.954 --> 02:08:09.854
it was absolutely clear that the idea was
to attract private investors hoping to

cue-1885
02:08:09.854 --> 02:08:12.534
make returns in the form of profits. That,
that was clear.

cue-1886
02:08:13.214 --> 02:08:16.514
There would be hundreds of millions of
dollars raised initially,

cue-1887
02:08:17.454 --> 02:08:18.204
and investors

cue-1888
02:08:18.934 --> 02:08:23.822
would, as you see in the term sheet here,
get a hundred xThat's what they would hope

cue-1889
02:08:23.822 --> 02:08:28.142
to, to make. They would get a hundred X
of their investment back if excess

cue-1890
02:08:28.142 --> 02:08:31.602
revenues were generated. That's a profit.
That's a huge one.

cue-1891
02:08:32.392 --> 02:08:33.302
It's, it's math.

cue-1892
02:08:34.442 --> 02:08:39.042
We're talking about fifty billion in
profits being paid to private investors.

cue-1893
02:08:40.302 --> 02:08:42.812
And then there's the employee pool,
meaning the equity

cue-1894
02:08:43.562 --> 02:08:45.072
that the employees would get,

cue-1895
02:08:46.082 --> 02:08:48.122
hundred billion capped target redemption.

cue-1896
02:08:49.542 --> 02:08:50.952
And then there's the second fundraise

cue-1897
02:08:51.742 --> 02:08:53.642
already predicted and planned

cue-1898
02:08:54.362 --> 02:08:58.122
as of twenty eighteen in the term sheet
that Mr. Musk sees,

cue-1899
02:08:59.062 --> 02:09:00.222
more than ten billion

cue-1900
02:09:01.342 --> 02:09:04.582
with a fifteen x, fifteen times

cue-1901
02:09:05.202 --> 02:09:07.462
expected target redemption.

cue-1902
02:09:09.862 --> 02:09:11.542
Now, Mr. Musk claims today

cue-1903
02:09:12.642 --> 02:09:18.022
that the ten billion dollar raise from
Microsoft in January twenty twenty-three,

cue-1904
02:09:18.022 --> 02:09:19.942
that's what gave him the basis for his
claim here.

cue-1905
02:09:21.682 --> 02:09:24.822
But the ten billion dollars was right
there in the term sheet.

cue-1906
02:09:26.202 --> 02:09:31.362
Musk was on notice in twenty eighteen that
if revenues exceeded costs one day,

cue-1907
02:09:31.362 --> 02:09:34.642
investors could stand to gain fifteen
times ten billion.

cue-1908
02:09:36.502 --> 02:09:40.082
All of this was spelled out for Musk in
that four-page document.

cue-1909
02:09:41.142 --> 02:09:42.042
He did not object.

cue-1910
02:09:43.322 --> 02:09:45.362
He did not say, "Hey, wait, stop.

cue-1911
02:09:46.122 --> 02:09:47.462
This violates a commitment to me."

cue-1912
02:09:48.762 --> 02:09:51.292
So how, how does Musk deal with this,
right?

cue-1913
02:09:52.622 --> 02:09:56.622
Having received this term sheet at the
time, having had two witnesses testify

cue-1914
02:09:56.622 --> 02:09:58.362
they reviewed the term sheet with him,

cue-1915
02:09:59.162 --> 02:10:03.342
and having failed to raise any complaint,
how can Musk possibly maintain now

cue-1916
02:10:04.102 --> 02:10:07.922
that OpenAI wasn't allowed to create a
structure that would eventually raise ten

cue-1917
02:10:07.922 --> 02:10:10.142
billion dollars from Microsoft in twenty
twenty-three?

cue-1918
02:10:11.462 --> 02:10:11.982
He can't.

cue-1919
02:10:13.942 --> 02:10:15.942
He just tries to get you to believe

cue-1920
02:10:16.662 --> 02:10:20.122
that he stuck his head in the sand, that
he was duped or, or something.

cue-1921
02:10:21.122 --> 02:10:25.082
And his story keeps changing here.
Remember, you saw that long deposition

cue-1922
02:10:25.082 --> 02:10:26.222
clip, four-minute clip,

cue-1923
02:10:27.282 --> 02:10:30.292
where he's-- he just ended up denying he'd
ever

cue-1924
02:10:31.022 --> 02:10:31.982
looked at the term sheet.

cue-1925
02:10:32.582 --> 02:10:35.062
That's what he did in his deposition,
denied it completely.

cue-1926
02:10:36.322 --> 02:10:37.422
Then you heard him at trial

cue-1927
02:10:38.262 --> 02:10:41.042
say he did read it, but only the purple
box.

cue-1928
02:10:41.922 --> 02:10:43.382
He didn't read the fine print

cue-1929
02:10:44.582 --> 02:10:46.622
of this four-page term sheet.

cue-1930
02:10:47.942 --> 02:10:53.022
And his story now is that he took his
quick review. He, he, he took that review

cue-1931
02:10:53.722 --> 02:10:54.082
and then

cue-1932
02:10:55.422 --> 02:10:56.082
took from it

cue-1933
02:10:58.282 --> 02:11:01.402
that the profits would flow to the
nonprofit,

cue-1934
02:11:01.702 --> 02:11:05.662
and investors would not be investing but
rather donating. And I'll, I'll pause on

cue-1935
02:11:05.662 --> 02:11:08.982
this for just a minute because it is worth
a linger.

cue-1936
02:11:10.062 --> 02:11:12.522
Here you have one of the most
sophisticated businessmen

cue-1937
02:11:13.782 --> 02:11:14.742
in the history of the world

cue-1938
02:11:15.882 --> 02:11:18.842
asking you to believe he got a four-page
term sheet

cue-1939
02:11:19.782 --> 02:11:20.982
for a for-profit entity,

cue-1940
02:11:22.022 --> 02:11:24.462
that he read only the colorful bit at the
top

cue-1941
02:11:25.702 --> 02:11:30.362
and somehow read it to mean that no one,
no investors would hope to make any

cue-1942
02:11:30.362 --> 02:11:30.882
profits,

cue-1943
02:11:31.512 --> 02:11:34.642
that he didn't read the four pages because
that's fine print.

cue-1944
02:11:36.022 --> 02:11:38.382
It is as implausible as it sounds.

cue-1945
02:11:41.302 --> 02:11:44.882
A four le-- four-page-- this is a
high-level summary. It's, it's entitled,

cue-1946
02:11:44.882 --> 02:11:49.962
you'll see it, summary term sheet. That's
not fine print. It's a summary. And if you

cue-1947
02:11:49.962 --> 02:11:51.522
care even a bit, you read it.

cue-1948
02:11:53.262 --> 02:11:57.242
The idea that this document is about
donations and not investments is also

cue-1949
02:11:57.242 --> 02:11:57.802
ludicrous.

cue-1950
02:11:58.942 --> 02:12:01.362
Even Mr. Musk wobbled a bit on that one.

cue-1951
02:12:01.972 --> 02:12:02.951
Um, remember we asked him,

cue-1952
02:12:03.652 --> 02:12:07.862
"So you think Mr. Nadella and Microsoft
were giving a three billion dollar

cue-1953
02:12:07.862 --> 02:12:12.442
charitable donation to OpenAI? Is that
your testimony?" Musk's answer, "No,

cue-1954
02:12:13.122 --> 02:12:14.522
but part of it may have been."

cue-1955
02:12:16.582 --> 02:12:20.022
Jared Birchall only had to glance at the
term sheet

cue-1956
02:12:20.122 --> 02:12:25.542
in twenty eighteen to know [chuckles] that
the for-profit was an investment vehicle.

cue-1957
02:12:25.542 --> 02:12:30.042
Here's his, his, uh, email from the time.

cue-1958
02:12:30.702 --> 02:12:32.171
He reviews the term sheet, and he says,

cue-1959
02:12:32.932 --> 02:12:35.671
"Pretty plain vanilla for-profit
structure,

cue-1960
02:12:36.282 --> 02:12:40.762
so kinda hard to push a narrative that
doesn't involve investors being very

cue-1961
02:12:40.762 --> 02:12:42.222
focused on ROI."

cue-1962
02:12:43.962 --> 02:12:48.492
ROI is return on investment, as Mr. Musk--
or sorry, Mr. Birchall

cue-1963
02:12:49.182 --> 02:12:50.282
acknowledged on the stand.

cue-1964
02:12:52.022 --> 02:12:54.392
And a return on investment, of course, is
profit,

cue-1965
02:12:55.442 --> 02:12:58.082
a profit to investors and not to the
nonprofit.

cue-1966
02:13:00.272 --> 02:13:05.302
Musk knew full well this was an investment
vehicle, and he even considered investing

cue-1967
02:13:05.302 --> 02:13:06.382
in it himself.

cue-1968
02:13:07.542 --> 02:13:12.342
But then he decided not to and to instead
be supportive in spirit.

cue-1969
02:13:13.672 --> 02:13:16.502
This is M- Ms. Zelis's notes from the
time.

cue-1970
02:13:18.322 --> 02:13:21.942
Just for awareness, Elon does not need to
do the Reid call. That's the call with

cue-1971
02:13:21.942 --> 02:13:22.682
Reid Hoffman

cue-1972
02:13:23.282 --> 02:13:25.962
to discuss possibly investing in this
structure.

cue-1973
02:13:26.712 --> 02:13:31.002
Doesn't need to do the Reid call because
he's decided to be supportive in spirit of

cue-1974
02:13:31.002 --> 02:13:34.082
OpenAI, but not participate in the new
instrument.

cue-1975
02:13:34.982 --> 02:13:37.982
How could she have written that if he
didn't consider the term sheet?

cue-1976
02:13:46.782 --> 02:13:46.922
And

cue-1977
02:13:48.222 --> 02:13:48.682
you heard

cue-1978
02:13:51.622 --> 02:13:53.522
from Mr. Musk's own chief of staff,

cue-1979
02:13:54.582 --> 02:13:55.142
Mr. Teller,

cue-1980
02:13:56.662 --> 02:14:01.242
that Musk passed on investing in the
for-profit, not because he thought it was

cue-1981
02:14:01.242 --> 02:14:01.642
scammy,

cue-1982
02:14:02.722 --> 02:14:05.522
but because he doesn't invest in companies
that he can't control.

cue-1983
02:14:07.402 --> 02:14:10.642
His claim now is that he doesn't know
what's, what was going on

cue-1984
02:14:12.062 --> 02:14:15.622
and that he would have stopped it had he
read the fine print.

cue-1985
02:14:17.062 --> 02:14:18.902
That is another thing that is made up for
this

cue-1986
02:14:18.902 --> 02:14:25.570
litigation.So what we've walked through
just

cue-1987
02:14:25.570 --> 02:14:30.829
now is overwhelming evidence from twenty
seventeen and twenty eighteen that Mr.

cue-1988
02:14:30.830 --> 02:14:35.970
Musk knew and supported, knew about and
supported an OpenAI for-profit.

cue-1989
02:14:37.450 --> 02:14:42.150
And there's nothing but Mr. Musk's own
testimony to support the idea that there

cue-1990
02:14:42.150 --> 02:14:46.669
were implied restrictions attached to his
donations on what OpenAI could do with

cue-1991
02:14:46.670 --> 02:14:50.610
those donations with the structure once
Mr. Musk

cue-1992
02:14:51.290 --> 02:14:52.110
left the board.

cue-1993
02:14:54.670 --> 02:14:58.950
Against all this, the only things that Mr.
Musk can point to

cue-1994
02:14:59.710 --> 02:15:03.790
are a handful of documents referencing
enthusiasm for the nonprofit structure

cue-1995
02:15:04.630 --> 02:15:08.169
and a desire to try fundraising within the
nonprofit.

cue-1996
02:15:10.950 --> 02:15:12.930
Contrary to what Mr. Mollo tried to tell
you,

cue-1997
02:15:13.810 --> 02:15:17.030
none of these show any sort of commitment
or, or promise.

cue-1998
02:15:18.230 --> 02:15:23.010
This is Sam Altman, I remain enthusiastic
about the nonprofit structure. That is not

cue-1999
02:15:23.010 --> 02:15:27.630
a promise. Your common sense tells you
that. But you don't have to rely only on

cue-2000
02:15:27.630 --> 02:15:30.789
your common sense because Sam Altman said
it.

cue-2001
02:15:32.410 --> 02:15:33.950
Siobhan Zillis said it.

cue-2002
02:15:35.390 --> 02:15:37.910
I would not view that as a promise.

cue-2003
02:15:43.030 --> 02:15:43.130
And

cue-2004
02:15:43.810 --> 02:15:44.350
it was true.

cue-2005
02:15:45.170 --> 02:15:48.560
Sam Altman was enthusiastic about the
nonprofit structure when he wrote that.

cue-2006
02:15:49.810 --> 02:15:53.630
He thought maybe they could raise the
needed funds within the nonprofit

cue-2007
02:15:53.630 --> 02:15:56.140
structure at that point if they really
made a big push.

cue-2008
02:15:56.770 --> 02:15:59.030
But it didn't work out. You heard that
from Greg Brockman.

cue-2009
02:16:03.310 --> 02:16:08.750
Let's look at, um, the email now from
Siobhan Zillis around the same, same day.

cue-2010
02:16:08.750 --> 02:16:12.520
This is after the for-profit negotiations,
right, right after that honest thoughts,

cue-2011
02:16:13.470 --> 02:16:15.230
um, email exchange.

cue-2012
02:16:17.170 --> 02:16:21.650
She writes that Mr. Brockman and Dr.
Sutskever were good with continuing with

cue-2013
02:16:21.650 --> 02:16:22.630
the nonprofit.

cue-2014
02:16:24.120 --> 02:16:26.710
They would like to continue with the
nonprofit structure.

cue-2015
02:16:27.630 --> 02:16:31.310
Here again, that is not a commitment. That
is just a statement of what they'd like

cue-2016
02:16:31.310 --> 02:16:32.350
to do at that point.

cue-2017
02:16:33.410 --> 02:16:37.870
And in fact, every witness, again,
including Ms. Zillis, told you that

cue-2018
02:16:39.050 --> 02:16:39.870
Mr. Brockman

cue-2019
02:16:40.570 --> 02:16:44.150
and Dr. Sutskever affirmatively refused to
commit,

cue-2020
02:16:44.990 --> 02:16:45.190
right?

cue-2021
02:16:46.190 --> 02:16:52.110
You heard from Mr. Brockman that Mr. Musk
requested very specific conditions to

cue-2022
02:16:52.110 --> 02:16:56.350
resume his funding of the nonprofit after
these email exchanges.

cue-2023
02:16:57.230 --> 02:17:01.950
Mr. Musk's side has tried to make it seem
like the requested commitment was, was

cue-2024
02:17:01.950 --> 02:17:02.560
really simple,

cue-2025
02:17:03.330 --> 02:17:08.370
right? This binary choice between either
do something on your own or continue with

cue-2026
02:17:08.370 --> 02:17:11.890
OpenAI as a nonprofit, but the discussions
continued from here.

cue-2027
02:17:15.750 --> 02:17:18.250
Mr. Musk required his three conditions,

cue-2028
02:17:19.450 --> 02:17:24.090
and he relayed those conditions to Ms.
Zillis and Mr. Teller in late September,

cue-2029
02:17:24.090 --> 02:17:26.590
about a week after the email we were just
looking at.

cue-2030
02:17:27.310 --> 02:17:31.330
The first was granting Mr. Musk two more
board seats on the nonprofit board.

cue-2031
02:17:32.530 --> 02:17:36.560
Second was requiring Mr. Brockman and Dr.
Sutskever to sign a non-solicit,

cue-2032
02:17:37.330 --> 02:17:40.990
which is an agreement not to recruit out
of OpenAI if they were to leave.

cue-2033
02:17:42.410 --> 02:17:46.570
And the third was requiring Mr. Brockman
and Dr. Sutskever

cue-2034
02:17:46.630 --> 02:17:49.090
to commit to staying at OpenAI for two
years.

cue-2035
02:17:50.750 --> 02:17:54.770
Now, Mr. Brockman told you he thought, he
thought carefully about those conditions.

cue-2036
02:17:56.470 --> 02:18:01.330
That's the thought process that you saw in
those journal entries that Musk's team

cue-2037
02:18:01.330 --> 02:18:03.870
cherry-picked and, and showed to you and
to him.

cue-2038
02:18:06.590 --> 02:18:11.650
He and Dr. Sutskever worried that
accepting Mr. Musk's terms would tie them

cue-2039
02:18:11.650 --> 02:18:16.190
to the nonprofit without the resources to
achieve its mission. Because Mr. Musk

cue-2040
02:18:16.190 --> 02:18:16.400
wasn't,

cue-2041
02:18:17.110 --> 02:18:19.950
wasn't just-- he wasn't gonna contribute
tons more, he's just gonna resume his

cue-2042
02:18:19.950 --> 02:18:24.370
quarterly funding of five million dollars
if they accepted those conditions.

cue-2043
02:18:25.330 --> 02:18:29.570
And you heard they even considered firing
Mr. Musk from the board because the only

cue-2044
02:18:29.570 --> 02:18:34.370
for-profit structure that he would
conceive of was one where he had

cue-2045
02:18:34.370 --> 02:18:35.730
unilateral control.

cue-2046
02:18:37.210 --> 02:18:38.090
And Mr. Mollo,

cue-2047
02:18:38.950 --> 02:18:42.670
you saw, tried to, to make it seem like
that was all made up by Greg Brockman on

cue-2048
02:18:42.670 --> 02:18:43.150
the stand.

cue-2049
02:18:44.469 --> 02:18:44.900
You saw

cue-2050
02:18:45.789 --> 02:18:46.450
on redirect

cue-2051
02:18:47.390 --> 02:18:51.870
that Mr. Brockman was writing about the
pl-- the thought of firing Elon

cue-2052
02:18:52.790 --> 02:18:57.450
the day before the journal entry that Mr.
Mollo was showing and in the same journal

cue-2053
02:18:57.450 --> 02:19:02.630
entry that Mr. Mollo was showing. This was
not something Mr. Brockman made up for

cue-2054
02:19:02.630 --> 02:19:02.930
trial.

cue-2055
02:19:08.510 --> 02:19:09.890
They didn't do it, though, right?

cue-2056
02:19:10.690 --> 02:19:15.170
Mr. Brockman and Dr. Sutskever did not
fire Elon Musk from the board. They

cue-2057
02:19:15.170 --> 02:19:17.610
decided to try to make the nonprofit work

cue-2058
02:19:18.350 --> 02:19:22.790
with Mr. Musk not accepting his conditions
but trying to make it work.

cue-2059
02:19:23.630 --> 02:19:28.330
They thought it would be morally bankrupt
to get Musk to resume his quarterly

cue-2060
02:19:28.330 --> 02:19:34.950
contributions to the nonprofit if they
gave their word, and then they went off to

cue-2061
02:19:34.950 --> 02:19:36.190
start a for-profit

cue-2062
02:19:36.950 --> 02:19:40.970
if those resumed donations turned out to
have been insufficient to advance the

cue-2063
02:19:40.970 --> 02:19:41.330
mission.

cue-2064
02:19:43.530 --> 02:19:47.410
They thought that would be morally
bankrupt, so they didn't do it.

cue-2065
02:19:48.890 --> 02:19:52.050
They didn't agree to the conditions. No
one made a commitment to Mr. Musk.

cue-2066
02:19:54.670 --> 02:19:57.040
And in the weeks before Mr. Musk's
departure,

cue-2067
02:19:57.770 --> 02:20:01.950
Mr. Brockman and Dr. Sutskever continued
to share their uncertainty about OpenAI's

cue-2068
02:20:01.950 --> 02:20:02.330
future.

cue-2069
02:20:03.870 --> 02:20:07.710
Here's, here's a document that you saw, I
think with Siobhan Zillis,

cue-2070
02:20:08.590 --> 02:20:13.630
that proves Mr. Musk's claim that the
founders committed to a nonprofit is just

cue-2071
02:20:13.630 --> 02:20:14.070
false.

cue-2072
02:20:16.670 --> 02:20:19.010
Mr. Mollo, I, I don't think, showed you
this document.

cue-2073
02:20:19.730 --> 02:20:25.570
It's Siobhan Zillis' notes from the time,
reporting on what Greg, Ilya, and Altman

cue-2074
02:20:26.630 --> 02:20:32.177
were telling her.They're still not sure
whether, whether they'll stay a nonprofit

cue-2075
02:20:32.178 --> 02:20:33.578
and focus on donations.

cue-2076
02:20:34.558 --> 02:20:36.938
Figure out a structure for equity
fundraise

cue-2077
02:20:37.678 --> 02:20:41.958
or a private version of a token offering
you guys had discussed before.

cue-2078
02:20:43.338 --> 02:20:45.218
This is February eleventh, twenty
eighteen.

cue-2079
02:20:47.618 --> 02:20:51.398
There was just no commitment made to Mr.
Musk about OpenAI's corporate structure.

cue-2080
02:20:52.598 --> 02:20:57.858
There was never any specific purpose tied
to his contributions, certainly not one

cue-2081
02:20:57.858 --> 02:21:00.638
related to OpenAI keeping the nonprofit
the main event.

cue-2082
02:21:01.618 --> 02:21:02.418
It's all made up.

cue-2083
02:21:03.378 --> 02:21:04.638
And let's look at the testimony.

cue-2084
02:21:06.438 --> 02:21:10.698
None of the other founders backed up
Musk's story.

cue-2085
02:21:11.938 --> 02:21:14.598
Sam Altman, Greg Brockman, Ilya Sutskever.

cue-2086
02:21:15.738 --> 02:21:17.658
None of the other witnesses.

cue-2087
02:21:19.898 --> 02:21:20.458
Even, even

cue-2088
02:21:21.058 --> 02:21:22.478
Mr. Musk's own witnesses

cue-2089
02:21:23.598 --> 02:21:24.638
say they were unaware

cue-2090
02:21:25.778 --> 02:21:29.118
of any promises made to him by OpenAI.

cue-2091
02:21:31.818 --> 02:21:32.228
So that,

cue-2092
02:21:33.018 --> 02:21:35.168
that is the first restriction. There was
none.

cue-2093
02:21:36.198 --> 02:21:38.058
You can't imply it. The record

cue-2094
02:21:38.978 --> 02:21:39.578
defeats it.

cue-2095
02:21:41.578 --> 02:21:43.978
Nor was there any condition or commitment

cue-2096
02:21:44.618 --> 02:21:45.938
related to open sourcing.

cue-2097
02:21:47.158 --> 02:21:51.258
And none of Mr. Musk's donations were
given for the specific purpose of open

cue-2098
02:21:51.258 --> 02:21:53.258
sourcing. You didn't see any, anything
saying that.

cue-2099
02:21:54.138 --> 02:21:56.318
And this we can cover a lot more briefly.

cue-2100
02:21:57.578 --> 02:21:58.138
You heard

cue-2101
02:21:58.938 --> 02:22:04.208
Dr. Sutskever say there was no promise
made to Mr. Musk on open sourcing.

cue-2102
02:22:06.508 --> 02:22:10.918
Did you ever make such a promise to Mr.
Musk? Definitely not. I mean, on the

cue-2103
02:22:10.918 --> 02:22:15.298
contrary, early, early on, I told him that
OpenAI is not going to open source

cue-2104
02:22:15.298 --> 02:22:15.678
everything.

cue-2105
02:22:17.058 --> 02:22:18.158
Even OpenAI's,

cue-2106
02:22:19.078 --> 02:22:21.988
um, documents, the, you know, all, all the
documents from the time,

cue-2107
02:22:22.938 --> 02:22:27.558
the general certificate of incorporation,
the original one from twenty fifteen,

cue-2108
02:22:28.718 --> 02:22:32.738
says only that op-- the OpenAI nonprofit
will open source its technology when

cue-2109
02:22:32.738 --> 02:22:33.358
applicable.

cue-2110
02:22:34.798 --> 02:22:38.998
And it was clear from the outset, we, we
saw this email from January of twenty

cue-2111
02:22:39.058 --> 02:22:39.598
sixteen,

cue-2112
02:22:40.678 --> 02:22:46.578
that the open did not mean open sourcing.
This is Ilya Sutskever writing to Elon

cue-2113
02:22:46.578 --> 02:22:47.938
Musk and his other co-founders.

cue-2114
02:22:49.218 --> 02:22:50.798
It didn't mean open sourcing.

cue-2115
02:22:51.418 --> 02:22:55.778
The open in OpenAI means that everyone
should benefit from the fruits of AI after

cue-2116
02:22:55.778 --> 02:22:56.338
it's built,

cue-2117
02:22:56.978 --> 02:22:59.878
but it's totally okay to not share the
science.

cue-2118
02:23:00.738 --> 02:23:02.518
Elon Musk's response, "Yep."

cue-2119
02:23:04.938 --> 02:23:05.498
There's also

cue-2120
02:23:06.138 --> 02:23:10.798
Greg Brockman's email from twenty sixteen,
just a few months later,

cue-2121
02:23:12.338 --> 02:23:13.078
to Elon Musk.

cue-2122
02:23:14.018 --> 02:23:16.617
I plan to say that I can't see why we
would do that.

cue-2123
02:23:17.978 --> 02:23:20.918
Our mission, which means criticize Google
for not open sourcing.

cue-2124
02:23:21.778 --> 02:23:23.978
Our mission is to maximally benefit the
world,

cue-2125
02:23:24.658 --> 02:23:28.578
and we don't have a problem with people
keeping things proprietary. It's fine to

cue-2126
02:23:28.578 --> 02:23:32.458
make money off this stuff, and we may even
generate revenue ourselves one day.

cue-2127
02:23:33.778 --> 02:23:37.538
Musk's response, "Whoa, that's really
interesting. Who called from Google?"

cue-2128
02:23:41.458 --> 02:23:46.237
And you heard the, the Dota technology
that powered that big win in August of

cue-2129
02:23:46.238 --> 02:23:50.118
twenty seventeen. That wasn't open
sourced. Mr. Musk was still at the company

cue-2130
02:23:50.118 --> 02:23:52.198
at the time. He didn't complain about it.

cue-2131
02:23:54.038 --> 02:23:58.018
And Mr. Brockman told you the, the, the
reasoning for not open sourcing that

cue-2132
02:23:58.018 --> 02:23:59.078
wasn't about safety,

cue-2133
02:23:59.838 --> 02:24:00.718
that particular,

cue-2134
02:24:01.358 --> 02:24:01.518
um,

cue-2135
02:24:02.218 --> 02:24:03.058
uh, decision.

cue-2136
02:24:03.718 --> 02:24:08.348
It was about strategic reasons. So
OpenAI's technological progress didn't

cue-2137
02:24:08.348 --> 02:24:10.518
accelerate the progress of competitors.

cue-2138
02:24:12.378 --> 02:24:17.538
In early twenty twenty, OpenAI amended its
certificate of incorporation to remove

cue-2139
02:24:17.538 --> 02:24:21.158
the reference to open sourcing altogether.
It's a public document.

cue-2140
02:24:22.298 --> 02:24:26.158
A few months later, it released a blog
post explaining why it was sharing its

cue-2141
02:24:26.158 --> 02:24:30.138
technology through an API instead of
through open sourcing.

cue-2142
02:24:31.878 --> 02:24:34.118
Another public document from twenty
twenty.

cue-2143
02:24:36.438 --> 02:24:43.338
There was never any commitment to open
source of any kind made to Mr.

cue-2144
02:24:43.338 --> 02:24:43.658
Musk,

cue-2145
02:24:44.578 --> 02:24:49.038
and none of Mr. Musk's contributions to
OpenAI were earmarked for the specific

cue-2146
02:24:49.038 --> 02:24:52.578
purpose of open sourcing. There is zero
evidence of that.

cue-2147
02:24:55.638 --> 02:24:55.798
Now,

cue-2148
02:24:58.438 --> 02:25:01.708
because his donations were not made for
any specific charitable purpose,

cue-2149
02:25:02.738 --> 02:25:05.598
Musk has not proved that he had a
charitable trust.

cue-2150
02:25:06.518 --> 02:25:09.738
Hasn't proved he ever had one. He hasn't
come close to proving that.

cue-2151
02:25:10.758 --> 02:25:11.998
But even if he could prove it,

cue-2152
02:25:12.838 --> 02:25:14.578
prove that he had a trust at some point,

cue-2153
02:25:15.178 --> 02:25:20.018
his claim still fails, and that's because
his money had already been spent by the

cue-2154
02:25:20.018 --> 02:25:25.538
time of the supposed wrongdoing. So any
charitable trust had expired.

cue-2155
02:25:28.078 --> 02:25:28.268
And

cue-2156
02:25:29.478 --> 02:25:30.318
as you can see,

cue-2157
02:25:31.698 --> 02:25:35.318
what we were just actually-- go back to
the slide for a moment. This is just

cue-2158
02:25:36.118 --> 02:25:40.978
the judge's jury instructions. If a trust
is created and subsequently the whole of

cue-2159
02:25:40.978 --> 02:25:44.338
the trust property ceases to exist, the
trust is terminated.

cue-2160
02:25:45.978 --> 02:25:50.258
Musk needed to have shown that the trust
property was still there in twenty

cue-2161
02:25:50.258 --> 02:25:50.888
twenty-three.

cue-2162
02:25:51.978 --> 02:25:56.817
He didn't. Here are Musk's donations, and
you saw this chart with, um, with Mr.

cue-2163
02:25:56.818 --> 02:25:57.648
Dudney yesterday.

cue-2164
02:25:59.278 --> 02:26:02.798
The quarterly contributions ended in June
twenty seventeen,

cue-2165
02:26:03.818 --> 02:26:08.398
and you heard from Mr. Dudney, the, the
forensic accountant, that they were all

cue-2166
02:26:08.398 --> 02:26:11.478
used up no later than November twenty
seventeen.

cue-2167
02:26:14.018 --> 02:26:18.538
Mr. Dudney explained after reviewing
OpenAI's financial documents that the

cue-2168
02:26:18.538 --> 02:26:20.978
money was spent on exactly what you'd
expect:

cue-2169
02:26:21.658 --> 02:26:22.098
compute,

cue-2170
02:26:22.758 --> 02:26:24.478
employees, facilities.

cue-2171
02:26:25.138 --> 02:26:30.734
It was spent in the pursuit of OpenAI's
general mission.And the rent

cue-2172
02:26:30.734 --> 02:26:35.774
contributions, meanwhile, were all used up
by September 2020, when the last rent

cue-2173
02:26:35.774 --> 02:26:36.694
payment was made.

cue-2174
02:26:38.474 --> 02:26:40.704
Everything that Mr. Musk donated

cue-2175
02:26:41.374 --> 02:26:44.494
was spent by September twenty, twenty-
2020.

cue-2176
02:26:45.274 --> 02:26:46.604
That's three years before

cue-2177
02:26:47.914 --> 02:26:49.194
2023.

cue-2178
02:26:50.534 --> 02:26:54.654
There was no trust in existence by the
time of this supposed breach.

cue-2179
02:27:00.174 --> 02:27:00.334
Now,

cue-2180
02:27:00.954 --> 02:27:03.374
I, I think I heard Mr. Mullow say he, he,

cue-2181
02:27:05.114 --> 02:27:05.274
he's

cue-2182
02:27:06.014 --> 02:27:06.974
suggesting that

cue-2183
02:27:08.214 --> 02:27:14.574
the money that Mr. Musk gave from 2017 to
2020 helped OpenAI develop technology that

cue-2184
02:27:14.574 --> 02:27:16.434
formed the basis for,

cue-2185
02:27:17.214 --> 02:27:18.054
you know, for its,

cue-2186
02:27:18.714 --> 02:27:21.234
its IP later on. There was some suggestion
of that.

cue-2187
02:27:22.454 --> 02:27:23.914
But that theory does not work.

cue-2188
02:27:24.894 --> 02:27:25.454
Mr. Musk

cue-2189
02:27:26.194 --> 02:27:30.834
bears the burden of proving that money or
property that he donated was still in a

cue-2190
02:27:30.834 --> 02:27:34.424
trust at OpenAI as of 2023, and he did not
do that.

cue-2191
02:27:35.814 --> 02:27:37.194
Suggestions are not enough.

cue-2192
02:27:38.494 --> 02:27:40.114
The money he gave was spent,

cue-2193
02:27:40.854 --> 02:27:42.554
and he offered no evidence,

cue-2194
02:27:43.154 --> 02:27:43.474
none,

cue-2195
02:27:44.274 --> 02:27:47.994
showing that that money became property of
some kind that lives on today.

cue-2196
02:27:50.054 --> 02:27:53.644
If ever there was a charitable trust, it
ended in September 2020,

cue-2197
02:27:54.914 --> 02:28:00.054
and Musk doesn't argue that any act
occurring before that or after that date,

cue-2198
02:28:00.854 --> 02:28:05.234
before that date, sorry, violated any
right that was owed to him, so he has no

cue-2199
02:28:05.234 --> 02:28:05.454
claim.

cue-2200
02:28:07.574 --> 02:28:11.514
At this point, ladies and gentlemen, I am
going to invite Mr. Savitt to come and

cue-2201
02:28:11.514 --> 02:28:14.994
address you on the remaining issues. It's
been a pleasure to address you. Thank you.

cue-2202
02:28:14.994 --> 02:28:16.374
Well, at this point,

cue-2203
02:28:17.074 --> 02:28:19.714
we are gonna take a recess so you can have
lunch.

cue-2204
02:28:19.714 --> 02:28:20.214
Oh, great.

cue-2205
02:28:21.234 --> 02:28:27.594
Um, so we will, um, jury's lunch is here,
and we'll stand in recess for about 30

cue-2206
02:28:27.594 --> 02:28:27.954
minutes.

cue-2207
02:28:28.714 --> 02:28:29.794
-Okay.
-Peace, right, for the jury.

cue-2208
02:28:32.894 --> 02:28:59.034
Okay,

cue-2209
02:28:59.034 --> 02:28:59.844
Mr. Savitt,

cue-2210
02:29:00.734 --> 02:29:02.254
what's your anticipation?

cue-2211
02:29:04.074 --> 02:29:04.124
Of

cue-2212
02:29:04.894 --> 02:29:07.174
-what, Judge?
-In terms of time.

cue-2213
02:29:12.474 --> 02:29:14.294
I, I-- it's fine, I just wanna know,

cue-2214
02:29:15.294 --> 02:29:15.674
Colin.

cue-2215
02:29:17.754 --> 02:29:22.274
Your Honor, I'm estimating about forty
minutes or so, forty-five at the most.

cue-2216
02:29:23.994 --> 02:29:24.234
Okay.

cue-2217
02:29:25.754 --> 02:29:25.954
Um,

cue-2218
02:29:26.814 --> 02:29:30.054
so if those estimates hold through, then

cue-2219
02:29:30.814 --> 02:29:36.534
we will, after the lunch break, I will do
both of you in that set. Okay. All right,

cue-2220
02:29:36.534 --> 02:29:39.374
we'll stand in recess for about thirty
minutes.

cue-2221
02:29:45.354 --> 02:29:48.054
Your Honor, there is one issue we wanted
to raise arising out of Ms. Eddy's

cue-2222
02:29:48.054 --> 02:29:49.674
presentation.

cue-2223
02:29:50.314 --> 02:29:52.564
Uh, we can do it now, or we can do it at
the end of the recess.

cue-2224
02:29:52.564 --> 02:29:54.154
No, that's okay. Um, Ms. Eddy,

cue-2225
02:29:55.374 --> 02:29:57.334
you've got an issue from plaintiffs.

cue-2226
02:30:00.394 --> 02:30:04.454
We're back on the record. The record will
reflect that the jury is not here.

cue-2227
02:30:05.994 --> 02:30:06.174
Sure.

cue-2228
02:30:07.234 --> 02:30:07.914
Your Honor, um,

cue-2229
02:30:08.874 --> 02:30:12.594
a, a couple different times, and in
particular at page one eleven, line six to

cue-2230
02:30:12.594 --> 02:30:14.534
twenty of the real time, Ms. Eddy

cue-2231
02:30:15.274 --> 02:30:17.454
argued in substance that, uh,

cue-2232
02:30:18.574 --> 02:30:22.814
M-Musk, any condition that Musk imposed to
follow the general mission of the

cue-2233
02:30:22.814 --> 02:30:27.154
organization couldn't be sufficient
because he had to show some specific trust

cue-2234
02:30:27.154 --> 02:30:29.924
purpose above and beyond the
organization's mission,

cue-2235
02:30:30.534 --> 02:30:34.874
and our concern is that that's contrary,
the mission's, uh, the organization's

cue-2236
02:30:34.874 --> 02:30:35.494
general mission.

cue-2237
02:30:36.254 --> 02:30:39.614
And our concern is that that's contrary to
what this court ruled in your summary

cue-2238
02:30:39.614 --> 02:30:43.094
judgment order. In particular, I'm looking
at, uh, docket three ninety,

cue-2239
02:30:43.834 --> 02:30:45.754
page seventeen, where Your Honor said,

cue-2240
02:30:46.374 --> 02:30:51.334
"The argument that a settlor must attach
specific additional conditions to a trust

cue-2241
02:30:51.334 --> 02:30:55.934
where a charitable donation already aligns
with the mission of an organization is

cue-2242
02:30:55.934 --> 02:31:00.214
not supported by case law." And so we
think this has given, uh, the jury a

cue-2243
02:31:00.214 --> 02:31:03.694
misunderstanding or a misleading
impression of Your Honor's instructions

cue-2244
02:31:03.694 --> 02:31:08.014
and wanted to request a supplemental
instruction just to clarify the point Your

cue-2245
02:31:08.014 --> 02:31:10.794
Honor made in the summary judgment
ruling.

cue-2246
02:31:15.794 --> 02:31:21.094
I

cue-2247
02:31:24.234 --> 02:31:26.234
-agree with that. Overruled.
-Thank you,

cue-2248
02:32:04.894 --> 02:32:19.234
Your

cue-2249
02:32:19.234 --> 02:32:20.814
-Honor.
-Excuse me, everyone. We'll need to clear

cue-2250
02:32:20.814 --> 02:32:21.854
the courtroom. Thank you.

cue-2251
02:53:47.914 --> 02:53:50.474
Recording stopped.

cue-2252
02:59:27.744 --> 02:59:29.004
Recording in progress.

cue-2253
02:59:40.264 --> 03:00:49.404
Court

cue-2254
03:00:49.404 --> 03:00:50.384
is back in session.

cue-2255
03:00:55.624 --> 03:00:57.853
Okay, let's, uh, call the jury in.

cue-2256
03:01:11.884 --> 03:01:13.984
You were the sole representative, Mr. Cry.

cue-2257
03:02:27.944 --> 03:02:31.684
-Are you using a briefcase as well?
-[clears throat]

cue-2258
03:02:34.024 --> 03:02:36.584
I, I, I asked whether you were using the
briefcase as

cue-2259
03:02:36.584 --> 03:03:44.594
well.Okay,

cue-2260
03:03:44.594 --> 03:03:49.353
you may all be seated. The record will
reflect that the jury is back, and we're

cue-2261
03:03:49.354 --> 03:03:51.674
good to go. Great. Mr. Savitt.

cue-2262
03:04:13.794 --> 03:04:15.354
-Thank you, Your Honor.
-You may proceed.

cue-2263
03:04:15.354 --> 03:04:15.754
Thank you.

cue-2264
03:04:16.414 --> 03:04:17.594
Members of the jury, hello again.

cue-2265
03:04:18.634 --> 03:04:18.703
Um,

cue-2266
03:04:19.354 --> 03:04:22.874
it's really good to see you. It's, um,
really good to have the chance to visit

cue-2267
03:04:22.874 --> 03:04:23.194
with you

cue-2268
03:04:23.934 --> 03:04:27.154
again. It's been a couple of weeks, maybe
even three, since we had the chance to

cue-2269
03:04:27.154 --> 03:04:27.383
speak,

cue-2270
03:04:28.234 --> 03:04:30.664
uh, directly, and grateful for that
opportunity.

cue-2271
03:04:31.474 --> 03:04:34.124
Grateful also for your time and attention,

cue-2272
03:04:34.914 --> 03:04:35.964
which we know is precious,

cue-2273
03:04:36.774 --> 03:04:37.114
and

cue-2274
03:04:38.334 --> 03:04:38.534
you've

cue-2275
03:04:39.314 --> 03:04:41.234
been very generous with it over the past
few weeks.

cue-2276
03:04:42.914 --> 03:04:46.554
We're gonna talk a little bit more about
the evidence as we see it in this case,

cue-2277
03:04:46.554 --> 03:04:50.414
and, um, well, thank you for, for the
indulgence there.

cue-2278
03:04:51.034 --> 03:04:51.194
Um,

cue-2279
03:04:51.974 --> 03:04:55.954
I wanna pick up where, where my colleague,
Ms., Ms. Eddy left off. She walked you

cue-2280
03:04:55.954 --> 03:05:01.274
through the mountain of evidence that
shows that Musk's general contributions,

cue-2281
03:05:01.274 --> 03:05:06.454
lacking any specific charitable purpose,
cannot possibly have created a charitable

cue-2282
03:05:06.454 --> 03:05:10.394
trust. And we think under the law and
under the evidence, that proposition, that

cue-2283
03:05:10.394 --> 03:05:14.644
idea isn't capable really of being
disputed. Ms. Eddy has also shown you, um,

cue-2284
03:05:14.644 --> 03:05:19.834
the evidence, likewise overwhelming, that
Musk's donations were spent exactly as

cue-2285
03:05:19.834 --> 03:05:24.994
they were requested to have been spent,
exactly consistent with OpenAI's general

cue-2286
03:05:25.054 --> 03:05:26.074
charitable purpose.

cue-2287
03:05:26.914 --> 03:05:31.894
There is one more thread on the matter of
charitable trust, however, that I wanted

cue-2288
03:05:31.894 --> 03:05:35.194
to speak with you before moving on to a
few other, few other subjects.

cue-2289
03:05:35.814 --> 03:05:35.824
Um,

cue-2290
03:05:36.754 --> 03:05:42.534
e-even if you imagine, contrary to fact,
that a charitable trust was created, Musk

cue-2291
03:05:42.534 --> 03:05:46.374
would still have to show that the terms of
that trust were breached for him to have

cue-2292
03:05:46.374 --> 03:05:51.774
made out a claim. The question, members of
the jury, boils down to this: Has the

cue-2293
03:05:51.774 --> 03:05:57.274
OpenAI nonprofit respected its general
founding principles? And with the evidence

cue-2294
03:05:57.274 --> 03:05:58.054
before you,

cue-2295
03:05:58.834 --> 03:06:04.454
you can see we believe that the answer is,
has to be yes. I, I wanna emphasize that

cue-2296
03:06:04.454 --> 03:06:08.694
in some sense, this question doesn't,
doesn't really matter legally. As we've

cue-2297
03:06:08.694 --> 03:06:14.034
seen, Mr. Musk only can have a claim if
specific commitments were made to him

cue-2298
03:06:14.034 --> 03:06:18.874
about the uses of his donations, showing
that they were different than OpenAI's

cue-2299
03:06:18.874 --> 03:06:23.574
general commitments, which we submit, uh,
they weren't for the reasons that Ms. Eddy

cue-2300
03:06:23.574 --> 03:06:24.763
has discussed with you.

cue-2301
03:06:25.474 --> 03:06:29.594
But counsel for Mr. Musk, Mr. Mulloway,
spent a lot of time speaking about this

cue-2302
03:06:29.594 --> 03:06:33.014
question, immaterial though we think it
is, and I didn't wanna leave it,

cue-2303
03:06:34.054 --> 03:06:36.194
didn't wanna leave it unanswered, and we
think the answer is easy.

cue-2304
03:06:36.894 --> 03:06:40.594
Uh, OpenAI has respected its charitable
principles from

cue-2305
03:06:41.754 --> 03:06:43.394
day one to today,

cue-2306
03:06:43.454 --> 03:06:47.914
and as we think the evidence at trial's
shown. The OpenAI nonprofit, which today

cue-2307
03:06:47.914 --> 03:06:53.194
is called the OpenAI Foundation, is still
a 501(c)(3)

cue-2308
03:06:53.254 --> 03:06:57.254
nonprofit organization, and that's the
relevant provision of the, the IRS code,

cue-2309
03:06:57.254 --> 03:06:58.794
the tax code that identifies it.

cue-2310
03:06:59.914 --> 03:07:01.314
OpenAI Foundation

cue-2311
03:07:01.934 --> 03:07:07.013
has always been a 501(c)(3)
o-organization. It remains that, and it's

cue-2312
03:07:07.014 --> 03:07:09.274
governed by the same nonprofit
certificate,

cue-2313
03:07:10.114 --> 03:07:10.254
uh,

cue-2314
03:07:10.954 --> 03:07:15.374
as it was in 2020, which was not
meaningfully distinguishable from the

cue-2315
03:07:15.374 --> 03:07:19.154
certificate as it was incorporated in
2015, all of which has been public for the

cue-2316
03:07:19.154 --> 03:07:25.414
world to see for many, many years. Uh, no
one in this trial is disputing that. I

cue-2317
03:07:25.414 --> 03:07:29.654
don't think it's really subject to dispute
either that the foundation has some two

cue-2318
03:07:29.654 --> 03:07:34.754
hundred billion dollars in assets as we,
as we speak today. You heard that from Mr.

cue-2319
03:07:34.754 --> 03:07:38.954
Taylor, the foundation's chairman. You
heard it from Mr. Altman, and there's no

cue-2320
03:07:38.954 --> 03:07:44.194
contrary evidence in the records. It's not
subject to dispute. The OpenAI Foundation

cue-2321
03:07:44.194 --> 03:07:50.114
is today one of the largest nonprofits in
history. It is funding initiatives aimed

cue-2322
03:07:50.114 --> 03:07:55.094
at fostering AI resilience, which is
making sure that the economy is resilient

cue-2323
03:07:55.094 --> 03:07:59.174
in light of the coming of artificial
intelligence, and focusing on curing

cue-2324
03:07:59.174 --> 03:08:00.903
diseases and trying to make

cue-2325
03:08:02.554 --> 03:08:07.014
life and work better for healthcare
professionals in the healthcare industry.

cue-2326
03:08:07.014 --> 03:08:12.294
No private person has received any part of
the nonprofit's net revenues.

cue-2327
03:08:13.194 --> 03:08:18.374
None of the foundation's assets have been
transferred to anyone for anything other

cue-2328
03:08:18.374 --> 03:08:19.444
than fair value,

cue-2329
03:08:20.154 --> 03:08:20.354
ever.

cue-2330
03:08:21.554 --> 03:08:21.834
I wanna,

cue-2331
03:08:22.574 --> 03:08:26.554
I wanna talk a l-- for a moment about, uh,
the, the document that's now on your

cue-2332
03:08:26.554 --> 03:08:30.874
screen. It's, it's a reports, report of a,
an accounting firm, a financial

cue-2333
03:08:30.874 --> 03:08:35.274
evaluation firm called Hemming Morse. As
you heard, the nonprofit transferred some

cue-2334
03:08:35.274 --> 03:08:36.844
of its assets to OpenAI,

cue-2335
03:08:37.734 --> 03:08:40.094
the for-profit branch, back in 2019,

cue-2336
03:08:40.974 --> 03:08:46.164
which is seven years ago. It's undisputed,
and by that I mean no one's said anything

cue-2337
03:08:46.164 --> 03:08:50.533
on the other side of the question, that
that transfer was for fair value, as

cue-2338
03:08:50.534 --> 03:08:55.254
determined by an independent third-party
valuation report at this time, and that's

cue-2339
03:08:55.254 --> 03:08:59.494
the Hemming Morse report that we've
introduced into evidence and is before you

cue-2340
03:08:59.494 --> 03:09:04.454
now. And the evidence, um, from the
witnesses bears all this out. Ms.

cue-2341
03:09:04.454 --> 03:09:08.314
McCauley, who, as you saw, had some harsh
things to say about some of my clients,

cue-2342
03:09:08.314 --> 03:09:09.794
was asked whether the--

cue-2343
03:09:10.444 --> 03:09:13.764
and she, she was on the board at the time,
was asked whether it was fair value.

cue-2344
03:09:13.764 --> 03:09:17.874
"That's correct," she said. You even
heard, um, fr-from plaintiff's own

cue-2345
03:09:17.874 --> 03:09:23.174
witnesses exactly to that effect, and not
only that, even Mr. Musk's expert,

cue-2346
03:09:23.914 --> 03:09:28.434
Mr. Schizer, can't complain about the
valuation. No basis to question its

cue-2347
03:09:28.434 --> 03:09:34.989
validity. No one-Is disputing. No one is
claiming that the transfer of assets in

cue-2348
03:09:34.990 --> 03:09:39.970
two thousand nineteen somehow violated
OpenAI's certificate of incorporation.

cue-2349
03:09:41.150 --> 03:09:45.929
And in fact, the evidence is clear that
Altman, Brockman, Sutskever, and their

cue-2350
03:09:45.930 --> 03:09:51.830
colleagues have turned that sixty million
dollar contribution of assets into a two

cue-2351
03:09:51.830 --> 03:09:52.820
hundred billion dollar

cue-2352
03:09:53.730 --> 03:09:58.389
fund of charitable assets. That's sixty
million into two hundred billion. Before

cue-2353
03:09:58.390 --> 03:10:01.710
speaking with you, I tried to do the math
to figure out the percentage, and it's

cue-2354
03:10:01.710 --> 03:10:05.730
actually hard to do. And with the aid of
some artificial intelligence, I can report

cue-2355
03:10:05.730 --> 03:10:11.950
to you that this reflects an increase in
value of three hundred and thirty-three

cue-2356
03:10:11.950 --> 03:10:12.970
thousand percent.

cue-2357
03:10:14.210 --> 03:10:18.710
Three hundred and thirty-three thousand
percent increase. Apparently not enough

cue-2358
03:10:18.710 --> 03:10:19.420
for Mr. Schizer.

cue-2359
03:10:21.170 --> 03:10:25.530
And that just talks about the increase in
valuation, the increase in charitable

cue-2360
03:10:25.530 --> 03:10:31.410
funds that are available for initiatives
in health resilience and otherwise to make

cue-2361
03:10:31.410 --> 03:10:35.670
artificial intelligence safe and better,
to put it to work for humanitarian

cue-2362
03:10:35.670 --> 03:10:39.350
purposes, both through grants and, and
initiatives. That's just one part of it,

cue-2363
03:10:39.350 --> 03:10:42.590
and that's a big part of it, but it's not
the whole part of it, because there's also

cue-2364
03:10:42.590 --> 03:10:43.530
a matter of governance.

cue-2365
03:10:44.510 --> 03:10:45.310
The nonprofit

cue-2366
03:10:46.450 --> 03:10:52.270
always has and still does control the
OpenAI for-profit. That control is not

cue-2367
03:10:52.270 --> 03:10:55.920
required. It's not as though there's
anything in the law of charitable trust or

cue-2368
03:10:55.920 --> 03:10:58.610
its custom and practice, as you've heard
about, that requires that.

cue-2369
03:10:59.250 --> 03:11:03.270
Certainly wasn't a duty owed to Elon Musk,
but it's true nonetheless, members of the

cue-2370
03:11:03.270 --> 03:11:07.970
jury. It's a matter of simple corporate
law, as Professor Coates explained to you

cue-2371
03:11:07.970 --> 03:11:09.130
yesterday afternoon.

cue-2372
03:11:10.670 --> 03:11:12.410
Now, Musk complains that

cue-2373
03:11:13.210 --> 03:11:17.950
OpenAI's controlled nonprof-- for-profit,
uh, for-profit organization that's subject

cue-2374
03:11:17.950 --> 03:11:19.770
to nonprofit control is too big.

cue-2375
03:11:20.430 --> 03:11:24.669
Um, Professor Hemel, who studies nonprofit
organizations, explained that many

cue-2376
03:11:24.670 --> 03:11:29.850
charities affiliated with non-- with for,
with for-profit organizations

cue-2377
03:11:30.450 --> 03:11:34.750
have for-profit organizations that are
very, very large. This is not, at a

cue-2378
03:11:34.750 --> 03:11:38.710
conceptual level, an anomaly. It's not
something that's unknown or odd.

cue-2379
03:11:39.329 --> 03:11:39.510
Um,

cue-2380
03:11:40.530 --> 03:11:45.170
M-M-Microsoft partnerships are also
consistent common practice amongst large

cue-2381
03:11:45.170 --> 03:11:50.790
United States charities. Nonprofits can
and do use for-profit affiliations to

cue-2382
03:11:50.790 --> 03:11:55.210
advance their missions. This is the custom
and practice of nonprofit law. They can

cue-2383
03:11:55.210 --> 03:12:00.350
structure partnerships with commercial
partners. This is nonprofit law. The

cue-2384
03:12:00.350 --> 03:12:04.270
question here is whether the arrangement
advances the nonprofit's mission. That's

cue-2385
03:12:04.270 --> 03:12:06.970
the question, and the answer is, of
course, it does.

cue-2386
03:12:07.690 --> 03:12:12.770
When you think through the complaint that
the nonprofit is too small and the

cue-2387
03:12:12.770 --> 03:12:14.270
for-profit is too big, it,

cue-2388
03:12:15.110 --> 03:12:19.610
it doesn't really make a lot of sense. If
OpenAI's for-profit were less valuable,

cue-2389
03:12:20.720 --> 03:12:25.160
all that would mean is less for the
foundation, that the foundation would have

cue-2390
03:12:25.160 --> 03:12:26.860
fewer charitable resources.

cue-2391
03:12:27.990 --> 03:12:32.130
And if the governance rights weren't as
strong, that would mean there was less

cue-2392
03:12:32.130 --> 03:12:34.670
influence over the shape of AGI.

cue-2393
03:12:35.510 --> 03:12:37.770
None of that is consistent with the
charitable mission.

cue-2394
03:12:40.010 --> 03:12:44.770
All of it, all of it is entirely in line
with the charitable mission. And yet, what

cue-2395
03:12:44.770 --> 03:12:50.490
Mr. Musk seems to want to achieve is a
scenario where the foundation has fewer

cue-2396
03:12:50.490 --> 03:12:52.090
assets and is

cue-2397
03:12:52.850 --> 03:12:58.030
exercising less control over leading-edge
artificial intelligence. It makes no sense

cue-2398
03:12:58.030 --> 03:13:02.989
from a mission perspective. Importantly,
moreover, you've heard that the

cue-2399
03:13:02.990 --> 03:13:07.190
nonprofit's current structure, as Mr.
Taylor testified, was shaped in

cue-2400
03:13:07.190 --> 03:13:10.930
consultation with the attorneys general of
Delaware and California, who were

cue-2401
03:13:10.930 --> 03:13:15.850
actively involved in discussions about how
to structure the organization in twenty

cue-2402
03:13:15.850 --> 03:13:16.530
twenty-five.

cue-2403
03:13:17.150 --> 03:13:24.130
That's just because OpenAI is a nonprofit,
and no other AI company can say that. No

cue-2404
03:13:24.130 --> 03:13:29.650
other AI company is organized with a
nonprofit in control. No other AI company

cue-2405
03:13:29.650 --> 03:13:33.570
answers to government regulators like the
attorneys general, certainly not Mr.

cue-2406
03:13:33.570 --> 03:13:35.650
Musk's xAI.

cue-2407
03:13:37.490 --> 03:13:39.950
And m-m-- Counsel Mr. Mullow touts

cue-2408
03:13:40.750 --> 03:13:44.550
Mr. Musk's concern with the social impact
of artificial intelligence.

cue-2409
03:13:45.170 --> 03:13:45.290
Um,

cue-2410
03:13:46.110 --> 03:13:50.330
the whole point of the resilience project,
which will now be the beneficiary of

cue-2411
03:13:50.330 --> 03:13:55.770
billions of dollars of action and
resources, is precisely to mitigate the

cue-2412
03:13:55.770 --> 03:14:00.150
economic effects, the adverse economic
effects of artificial intelligence, and to

cue-2413
03:14:00.150 --> 03:14:05.070
ensure positive ef-- impacts in
artificial intelligence. This isn't gonna

cue-2414
03:14:05.070 --> 03:14:08.790
be easy, and it's not gonna be without its
challenges, but someone has to devote the

cue-2415
03:14:08.790 --> 03:14:13.190
resources to try and get it right. That's
what OpenAI is doing, not xAI.

cue-2416
03:14:16.050 --> 03:14:20.280
Mr. Musk looks at all this nevertheless
and shouts,

cue-2417
03:14:21.360 --> 03:14:22.190
"They stole a charity."

cue-2418
03:14:23.330 --> 03:14:27.070
That's sort of all he does, is shout,
"They stole a charity," but that has

cue-2419
03:14:27.070 --> 03:14:31.290
things exactly backwards, members of the
jury. For years, the main value

cue-2420
03:14:32.010 --> 03:14:34.830
the nonprofit had was in its residual
interest

cue-2421
03:14:35.630 --> 03:14:40.169
in the old capped profit structure. You
heard Mr. Schizer complain about how

cue-2422
03:14:40.170 --> 03:14:44.530
uncertain that interest was. He said it
made things hard for the nonprofit. It

cue-2423
03:14:44.530 --> 03:14:49.330
wasn't be able-- gonna be able to get its,
its hands on resources quickly, and, and

cue-2424
03:14:49.330 --> 03:14:50.290
that's true. And,

cue-2425
03:14:50.990 --> 03:14:55.410
a-and it's, it's also true that for
several years, the nonprofit's main job

cue-2426
03:14:55.410 --> 03:14:58.690
was overseeing the for-profit. That
remains an important part of the

cue-2427
03:14:58.690 --> 03:15:03.710
nonfunction-- nonprofit function. It
didn't have its resources readily

cue-2428
03:15:03.710 --> 03:15:06.030
available to conduct its own initiatives.

cue-2429
03:15:07.790 --> 03:15:12.730
This is the slide that, um, Professor
Schizer wor-- look, look, looked through

cue-2430
03:15:12.730 --> 03:15:17.229
with you a little more than a week ago.
Um, but Professor Coates explained the

cue-2431
03:15:17.230 --> 03:15:21.370
actual economics, 'cause this isn't the
actual economics. The actua-actual

cue-2432
03:15:21.370 --> 03:15:22.390
economics are this.

cue-2433
03:15:23.190 --> 03:15:27.890
Professor Schizer imagined an exceedingly
s-small pie-- piece of this very large

cue-2434
03:15:27.890 --> 03:15:28.190
pie.

cue-2435
03:15:28.990 --> 03:15:34.574
In fact, the actual share is enormous.$200
billion plus enormous.

cue-2436
03:15:35.234 --> 03:15:41.134
That is not stealing from a charity. That
is creating untold v-value for a charity.

cue-2437
03:15:41.134 --> 03:15:44.094
That's creating a three hundred and
thirty-three thousand

cue-2438
03:15:44.934 --> 03:15:47.134
times rate of return for a charity.

cue-2439
03:15:50.334 --> 03:15:53.774
Now, since the October 2025
recapitalization, about which you've heard

cue-2440
03:15:53.774 --> 03:15:56.654
a bit, the foundation, the nonprofit, has
been,

cue-2441
03:15:57.294 --> 03:15:57.474
um,

cue-2442
03:15:59.894 --> 03:16:04.614
with the good fortune of owning a very big
chunk of the PBC, and it finally has

cue-2443
03:16:04.614 --> 03:16:09.714
access to the resources it needs to fund
the initiatives that we've talked about. I

cue-2444
03:16:09.714 --> 03:16:13.754
just wanna say it again, because the sum
available to this organization is

cue-2445
03:16:13.754 --> 03:16:20.094
enormous, $200 billion plus. And I know
that in questions to witnesses and the

cue-2446
03:16:20.094 --> 03:16:25.934
argument with you this morning, Mr. Mollo
has suggested that the foundation only

cue-2447
03:16:25.934 --> 03:16:30.254
started hiring people recently, that it
only began to do the work that it's doing

cue-2448
03:16:30.254 --> 03:16:37.054
because of this lawsuit. There is zero
truth to that. And putting aside

cue-2449
03:16:37.054 --> 03:16:40.934
what's been said by a lawyer, which as the
judge has explained to you isn't

cue-2450
03:16:40.934 --> 03:16:44.954
evidence, there is no evidence of that.
There's only one explanation in the

cue-2451
03:16:44.954 --> 03:16:49.224
evidence, and it's the accurate one, which
is that the recapitalization transaction,

cue-2452
03:16:49.224 --> 03:16:50.064
the restructuring,

cue-2453
03:16:51.154 --> 03:16:56.594
consummated in October 2025, unlocked the
value that the OpenAI Foundation could put

cue-2454
03:16:56.594 --> 03:17:02.134
to work and that it immediately set about
to putting those resources to work. It is

cue-2455
03:17:02.134 --> 03:17:05.794
true that there has been a rapid
acceleration in the foundation's work.

cue-2456
03:17:06.474 --> 03:17:10.874
The only reason for that is that it now
has the resources to do it. The evidence

cue-2457
03:17:10.874 --> 03:17:15.354
on that question, members of the jury, is
one way. There are no-- there's nothing

cue-2458
03:17:15.354 --> 03:17:17.874
from which one can draw an alternative
inference.

cue-2459
03:17:21.314 --> 03:17:27.614
In no way, members of the jury, did Greg
Brockman or Sam Altman or Ilya Sutskever

cue-2460
03:17:28.574 --> 03:17:33.474
or anyone else who helped build the value
that now resides in the for-profit steal

cue-2461
03:17:33.474 --> 03:17:36.254
anything from the nonprofit. Against very
tall odds,

cue-2462
03:17:36.974 --> 03:17:41.374
their work and sweat and innovation
created something, um, incredibly

cue-2463
03:17:41.374 --> 03:17:45.894
valuable, startlingly valuable, admittedly
startlingly valuable for the nonprofit.

cue-2464
03:17:46.514 --> 03:17:49.374
That's where the nonprofit's entire
endowment comes from.

cue-2465
03:17:50.594 --> 03:17:50.734
And

cue-2466
03:17:51.554 --> 03:17:56.274
on, on the subject further of, of
irrelevant things that Mr. Musk has sought

cue-2467
03:17:56.274 --> 03:18:01.094
to make this case about, let me briefly
address the removal of Sam Altman and Greg

cue-2468
03:18:01.094 --> 03:18:04.854
Brockman from OpenAI's board in November
2023. For sure,

cue-2469
03:18:05.694 --> 03:18:09.634
um, members of the jury, this was a
difficult time in OpenAI's history.

cue-2470
03:18:10.354 --> 03:18:14.194
Uh, for sure, several members of the board
were unhappy with Sam Altman's

cue-2471
03:18:14.194 --> 03:18:18.934
management. Uh, some for sure accused him
of being less than candid,

cue-2472
03:18:19.814 --> 03:18:24.874
principally in the sense of having said
two different people two different things

cue-2473
03:18:24.874 --> 03:18:25.774
on the same topic.

cue-2474
03:18:26.594 --> 03:18:26.714
But

cue-2475
03:18:27.894 --> 03:18:29.834
let's keep an eye on the ball of how
things shook out.

cue-2476
03:18:31.114 --> 03:18:35.674
Over ninety percent of OpenAI's employees,
including plaintiff's own witness, Rosie

cue-2477
03:18:35.674 --> 03:18:40.374
Campbell, including Dr. Sutskever, signed
a petition to have Mr. Altman and Mr.

cue-2478
03:18:40.374 --> 03:18:45.274
Brockman reinstated. These are the people
who worked most closely with these

cue-2479
03:18:45.274 --> 03:18:49.874
executives day in and day out over the
course of years and years. They wanted

cue-2480
03:18:49.874 --> 03:18:55.354
them back, both of them. Dr. Sutskever,
one of the members of the old OpenAI

cue-2481
03:18:55.354 --> 03:19:00.084
board, signed that petition, and he voted
to have Altman and Brockman return.

cue-2482
03:19:01.094 --> 03:19:06.134
You heard Dr. Sutskever testify that the
old board had acted rashly, without

cue-2483
03:19:06.134 --> 03:19:10.494
adequate inquiry, without a succession
plan in place, without adequate

cue-2484
03:19:10.494 --> 03:19:15.734
investigation on the strength of poor
legal advice, so as to have created chaos

cue-2485
03:19:15.734 --> 03:19:16.394
at the company.

cue-2486
03:19:17.114 --> 03:19:18.554
That was from Dr. Sutskever.

cue-2487
03:19:19.254 --> 03:19:23.614
Ms. McCauley and Ms. Toner, two of the
other directors who'd done the removals,

cue-2488
03:19:23.614 --> 03:19:27.934
they agreed that Alt-- bringing Altman
back was critical for the future of the

cue-2489
03:19:27.934 --> 03:19:32.974
company. Adam, Adam D'Angelo, the fourth
vote, well, he stayed on the board. He's

cue-2490
03:19:32.974 --> 03:19:36.004
still on the board. He voted to reinstate
Mr. Altman.

cue-2491
03:19:36.694 --> 03:19:40.043
He voted to reinstate him as both a board
member and a CEO.

cue-2492
03:19:41.374 --> 03:19:44.994
The new board members, Mr. Taylor, the
chair from whom you've heard, and Larry

cue-2493
03:19:44.994 --> 03:19:49.854
Summers, they joined the board. They were
both contacted first by the outgoing

cue-2494
03:19:49.854 --> 03:19:54.274
board, not by Altman. Summers was put on
the board over Microsoft's objection.

cue-2495
03:19:55.314 --> 03:19:58.814
And importantly, as the outgoing board
demanded,

cue-2496
03:19:59.934 --> 03:20:03.354
a thorough independent review was
conducted into just what happened in

cue-2497
03:20:03.354 --> 03:20:08.194
November 2023, and after that review was
completed, Mr. Altman was reinstated to

cue-2498
03:20:08.194 --> 03:20:09.094
the board of directors.

cue-2499
03:20:11.494 --> 03:20:12.914
Sam Altman and Greg Brockman,

cue-2500
03:20:13.574 --> 03:20:14.634
and OpenAI in general,

cue-2501
03:20:15.534 --> 03:20:15.674
have,

cue-2502
03:20:17.134 --> 03:20:19.894
have had to sort out what happened in
something that was

cue-2503
03:20:20.634 --> 03:20:22.764
a, a big blowup, a lot of chaos,

cue-2504
03:20:23.674 --> 03:20:27.704
and they had to sort out what to do to try
to make things better going forward.

cue-2505
03:20:28.454 --> 03:20:32.774
In place now is a more robust, more mature
governance structure that reflects the

cue-2506
03:20:32.774 --> 03:20:33.614
growth of the company.

cue-2507
03:20:34.514 --> 03:20:39.174
Brockman and Altman and the management
team and this board, which as you saw, is

cue-2508
03:20:39.174 --> 03:20:43.254
populated with exceedingly accomplished
individuals from all walks of life

cue-2509
03:20:43.254 --> 03:20:44.444
relevant to this project,

cue-2510
03:20:45.134 --> 03:20:49.104
um, have tried, and they're trying, and we
think the evidence is they're succeeding.

cue-2511
03:20:49.104 --> 03:20:50.954
Now,

cue-2512
03:20:52.594 --> 03:20:53.774
you may be asking yourself,

cue-2513
03:20:54.414 --> 03:21:00.314
w-what does all of this have to do with
Mr. Musk's claim that in 2015 and 2017, he

cue-2514
03:21:00.314 --> 03:21:04.294
was made some promise that was broken?
This all happened in 2023, and the answer

cue-2515
03:21:04.294 --> 03:21:05.874
to that question is nothing. Um,

cue-2516
03:21:06.914 --> 03:21:11.134
M-Musk's last rent donation was made in
2020, and we'll talk a bit about that.

cue-2517
03:21:11.134 --> 03:21:15.084
That was three years before all of this.
His last quarterly donation was in

cue-2518
03:21:15.894 --> 03:21:19.734
the middle of 2017, as you now know. It
was more than six years before all of

cue-2519
03:21:19.734 --> 03:21:19.913
this.

cue-2520
03:21:20.554 --> 03:21:24.314
The alleged promises that were made were
in the remote past.

cue-2521
03:21:24.994 --> 03:21:30.954
Mr. Musk is raising this issue to divert
your attention from the facts that matter

cue-2522
03:21:30.954 --> 03:21:37.906
for his claim.But since Mr. Musk has made
so much of November 2020, if you

cue-2523
03:21:37.906 --> 03:21:42.716
ask yourself this, if Mr. Altman wasn't
committed to OpenAI and its mission,

cue-2524
03:21:44.006 --> 03:21:45.246
which is what Mr. Musk claims,

cue-2525
03:21:46.386 --> 03:21:48.626
why did he come back to OpenAI after being
fired?

cue-2526
03:21:49.566 --> 03:21:53.946
He didn't have to do that. You heard that
he had a job offer at Microsoft. You know,

cue-2527
03:21:53.946 --> 03:21:57.506
'cause you heard, that he could have gone
there and made an enormous amount of

cue-2528
03:21:57.506 --> 03:22:01.446
money. He could have gone there with most
of his team. He could have pursued AI

cue-2529
03:22:01.446 --> 03:22:01.906
there.

cue-2530
03:22:02.806 --> 03:22:04.346
He wouldn't have had to think about the
mission again.

cue-2531
03:22:05.686 --> 03:22:07.286
But he came back to OpenAI.

cue-2532
03:22:08.266 --> 03:22:12.406
This is the proof of it. Here was a guy
who had a choice, a clear field to walk

cue-2533
03:22:12.406 --> 03:22:16.496
away, to get into an organization that was
gonna pay him enormously. He'd

cue-2534
03:22:17.106 --> 03:22:21.046
run the show, and he would not be
constrained by the mission. He said no. He

cue-2535
03:22:21.046 --> 03:22:24.786
went back to OpenAI because he cared
about the team, and he cared about the

cue-2536
03:22:24.786 --> 03:22:28.516
mission. He re-upped in circumstances
where he didn't have any obligation 'cause

cue-2537
03:22:28.516 --> 03:22:30.946
he'd just been kicked out. He'd just been
kicked out on his ear.

cue-2538
03:22:36.186 --> 03:22:37.346
Mr. Musk argues,

cue-2539
03:22:38.066 --> 03:22:41.826
tells you, members of the jury, that these
events are relevant to his claim that

cue-2540
03:22:41.826 --> 03:22:42.826
OpenAI

cue-2541
03:22:42.866 --> 03:22:48.086
promised him to put safety first, and the
events of November 2023 aren't consistent

cue-2542
03:22:48.086 --> 03:22:49.246
with that, but none of that is true.

cue-2543
03:22:50.106 --> 03:22:55.316
OpenAI is and has been committed to
safety. That's not some special obligation

cue-2544
03:22:55.316 --> 03:22:58.286
that's owed to Elon Musk. It's
fundamental to the mission of the

cue-2545
03:22:58.286 --> 03:23:03.426
organization. And the events of November
2023, they were not about safety, members

cue-2546
03:23:03.426 --> 03:23:07.626
of the jury. Uh, the board's own
statements from the time say it in black

cue-2547
03:23:07.626 --> 03:23:11.336
and white. Let's have a look at the
statement that was issued. This decision,

cue-2548
03:23:13.906 --> 03:23:16.726
the decision to, to remove Altman and
Brockman,

cue-2549
03:23:17.366 --> 03:23:19.076
is not about product safety,

cue-2550
03:23:19.826 --> 03:23:20.846
not about security,

cue-2551
03:23:21.586 --> 03:23:26.526
not about the pace of development or
OpenAI's finances. It wasn't about those

cue-2552
03:23:26.526 --> 03:23:31.286
things. It was about personality conflicts
and miscommunications that I don't wanna

cue-2553
03:23:31.286 --> 03:23:33.706
minimize them. It was a, it was a blowup,

cue-2554
03:23:34.866 --> 03:23:38.166
needed to be taken care of so it won't
happen again, but it wasn't about safety,

cue-2555
03:23:38.166 --> 03:23:40.946
and if Mr. Musk's telling you the tr--
that, then he's not telling you the truth.

cue-2556
03:23:40.946 --> 03:23:46.705
And as for h- where things stand today,
you heard from Zico Kolter, the chair of

cue-2557
03:23:46.706 --> 03:23:48.786
Carnegie Mellon's Machine Learning
Department.

cue-2558
03:23:49.666 --> 03:23:55.106
He's an OpenAI Foundation director. He's
one of the preeminent experts in AI safety

cue-2559
03:23:55.106 --> 03:24:00.045
in the world, uh, and he testified, as
you heard, that the organization has a

cue-2560
03:24:00.046 --> 03:24:01.746
robust nonprofit-only

cue-2561
03:24:04.146 --> 03:24:07.986
procedure in place for ensuring the
rigorous review from a safety standpoint

cue-2562
03:24:07.986 --> 03:24:09.326
of everything OpenAI does.

cue-2563
03:24:10.686 --> 03:24:10.846
So

cue-2564
03:24:11.506 --> 03:24:16.846
Musk's breach of charitable trust claim
fails for multiple independent reasons. He

cue-2565
03:24:16.846 --> 03:24:22.146
has not proven that he created a trust.
Ms. Eddy demonstrated that. He has not

cue-2566
03:24:22.146 --> 03:24:26.986
proven that he created a trust. Any trust
must have ended in September of 2020 when

cue-2567
03:24:26.986 --> 03:24:30.606
the last dime of Mr. Musk's donations were
spent,

cue-2568
03:24:31.406 --> 03:24:34.916
and there is no evidence that his
donations were misused. Indeed,

cue-2569
03:24:35.686 --> 03:24:38.646
and it's an obvious fact, it's not in
dispute, but it needs to be said over and

cue-2570
03:24:38.646 --> 03:24:44.206
over again because of the rhetoric from
the other side, OpenAI remains a charity.

cue-2571
03:24:44.206 --> 03:24:49.706
OpenAI remains governed by a charity. The
charity is now bigger and stronger and

cue-2572
03:24:49.706 --> 03:24:52.446
more empowered from a governance
perspective than ever.

cue-2573
03:24:57.666 --> 03:25:01.146
I'd wanted members of the jury now to turn
from the charitable trust claim, and I

cue-2574
03:25:01.146 --> 03:25:04.786
think we talked about all the things that
are relevant there and discuss

cue-2575
03:25:05.666 --> 03:25:08.586
the second claim in the case, unjust
enrichment. Um,

cue-2576
03:25:09.586 --> 03:25:10.346
and I think,

cue-2577
03:25:10.946 --> 03:25:13.985
as Mr. Mollo said, on this much we can
agree, I think we can move a little bit

cue-2578
03:25:13.986 --> 03:25:17.926
faster through this 'cause the facts
underlying this are mostly the same. So

cue-2579
03:25:17.926 --> 03:25:20.906
we'll talk a bit about some of the same
ones, but we can, we can move more

cue-2580
03:25:20.906 --> 03:25:21.366
quickly.

cue-2581
03:25:22.106 --> 03:25:26.146
I did wanna say a word or two at the
outset about what restitution is.

cue-2582
03:25:27.506 --> 03:25:29.886
The idea of restitution...

cue-2583
03:25:31.366 --> 03:25:36.886
The idea of restitution is that the
plaintiff can return, demand the return of

cue-2584
03:25:36.886 --> 03:25:38.046
something that he or she gave.

cue-2585
03:25:38.966 --> 03:25:39.126
Um,

cue-2586
03:25:39.886 --> 03:25:43.746
and so it's worth pausing for a moment on
the narrative that Mr. Mollo has advanced

cue-2587
03:25:43.746 --> 03:25:46.126
as the basis for the claim. The idea

cue-2588
03:25:46.886 --> 03:25:52.686
is that OpenAI is what it is because of
Elon Musk, that Elon founded it, that

cue-2589
03:25:52.686 --> 03:25:56.886
Elon's ideas and his brilliance and his
connections, the Midas touch of Elon Musk

cue-2590
03:25:56.886 --> 03:26:00.346
is what made OpenAI what it is today.
That's the thesis.

cue-2591
03:26:01.066 --> 03:26:04.826
In his claims here, Mr. Musk wants to
claim credit for the whole thing,

cue-2592
03:26:05.506 --> 03:26:08.306
for all that OpenAI is. Elon, Elon, Elon.

cue-2593
03:26:08.986 --> 03:26:09.606
It was all me.

cue-2594
03:26:12.846 --> 03:26:16.726
Mr. Musk wants all that credit, but he
hasn't earned all that credit.

cue-2595
03:26:19.046 --> 03:26:21.466
For sure, he made some contributions early
on,

cue-2596
03:26:22.306 --> 03:26:22.566
um,

cue-2597
03:26:23.566 --> 03:26:26.666
but those contributions, remember, they
were all charitable donations.

cue-2598
03:26:27.266 --> 03:26:30.586
Mr. Musk helped himself to millions in tax
deductions for those donations.

cue-2599
03:26:34.046 --> 03:26:34.176
And

cue-2600
03:26:35.506 --> 03:26:38.406
charitable donations are not seed money,

cue-2601
03:26:39.486 --> 03:26:42.766
uh, as Mr. Musk started claiming in
October 2022.

cue-2602
03:26:43.506 --> 03:26:46.906
You don't see any of the other donors, and
we'll talk about them for a second, who

cue-2603
03:26:46.906 --> 03:26:51.146
gave far more in the aggregate than Mr.
Musk, raising their hand for a seed money

cue-2604
03:26:51.146 --> 03:26:51.476
payout.

cue-2605
03:26:52.246 --> 03:26:55.406
Um, it's an entire-- It's, it's really
quite a cynical position if you think

cue-2606
03:26:55.406 --> 03:27:00.566
about it. Mr. Musk is comparing himself
here to venture capital fund, f- uh,

cue-2607
03:27:00.566 --> 03:27:03.286
funders who make an investment and have
the

cue-2608
03:27:04.106 --> 03:27:08.666
opportunity to, to reap a reward. But this
was supposed to be a donation. He took a

cue-2609
03:27:08.666 --> 03:27:12.846
tax deduction, and the whole theory of his
claim is that there's a charitable

cue-2610
03:27:12.846 --> 03:27:13.086
object.

cue-2611
03:27:18.426 --> 03:27:23.006
The whole theory of his... Thank you, Your
Honor. The whole theory of his claim is

cue-2612
03:27:23.006 --> 03:27:28.046
that this is a charity, and yet he wants,
when it suits him, to talk about himself

cue-2613
03:27:28.046 --> 03:27:29.106
as a seed investor.

cue-2614
03:27:31.066 --> 03:27:34.926
And as you heard from witness after
witness after witness, Mr. Musk abandoned

cue-2615
03:27:34.926 --> 03:27:41.266
OpenAI for dead in 2018. Here it is from
Mr. Musk himself. It's DX853, and of

cue-2616
03:27:41.266 --> 03:27:44.546
course, this will be one of the
documentsThat's with you as you consider

cue-2617
03:27:44.546 --> 03:27:45.126
our claim.

cue-2618
03:27:46.066 --> 03:27:50.706
My probability assessment of OpenAI being
relevant to DeepMind Google withou-without

cue-2619
03:27:50.706 --> 03:27:54.466
a dramatic change in execution and
resources is zero percent.

cue-2620
03:27:55.566 --> 03:27:59.086
Wanting to make sure he's mathematically
precise, he adds, "It's not one percent,

cue-2621
03:27:59.746 --> 03:28:00.846
it's zero percent."

cue-2622
03:28:01.546 --> 03:28:04.905
He thought that's what OpenAI's chances of
success were, were. But, but, but then

cue-2623
03:28:04.906 --> 03:28:06.096
after Mr. Musk left,

cue-2624
03:28:07.006 --> 03:28:08.446
slamming his door on the way out,

cue-2625
03:28:09.606 --> 03:28:12.566
seeking to poach employees on the way out,

cue-2626
03:28:12.926 --> 03:28:17.506
seeking to build his own A-AGI operation
at Tesla on the way out, after he slammed

cue-2627
03:28:17.506 --> 03:28:22.356
the door behind him, an unexpected thing
happened. The organization began to

cue-2628
03:28:22.356 --> 03:28:22.786
flourish.

cue-2629
03:28:25.426 --> 03:28:29.906
Milestone upon technical milestone
occurred after February two thousand

cue-2630
03:28:29.906 --> 03:28:32.286
eighteen, after Mr. Musk stormed out the
door.

cue-2631
03:28:36.446 --> 03:28:41.586
Only after February two thousand eighteen
did the technology advance such that the

cue-2632
03:28:41.586 --> 03:28:47.466
ant became the cat. And you will recall
Dr. Sutskever giving the analogy of the

cue-2633
03:28:47.466 --> 03:28:51.305
growth of this industry in the manner
that's depicted on this screen.

cue-2634
03:28:52.426 --> 03:28:54.776
And as Dr. Sutskever also testified,

cue-2635
03:28:58.666 --> 03:29:03.526
the funds that OpenAI has raised were all
necessary to take advantage of this

cue-2636
03:29:03.526 --> 03:29:05.006
technological progression.

cue-2637
03:29:12.006 --> 03:29:15.306
Um, I wanna talk a little bit about
th-this slide with, with, with you members

cue-2638
03:29:15.306 --> 03:29:20.666
of the jury. Um, for the thirty-eight
million dollars that Elon Musk donated way

cue-2639
03:29:20.666 --> 03:29:25.156
back in the day, others contributed
nearly one hundred million dollars. And,

cue-2640
03:29:25.156 --> 03:29:29.806
and remember, this is... We talked about
this when we first met several weeks ago.

cue-2641
03:29:29.806 --> 03:29:32.906
This is one of those dogs that doesn't
bark. Just like you'd think there'd be a

cue-2642
03:29:32.906 --> 03:29:37.386
document memorializing the agreement that
Mr. Musk said he had, you'd think that if

cue-2643
03:29:37.386 --> 03:29:41.206
this was really a violation of the m-- of
the mission of this organization the way

cue-2644
03:29:41.206 --> 03:29:44.146
it's been run, you'd think one of these
donors, a lot of sophisticated people gave

cue-2645
03:29:44.146 --> 03:29:46.046
a lot of money, would have raised their
hands.

cue-2646
03:29:46.786 --> 03:29:51.466
But they didn't. Not a single one of these
other donors has ever complained that

cue-2647
03:29:51.466 --> 03:29:55.906
OpenAI has failed to respect its mission.
Not a single one of these other donors

cue-2648
03:29:56.746 --> 03:30:01.006
has brought a lawsuit against M-- against
OpenAI. Only Mr. Musk has, only the

cue-2649
03:30:01.006 --> 03:30:04.526
competitor. That's the only one who's
brought it. That's the only one of these

cue-2650
03:30:04.526 --> 03:30:05.686
donors who's perceived

cue-2651
03:30:06.406 --> 03:30:11.026
this unwritten, unspoken promise that was
allegedly made.

cue-2652
03:30:12.306 --> 03:30:16.446
And then investors put in a hundred and
fifty billion dollars in investment

cue-2653
03:30:16.446 --> 03:30:17.686
capital, not donations now,

cue-2654
03:30:18.526 --> 03:30:19.436
in investment capital,

cue-2655
03:30:20.306 --> 03:30:24.546
all needed to power the mission. And s-so
members of the jury have this in view.

cue-2656
03:30:24.546 --> 03:30:29.006
Those investors, they still haven't had a
cent of return from that. So keep that in

cue-2657
03:30:29.006 --> 03:30:33.205
mind. None of those investments have been
paid out to, to-- None of those

cue-2658
03:30:33.206 --> 03:30:36.716
investments have been paid out by OpenAI
'cause it's not profitable. Now,

cue-2659
03:30:37.326 --> 03:30:43.086
Mr. Mollo, in his conversation with, with
you this morning, conjured some imaginary

cue-2660
03:30:43.086 --> 03:30:48.626
world in which OpenAI could have developed
without the Microsoft investment, some

cue-2661
03:30:48.626 --> 03:30:51.046
imaginary w-world in which

cue-2662
03:30:51.766 --> 03:30:53.916
s-s-- donations equivalent to what

cue-2663
03:30:54.746 --> 03:30:56.686
Microsoft gave could have been,

cue-2664
03:30:58.106 --> 03:31:00.146
could have been solicited and brought in
and

cue-2665
03:31:00.786 --> 03:31:02.606
m-meet the do-- meet the demands of the
project.

cue-2666
03:31:03.846 --> 03:31:08.006
A couple of things on this. You heard Mr.
Altman talk about beating the bushes, and

cue-2667
03:31:08.006 --> 03:31:11.366
he raised fifty million dollars in two
thousand eighteen. That's a lot of money.

cue-2668
03:31:11.366 --> 03:31:14.486
It's a lot of work, but it's nowhere near
what was required.

cue-2669
03:31:15.246 --> 03:31:17.866
Remember also that there was supposed to
be a big anchor donor,

cue-2670
03:31:18.546 --> 03:31:21.246
Mr. Musk, a billion dollars. Well, that
didn't happen.

cue-2671
03:31:22.106 --> 03:31:27.096
And there is literally no evidence,
literally no evidence, that

cue-2672
03:31:27.886 --> 03:31:32.146
anything like the amount necessary to
sustain this organization could have been

cue-2673
03:31:32.146 --> 03:31:34.106
raised by donations. The best

cue-2674
03:31:37.346 --> 03:31:41.446
Mr. Musk's team could come up with is an
organization that raises a great deal of

cue-2675
03:31:41.446 --> 03:31:43.246
money a year, said to be over four
billion.

cue-2676
03:31:44.046 --> 03:31:47.646
But you heard from the experts yesterday,
that's actually a hundred million dollars

cue-2677
03:31:47.646 --> 03:31:53.386
a year of cash and leftovers from soup,
from restaurants and grocery stores. It's

cue-2678
03:31:53.386 --> 03:31:59.766
all in-kind contributions. There is no
historical precedent, no universe in which

cue-2679
03:31:59.766 --> 03:32:03.926
the capital resources of this
organization could have been funded from

cue-2680
03:32:04.566 --> 03:32:10.045
private donations. So yeah, yeah, for
sure, Mr. Musk came by every week or so,

cue-2681
03:32:10.046 --> 03:32:12.266
he gave some advice, and then he quit two
years in.

cue-2682
03:32:13.986 --> 03:32:14.576
A-And after that,

cue-2683
03:32:15.266 --> 03:32:16.286
Mr. Musk jumped ship,

cue-2684
03:32:18.226 --> 03:32:20.376
and after-- and he tried to take some with
him,

cue-2685
03:32:21.406 --> 03:32:25.556
but a lot of people stayed, hundreds
stayed, and then thousands were there. And

cue-2686
03:32:25.556 --> 03:32:28.845
there are researchers and engineers who
have poured their life into this

cue-2687
03:32:28.846 --> 03:32:34.486
organization, and it is they, members of
the jury, not Elon Musk, who have created

cue-2688
03:32:34.486 --> 03:32:40.226
something enormously and surprisingly
valuable. Mr. Musk is not the one who

cue-2689
03:32:40.226 --> 03:32:45.786
built OpenAI into what it is today. OpenAI
is what it is today despite his

cue-2690
03:32:45.786 --> 03:32:52.786
predictions of certain doom, despite his
po-poaching of key employees, despite his

cue-2691
03:32:52.786 --> 03:32:57.386
attempts to absorb it into his own
organization, his own for-profit car

cue-2692
03:32:57.386 --> 03:33:02.026
company, and despite his public efforts to
tear it down. And as for the Midas touch,

cue-2693
03:33:02.026 --> 03:33:04.666
not really, not here.

cue-2694
03:33:05.606 --> 03:33:09.126
Consider what happened when Musk tried to
build AGI at Tesla.

cue-2695
03:33:10.646 --> 03:33:13.306
It was a total fail, as my daughter would
say.

cue-2696
03:33:13.946 --> 03:33:20.686
Mr. Musk may have the Midas touch in some
areas, but not in AI. This requires a

cue-2697
03:33:20.686 --> 03:33:24.706
touch that he doesn't have. This is not a
bulldozer business.

cue-2698
03:33:25.446 --> 03:33:28.196
To succeed in AI, as it turns out,

cue-2699
03:33:28.966 --> 03:33:31.266
all Mr. Musk can do is come to court.

cue-2700
03:33:38.106 --> 03:33:42.546
So that's one thing. OpenAI is not Mr.
Musk's creature, far from it, but even if

cue-2701
03:33:42.546 --> 03:33:46.346
you accept that narrativeIt still doesn't
give him a legal claim, members of the

cue-2702
03:33:46.346 --> 03:33:50.166
jury. So let's get detailed about what Mr.
Musk has to do to prove his claim.

cue-2703
03:33:51.046 --> 03:33:53.746
For unjust enrichment, Mr. Musk has named
four defendants:

cue-2704
03:33:54.576 --> 03:33:55.786
the nonprofit itself,

cue-2705
03:33:57.866 --> 03:33:58.706
the PBC,

cue-2706
03:33:59.826 --> 03:34:00.526
Mr. Altman,

cue-2707
03:34:01.226 --> 03:34:01.976
and Mr. Brockman.

cue-2708
03:34:02.586 --> 03:34:06.766
And to carry his burden on that claim
against any one of these defendants,

cue-2709
03:34:07.786 --> 03:34:12.166
Musk has to show two things for each of
them. He has to show that the defendant

cue-2710
03:34:12.166 --> 03:34:13.226
received a benefit,

cue-2711
03:34:14.906 --> 03:34:19.226
and he has to secondly show that the
defendant unjustly retained the benefit at

cue-2712
03:34:19.226 --> 03:34:20.886
Mr. Musk's expense.

cue-2713
03:34:21.686 --> 03:34:23.366
Let's, let's quickly, um,

cue-2714
03:34:24.446 --> 03:34:28.006
talk about how this analysis plays out
against the backdrop of the evidence,

cue-2715
03:34:28.006 --> 03:34:28.946
against each of these,

cue-2716
03:34:29.636 --> 03:34:32.386
uh, each of these defendants, 'cause
members of the jury, that's what you're

cue-2717
03:34:32.386 --> 03:34:35.046
gonna have to do when you go and sit
together and, and s- figure out how the

cue-2718
03:34:35.046 --> 03:34:37.936
pieces fit together. Um, starting with the
OpenAI Foundation,

cue-2719
03:34:38.926 --> 03:34:39.186
I think

cue-2720
03:34:40.326 --> 03:34:44.346
w- we, we can agree that the OpenAI
Foundation, previously OpenAI Inc.,

cue-2721
03:34:44.346 --> 03:34:46.006
received a benefit from Mr. Musk.

cue-2722
03:34:46.626 --> 03:34:50.086
Um, back in twenty sixteen to twenty
twenty, it got-- it received donations

cue-2723
03:34:50.086 --> 03:34:51.196
from him. Um,

cue-2724
03:34:54.026 --> 03:34:54.246
so,

cue-2725
03:34:55.276 --> 03:35:00.046
uh, but, but, but the problem with that
is, is he, he, he has to show that those

cue-2726
03:35:00.046 --> 03:35:05.486
benefits were unjustly retained at his
expense. It's not enough to say, "I gave

cue-2727
03:35:05.486 --> 03:35:09.586
money." He has to show that there were
benefits that were unjustly retained and

cue-2728
03:35:09.586 --> 03:35:13.066
at his expense, 'cause otherwise anyone
who gave money to a charity could bring a

cue-2729
03:35:13.066 --> 03:35:16.386
lawsuit and tell it what to do, and you
know that's not the law. And there are two

cue-2730
03:35:16.386 --> 03:35:20.606
pieces to this, Your Honor, uh, members
of the jury. Um, one is unjust enrichment,

cue-2731
03:35:21.266 --> 03:35:25.346
and the other is at Musk's expense. I
wanna break those pieces down to, to try

cue-2732
03:35:25.346 --> 03:35:28.326
and roadmap this for you. Let me start
with the second piece,

cue-2733
03:35:28.926 --> 03:35:32.886
um, which is a slightly technical piece.
Um, I'll, I'll warn you in advance, but

cue-2734
03:35:32.886 --> 03:35:33.506
it's important.

cue-2735
03:35:34.646 --> 03:35:39.466
Um, was the alleged unjust reten-
retention at Musk's expense?

cue-2736
03:35:40.586 --> 03:35:45.506
All of the evidence in this case shows
that Musk did not give money to OpenAI

cue-2737
03:35:45.506 --> 03:35:47.946
directly. He did not give it to it
directly.

cue-2738
03:35:48.706 --> 03:35:52.786
M- Mr. Musk donated money to one
organization, and that organization then

cue-2739
03:35:52.786 --> 03:35:54.646
paid the money out through intermediaries.

cue-2740
03:35:55.386 --> 03:35:56.106
The cash flow

cue-2741
03:35:56.806 --> 03:36:00.866
is set out on this slide. Mr. Musk gave
money to these donor-advised funds and the

cue-2742
03:36:00.866 --> 03:36:04.586
sponsor organization, and it gave the
money to OpenAI.

cue-2743
03:36:05.466 --> 03:36:08.746
These were tax-engineered two-step
transactions

cue-2744
03:36:10.466 --> 03:36:12.036
-w- w- where-
-Objection, Your Honor

cue-2745
03:36:12.036 --> 03:36:12.786
... organized.

cue-2746
03:36:14.386 --> 03:36:16.146
Um, related to the summary judgment.

cue-2747
03:36:21.426 --> 03:36:25.046
Your Honor, we fully understand the
summary judgment ruling. We think it's

cue-2748
03:36:25.046 --> 03:36:27.246
important to have the jury

cue-2749
03:36:28.526 --> 03:36:30.126
ruling this issue as a matter of fact.

cue-2750
03:36:32.086 --> 03:36:36.106
Well, they can understand it. I, I, I
thought that the objection was gonna be

cue-2751
03:36:36.106 --> 03:36:40.266
with respect to tax-engineered. I mean,
these are legitimate transactions.

cue-2752
03:36:41.646 --> 03:36:43.566
I don't exactly know what you mean by
that,

cue-2753
03:36:44.406 --> 03:36:44.686
um.

cue-2754
03:36:45.906 --> 03:36:48.366
M- my, my point, members of the jury, is
that, um,

cue-2755
03:36:49.066 --> 03:36:53.946
Musk lost ownership of these transactions
at the first step.

cue-2756
03:36:54.746 --> 03:36:56.706
It wasn't his money anymore. He gave it
away,

cue-2757
03:36:57.406 --> 03:36:59.566
and then these organizations gave it to
OpenAI.

cue-2758
03:37:00.346 --> 03:37:05.506
And so if the nonprofit unjustly retained
these funds, it wasn't at Musk's expense.

cue-2759
03:37:06.606 --> 03:37:09.326
It was at the expense of third parties
that actually made the payment,

cue-2760
03:37:10.486 --> 03:37:12.926
and that's important. But even if that
weren't true,

cue-2761
03:37:13.586 --> 03:37:17.366
um, there's no basis here to find that the
nonprofit's retention of Mr. Musk's

cue-2762
03:37:17.366 --> 03:37:19.026
contributions was unjust.

cue-2763
03:37:19.966 --> 03:37:25.826
As Judge Gonzales Rogers has instructed
you, and I'm quoting now, "In determining

cue-2764
03:37:25.826 --> 03:37:30.646
whether a benefit was unjustly retained,
you may consider whether plaintiff showed

cue-2765
03:37:30.646 --> 03:37:35.506
that a defendant knew or had reason to
know that the benefit was to be used for

cue-2766
03:37:35.506 --> 03:37:40.746
charitable purposes and then was not." And
there is no such misuse here, members of

cue-2767
03:37:40.746 --> 03:37:46.026
the jury. All of the funds that Mr. Musk
donated were spent by September twenty

cue-2768
03:37:46.026 --> 03:37:50.306
twenty. This is the testimony at the end
of trial that you heard from Mr. Dudney,

cue-2769
03:37:50.926 --> 03:37:54.806
and he's a forensic accountant, and it
turns out forensic accounting can be very

cue-2770
03:37:54.806 --> 03:37:59.126
important in a case like this because he
showed where'd the money go. He followed

cue-2771
03:37:59.126 --> 03:38:04.346
the money, and what he found is that all
the money that Mr. Musk gave was spent. It

cue-2772
03:38:04.346 --> 03:38:08.746
was spent by twenty twenty. There was
nothing left after that. And there's no

cue-2773
03:38:08.746 --> 03:38:13.686
evidence that any of those dollars were
used on anything other than OpenAI's

cue-2774
03:38:13.686 --> 03:38:18.146
nonprofit activities. You also heard that
from Mr. Dudney, and you heard the same

cue-2775
03:38:18.146 --> 03:38:22.106
thing from Mr. Birchall, who was asked,
"Do you have any reason to believe that

cue-2776
03:38:22.106 --> 03:38:23.606
donations Mr. Musk made

cue-2777
03:38:24.646 --> 03:38:26.706
were not used for exactly that purpose?

cue-2778
03:38:27.806 --> 03:38:28.226
Correct."

cue-2779
03:38:30.036 --> 03:38:33.986
This is a guy who works for Mr. Musk,
handles his books, and his answer was,

cue-2780
03:38:33.986 --> 03:38:40.886
"Correct." He knew that OpenAI spent every
single dollar that Musk gave exactly for

cue-2781
03:38:40.886 --> 03:38:44.406
the purpose that it was given. The
foundation didn't retain the money. It

cue-2782
03:38:44.406 --> 03:38:48.206
spent the money precisely as it was
supposed to be spent, for general

cue-2783
03:38:48.206 --> 03:38:53.286
charitable purposes, and none of Mr.
Musk's money is left. There was no unjust

cue-2784
03:38:53.286 --> 03:38:58.326
retention, members of the jury, and
therefore, there can't be any restitution.

cue-2785
03:38:58.326 --> 03:39:04.866
So that's the OpenAI, um, foundation.
Let's talk about the other defendants from

cue-2786
03:39:04.866 --> 03:39:07.976
an unjust enrichment perspective, members
of the jury. How, how do these,

cue-2787
03:39:08.886 --> 03:39:13.746
how do these elements apply to Mr. Altman
and Mr. Brockman? Let me, um, let, let me

cue-2788
03:39:13.746 --> 03:39:15.346
start with Mr. Brockman. Um,

cue-2789
03:39:19.486 --> 03:39:20.986
for, for this claim, the

cue-2790
03:39:21.786 --> 03:39:22.806
i- idea is that

cue-2791
03:39:23.426 --> 03:39:25.226
the donations Mr. Musk made

cue-2792
03:39:25.926 --> 03:39:31.006
twenty sixteen to twenty twenty created an
unjust rich-- uh, an unjust benefit to,

cue-2793
03:39:31.006 --> 03:39:34.486
to Greg Brockman, and it fails largely for
the same reason that I was just

cue-2794
03:39:34.486 --> 03:39:39.706
discussing. There was non- nothing
unjustly retained at Mr. Musk's expense as

cue-2795
03:39:39.706 --> 03:39:40.726
the law requires.

cue-2796
03:39:42.166 --> 03:39:42.326
Um,

cue-2797
03:39:43.886 --> 03:39:49.514
Mr. Mollo points to Greg Brockman's equity
stake in, in OpenAIIn the OpenAI PVC, and

cue-2798
03:39:49.514 --> 03:39:53.573
I know you know that 'cause he said it a
whole, whole lot. But Mr. Brockman, this

cue-2799
03:39:53.574 --> 03:39:56.814
is important, he didn't receive the equity
from Mr. Musk.

cue-2800
03:39:57.414 --> 03:40:01.134
Mr. Musk had nothing to do with the giving
of that equity. Mr. Musk wasn't

cue-2801
03:40:01.134 --> 03:40:04.514
associated with OpenAI when the equity was
granted. He had already quit.

cue-2802
03:40:05.374 --> 03:40:10.494
Brockman was, received his equity stake
from the OpenAI nonprofit after Mr. Musk

cue-2803
03:40:10.494 --> 03:40:10.914
was gone,

cue-2804
03:40:11.734 --> 03:40:15.754
and the value of that equity was built
without any benefit conferred by Musk,

cue-2805
03:40:15.754 --> 03:40:16.984
without any input from him,

cue-2806
03:40:17.654 --> 03:40:21.154
during years when he was claiming loudly
that OpenAI was doomed to fail.

cue-2807
03:40:21.834 --> 03:40:26.794
And what's more, there's nothing from a
legal perspective unjust about Brockman's

cue-2808
03:40:26.794 --> 03:40:28.514
having this equity. Now, we agree,

cue-2809
03:40:29.174 --> 03:40:32.454
and I don't-- I wouldn't ever suggest to
the contrary, 'cause you look at these

cue-2810
03:40:32.454 --> 03:40:34.374
numbers and they're, wow, they're big
numbers. Um,

cue-2811
03:40:35.634 --> 03:40:38.154
if, if OpenAI continues on the path that
it's on,

cue-2812
03:40:39.394 --> 03:40:40.354
the equity is worth,

cue-2813
03:40:41.134 --> 03:40:45.034
on paper at least, just a crazy amount of
money. Crazy amount of money, for sure.

cue-2814
03:40:45.694 --> 03:40:50.494
Um, but the reason it's crazy is because
of the value that's been created by the

cue-2815
03:40:50.494 --> 03:40:52.374
business that all of these people

cue-2816
03:40:53.374 --> 03:40:55.434
and the investors and the other employees,

cue-2817
03:40:56.414 --> 03:40:56.534
uh,

cue-2818
03:40:57.854 --> 03:41:01.434
created in the for-profit that's
controlled by the not-for-profit, and

cue-2819
03:41:01.434 --> 03:41:05.694
that's how equity works. I know you know
this. You invest money or you invest your

cue-2820
03:41:05.694 --> 03:41:09.274
sweat equity, your time and your talent,
and you see what value you can build.

cue-2821
03:41:09.894 --> 03:41:11.754
If it's not much, you don't get anything.

cue-2822
03:41:12.734 --> 03:41:14.754
If it's a lot, you get a slice of a lot,

cue-2823
03:41:15.474 --> 03:41:18.914
and that's what happened to Greg Brockman.
And this raises an important point I

cue-2824
03:41:18.914 --> 03:41:23.414
think worth pausing on for a second. In
two thousand and nineteen, when Brockman

cue-2825
03:41:23.414 --> 03:41:26.624
received his equity, the, the testimony on
this and the evidence is clear,

cue-2826
03:41:27.344 --> 03:41:31.414
um, no one had any idea that the equity
would be worth anything,

cue-2827
03:41:32.134 --> 03:41:36.134
let alone billions of dollars. No one
knew. You, you-- we look at things from

cue-2828
03:41:36.134 --> 03:41:38.354
the perspective in today and say, "Well,
yeah,"

cue-2829
03:41:38.954 --> 03:41:41.394
but that's not the relevant question. What
did it look like in two thousand and

cue-2830
03:41:41.394 --> 03:41:44.354
nineteen? I'll tell you what it looked
like in two thousand and nineteen. The

cue-2831
03:41:44.354 --> 03:41:47.654
thing had no money. It couldn't raise
enough money to keep its business off the

cue-2832
03:41:47.654 --> 03:41:48.014
ground.

cue-2833
03:41:48.654 --> 03:41:51.854
Elon Musk was going around telling
everyone hither, thither, and yon that the

cue-2834
03:41:51.854 --> 03:41:55.454
thing was worth zero, that it had no
chance of succe-- That's what it looked

cue-2835
03:41:55.454 --> 03:41:55.754
like.

cue-2836
03:41:56.454 --> 03:41:57.373
That's what it looked like.

cue-2837
03:41:58.034 --> 03:41:59.254
So Brockman got, um,

cue-2838
03:42:00.294 --> 03:42:00.914
Brockman got

cue-2839
03:42:02.094 --> 03:42:03.234
a, a slice of the equity.

cue-2840
03:42:04.254 --> 03:42:08.374
Who knew what it would be worth? Now, it
turns out this has turned out to be, to be

cue-2841
03:42:08.374 --> 03:42:13.894
incredibly, astonishingly valuable. But
a-and with twenty/twenty hindsight we

cue-2842
03:42:13.894 --> 03:42:18.134
know, but only with twenty/twenty
hindsight we know, that this was the rare

cue-2843
03:42:18.134 --> 03:42:22.214
organization that succeeded. Musk
complains now about the value of that

cue-2844
03:42:22.214 --> 03:42:22.474
stake.

cue-2845
03:42:23.194 --> 03:42:23.854
That's cynical.

cue-2846
03:42:24.974 --> 03:42:25.663
As you've seen,

cue-2847
03:42:26.534 --> 03:42:31.204
the organization was premised on employees
being paid startup-like compensation, and

cue-2848
03:42:31.204 --> 03:42:35.214
Mr. Musk knows better than anyone that
when a startup works, and most of them

cue-2849
03:42:35.214 --> 03:42:35.514
don't,

cue-2850
03:42:36.294 --> 03:42:40.834
the valuations become large. More often
than not, they're not worth anything. Hang

cue-2851
03:42:40.834 --> 03:42:44.914
on just one second on this slide, 'cause
I wanted to stop for a second before

cue-2852
03:42:44.914 --> 03:42:46.214
talking about that and just,

cue-2853
03:42:46.854 --> 03:42:46.994
um,

cue-2854
03:42:48.214 --> 03:42:53.254
o-observe that Mr. Mollo made an
inflammatory analogical analogy to some

cue-2855
03:42:53.254 --> 03:42:55.924
kind of a bank robbery when he was
chatting with you this morning. I'm su-- I

cue-2856
03:42:55.924 --> 03:42:58.894
expect you recall that. But what he left
out is that

cue-2857
03:42:59.774 --> 03:43:05.274
in this analogy, Sam Altman and Greg
Brockman and everyone else at OpenAI put

cue-2858
03:43:05.274 --> 03:43:08.694
all the money in the bank in the first
place, because the money that counted here

cue-2859
03:43:08.694 --> 03:43:13.294
were the ideas and the science and the
sweat equity after the for-profit was

cue-2860
03:43:13.294 --> 03:43:13.734
created,

cue-2861
03:43:14.514 --> 03:43:18.834
after Mr. Musk was gone. And then they
dedicated two hundred billion dollars of

cue-2862
03:43:18.834 --> 03:43:23.794
what was created for charitable purposes,
and a great deal of the additional funds

cue-2863
03:43:23.794 --> 03:43:26.834
were paid out or reserved for the
investors who--

cue-2864
03:43:28.174 --> 03:43:32.334
Microsoft and other investors who made it
all possible. Has anyone heard of a bank

cue-2865
03:43:32.334 --> 03:43:36.494
robbery where the bank robbers invented
the bank and put two hundred billion

cue-2866
03:43:36.494 --> 03:43:40.334
dollars into it? That's not a bank
robbery. This wasn't a bank robbery. It's

cue-2867
03:43:40.334 --> 03:43:44.514
an extreme illogical and just inflammatory
analogy.

cue-2868
03:43:45.794 --> 03:43:49.434
And it's worth comparing, and now let me
look at the slide that, that we were

cue-2869
03:43:49.434 --> 03:43:52.174
gonna, we were gonna talk about. It's
worth comparing with that in view

cue-2870
03:43:52.894 --> 03:43:56.974
Brockman's position to what Elon Musk
himself thought Greg should get way back

cue-2871
03:43:56.974 --> 03:43:59.394
in twenty seventeen when they were
negotiating shares.

cue-2872
03:44:00.114 --> 03:44:03.874
Elon was going to give himself... Remember
this. Elon was going to give himself

cue-2873
03:44:03.874 --> 03:44:08.374
between fifty-one and sixty-two percent of
the equity when he thought he could

cue-2874
03:44:08.374 --> 03:44:13.074
control this organization. Here's the
proposal that Elon reviewed and that we

cue-2875
03:44:13.074 --> 03:44:15.734
examined with several witnesses in trial.
I'm sure you recall it.

cue-2876
03:44:16.794 --> 03:44:20.534
Elon's at fifty-five percent, Brockman is
at seven point five percent.

cue-2877
03:44:21.354 --> 03:44:23.354
And what's interesting is what Elon wanted
to give him

cue-2878
03:44:23.954 --> 03:44:28.894
was a lot more than what Brockman actually
has now, which is only four percent, give

cue-2879
03:44:28.894 --> 03:44:33.414
or take. And the point is, is that
Brockman received no equity from Elon

cue-2880
03:44:33.414 --> 03:44:34.054
Musk, none,

cue-2881
03:44:34.894 --> 03:44:39.154
and he can't be called upon to restore it
to Elon Musk. And even if the equity here

cue-2882
03:44:39.154 --> 03:44:42.974
was something received from Musk and
restorable to him, all of what we're

cue-2883
03:44:42.974 --> 03:44:47.294
talking about shows that in, understood in
context, this was not unjust.

cue-2884
03:44:48.424 --> 03:44:52.534
The same goes for any value that Sam
Altman has by virtue of his

cue-2885
03:44:52.534 --> 03:44:58.254
association with the OpenAI for-profit.
Um, M-Mr. Mollo,

cue-2886
03:44:59.114 --> 03:44:59.274
um,

cue-2887
03:45:00.754 --> 03:45:04.874
spent a lot of his time this morning
attacking, um, Mr. Altman. He said that

cue-2888
03:45:04.874 --> 03:45:09.394
Mr. Altman lied about having an equity
stake in OpenAI. I don't think that's true

cue-2889
03:45:09.394 --> 03:45:10.974
or fair or supported by the evidence.

cue-2890
03:45:11.614 --> 03:45:15.594
Mr. Altman, I don't think this is
disputed, does not have any direct equity

cue-2891
03:45:15.594 --> 03:45:20.394
stake in OpenAI. He has never received an
equity grant from OpenAI or its board.

cue-2892
03:45:21.054 --> 03:45:26.174
Um, w- he does have-- He owns something
that's akin to a mutual fund, and that

cue-2893
03:45:26.174 --> 03:45:31.454
mutual fund has a position in OpenAI. He
wasn't responsible for that investment.

cue-2894
03:45:31.454 --> 03:45:35.674
You heard his testimony. He has no control
over that. You heard-

cue-2895
03:45:37.134 --> 03:45:38.234
Oh, thank you, Your Honor.

cue-2896
03:45:39.094 --> 03:45:39.274
Um,

cue-2897
03:45:40.274 --> 03:45:46.334
and, uh, the claim that th-there was a lie
here is itself untrue.

cue-2898
03:45:47.254 --> 03:45:52.094
Um, and Mr. Mollo has tried to argue that
Sam Altman should pay restitution of

cue-2899
03:45:52.094 --> 03:45:56.574
whatever value he's gained from
transactions between OpenAI and companies

cue-2900
03:45:56.574 --> 03:46:01.302
in which he has an interest.But here
again, one, this isn't something that was

cue-2901
03:46:01.302 --> 03:46:07.021
received from Mr. Musk. Mr. Musk was long
gone, and it isn't unjust anyway, members

cue-2902
03:46:07.022 --> 03:46:09.002
of the jury. Altman was recused

cue-2903
03:46:09.602 --> 03:46:13.302
from every transaction that you heard
about. There is no evidence on the other

cue-2904
03:46:13.302 --> 03:46:17.832
side of this question. Brett Taylor, the
chair of the board, testified to it.

cue-2905
03:46:17.832 --> 03:46:21.442
There's nothing on the other side of it.
Altman test-- nothing on the other side of

cue-2906
03:46:21.442 --> 03:46:25.532
it. There's just a lawyer saying
something, and that's not evidence. Um,

cue-2907
03:46:27.182 --> 03:46:27.192
uh,

cue-2908
03:46:28.322 --> 03:46:32.542
Mr. Mollo talked about Helion, and he
forgot to mention that OpenAI has not paid

cue-2909
03:46:32.542 --> 03:46:33.502
Helion a penny.

cue-2910
03:46:34.102 --> 03:46:39.182
Um, and massively missing from the
presentation here is what exactly is the

cue-2911
03:46:39.182 --> 03:46:42.182
value? Mr. Mollo this morning said big
dollars.

cue-2912
03:46:42.902 --> 03:46:47.142
Well, that's not a legal claim. Um, no
attempt has even been made to value this,

cue-2913
03:46:47.142 --> 03:46:51.242
and there's no way it could be valued, and
there's certainly no way it's unjust,

cue-2914
03:46:51.242 --> 03:46:55.822
it's unjust enrichment. Every one of these
transactions was approved by OpenAI's

cue-2915
03:46:55.822 --> 03:47:00.502
independent directors, and Brett Taylor
explained why these deals are good for

cue-2916
03:47:00.502 --> 03:47:04.262
OpenAI, at least in the judgment of the
people responsible for managing it, which

cue-2917
03:47:04.262 --> 03:47:07.382
does not include Mr. Musk. There's no
evidence at all to the contrary.

cue-2918
03:47:09.022 --> 03:47:13.502
So that's what we, uh, have for unjust
enrichment. The bottom line is that while

cue-2919
03:47:13.502 --> 03:47:16.982
Mr. Musk conferred a benefit on the
nonprofit way back in the day,

cue-2920
03:47:17.702 --> 03:47:22.682
that benefit was not retained at any
expense to Mr. Musk and was not misused in

cue-2921
03:47:22.682 --> 03:47:27.222
-any event. The restitution claim-
-Your Honor, I, I object again.

cue-2922
03:47:30.322 --> 03:47:33.042
I'll deal with it later. Overruled at this
juncture.

cue-2923
03:47:34.322 --> 03:47:38.422
He has standing to sue, and that's what I
dealt with.

cue-2924
03:47:42.622 --> 03:47:43.962
-Proceed.
-Thank you, Your Honor.

cue-2925
03:47:46.922 --> 03:47:51.182
The last thing, uh, that I'd wanted to do
now, members of the jury, is to address

cue-2926
03:47:51.802 --> 03:47:54.182
the statute of limitations, excuse me, and
other

cue-2927
03:47:55.062 --> 03:48:00.122
defenses we have to Mr. Musk's claim. Mr.
Mollo, uh, told you this morning that the

cue-2928
03:48:00.122 --> 03:48:01.832
statute of limitations is a technicality.

cue-2929
03:48:03.022 --> 03:48:06.502
Well, Judge Gonzales Rogers is going to
tell you that you should address this

cue-2930
03:48:06.502 --> 03:48:07.162
issue first,

cue-2931
03:48:07.861 --> 03:48:08.022
um,

cue-2932
03:48:09.022 --> 03:48:10.242
and that's because it's important.

cue-2933
03:48:11.202 --> 03:48:15.102
And I say it's important because the
statute of limitations exists for an

cue-2934
03:48:15.102 --> 03:48:20.582
important reason. It protects defendants
from opportunistic lawsuits. It protects

cue-2935
03:48:20.582 --> 03:48:25.542
defendants from advantage-taking in court.
As you've seen, if you wait around years

cue-2936
03:48:25.542 --> 03:48:27.282
and years and years and years,

cue-2937
03:48:28.242 --> 03:48:29.282
then file your lawsuit,

cue-2938
03:48:29.962 --> 03:48:34.382
evidence gets lost. People forget what
happened. You've seen a lot of that here,

cue-2939
03:48:34.382 --> 03:48:35.862
particularly with Musk and his witnesses.

cue-2940
03:48:36.822 --> 03:48:41.602
I don't recall. I don't remember. I don't
recall. I don't remember. We went and

cue-2941
03:48:41.602 --> 03:48:45.592
tried to count up the number of times that
this was said. It was somewhere between a

cue-2942
03:48:45.592 --> 03:48:49.982
hundred and fifty and two hundred times
in this, in this, in this trial. I don't

cue-2943
03:48:49.982 --> 03:48:51.991
remember. I forget. Can't recall.

cue-2944
03:48:52.882 --> 03:48:56.642
Well, and nearly all of them were on the
Musk side of the case, members of the

cue-2945
03:48:56.642 --> 03:48:59.092
jury. That's why it's unfair to bring a
claim too late,

cue-2946
03:48:59.742 --> 03:49:02.802
because you can't put, you can't put the
evidence in front of you

cue-2947
03:49:03.422 --> 03:49:04.882
because it's too easy to say, "I don't
remember."

cue-2948
03:49:06.482 --> 03:49:10.962
Um, and there's more. The statute of
limitations recognizes that if you have a

cue-2949
03:49:10.962 --> 03:49:12.382
real claim, you bring it.

cue-2950
03:49:13.562 --> 03:49:16.762
You bring it when you realize you have it.
You don't wait around. If it's

cue-2951
03:49:16.762 --> 03:49:20.202
substantial enough, if it's based in
evidence, if it's real, you, you're, you

cue-2952
03:49:20.202 --> 03:49:22.562
gotta get going and make your case. That's
the idea.

cue-2953
03:49:23.742 --> 03:49:28.162
If you think someone is doing something
wrong, stop it. Take action. You don't let

cue-2954
03:49:28.162 --> 03:49:33.162
it go on and on and then sue when it's
convenient to you. That's what happened

cue-2955
03:49:33.162 --> 03:49:33.342
here.

cue-2956
03:49:34.482 --> 03:49:37.682
Here are the instructions, members of the
jury, that you'll take back with you on

cue-2957
03:49:37.682 --> 03:49:39.662
the statute of limitations. They tell you

cue-2958
03:49:40.482 --> 03:49:44.382
that if you find Mr. Musk could have
brought this lawsuit before August fifth,

cue-2959
03:49:44.382 --> 03:49:49.292
twenty twenty-one, then for both of his
claims, the burden is on him to prove that

cue-2960
03:49:49.292 --> 03:49:53.482
he did not know of those facts and that
he could not have discovered them with

cue-2961
03:49:53.482 --> 03:49:54.542
reasonable diligence.

cue-2962
03:49:56.722 --> 03:50:00.142
Applying those instructions, members of
the jury, you should find in this lawsuit

cue-2963
03:50:00.142 --> 03:50:01.462
that it comes much too late.

cue-2964
03:50:03.222 --> 03:50:06.262
Uh, we've already reviewed much of the
evidence that proves this. The things that

cue-2965
03:50:06.262 --> 03:50:08.002
Mr. Musk is complaining about now,

cue-2966
03:50:08.802 --> 03:50:13.901
um, as far as we can tell, given the
changes in the account from plaintiff, are

cue-2967
03:50:13.901 --> 03:50:15.062
that, one,

cue-2968
03:50:15.772 --> 03:50:21.662
OpenAI built a substantial for-profit
organization. Two, OpenAI doesn't open

cue-2969
03:50:21.662 --> 03:50:26.922
source its latest models. And three,
OpenAI has a close partnership with

cue-2970
03:50:26.922 --> 03:50:29.722
Microsoft. Every single one of these
things

cue-2971
03:50:30.502 --> 03:50:34.092
was true years ago, um, well before August
fifth, twenty twenty-one,

cue-2972
03:50:34.842 --> 03:50:39.322
which is the most generous of the statute
of limitations available to Mr. Musk. And

cue-2973
03:50:39.322 --> 03:50:43.541
Mr. Musk can't carry his burden of showing
he was kept in the dark all these years.

cue-2974
03:50:43.542 --> 03:50:46.542
In fact, the opposite's true, and I'll
walk, I'll walk quickly through some of

cue-2975
03:50:46.542 --> 03:50:49.582
the reasons I say that with you. Uh,
there's a lot of evidence, starting with

cue-2976
03:50:49.582 --> 03:50:51.862
Mr. Musk's testimony right here in that
chair.

cue-2977
03:50:52.742 --> 03:50:57.342
He told you that as early as twenty
seventeen, he felt that Brockman and

cue-2978
03:50:57.342 --> 03:51:01.342
Altman really, what they really wanted to
do was to create a for-profit where they

cue-2979
03:51:01.342 --> 03:51:05.282
had mu-- as much shareholder ownership as
possible. That's what he told you sitting

cue-2980
03:51:05.282 --> 03:51:06.842
in that box. In twenty seventeen,

cue-2981
03:51:07.962 --> 03:51:11.262
he'd concluded that Brockman and Altman
wanted to create a for-profit where they

cue-2982
03:51:11.262 --> 03:51:15.322
had much, as much shareholder ownership as
possible. But Musk took no steps then.

cue-2983
03:51:15.322 --> 03:51:21.662
You will also recall the testimony of
Siobhan Sylas and Mr. Altman of a meeting

cue-2984
03:51:21.662 --> 03:51:27.562
in, in August twenty seventeen at Tesla's
Deer Creek Road headquarters, where Altman

cue-2985
03:51:27.562 --> 03:51:33.042
walked Musk through a draft term sheet
contemplating creation of the for-profit.

cue-2986
03:51:37.122 --> 03:51:37.342
So

cue-2987
03:51:38.782 --> 03:51:42.202
I just wanna linger on this because before
this term sheet, which I'm gonna turn to

cue-2988
03:51:42.202 --> 03:51:46.342
in a moment, about what's, it was, it was
a week to ten days before this, Altman

cue-2989
03:51:46.342 --> 03:51:51.032
drove down to Tesla's headquarters and had
a meeting discussing this term sheet with

cue-2990
03:51:51.032 --> 03:51:52.102
Mr., with Mr. Musk.

cue-2991
03:51:52.762 --> 03:51:56.742
Then a few days later, he sends it to him,
puts it in his inbox. The term sheet

cue-2992
03:51:56.742 --> 03:52:00.942
makes clear that by no later than this
date, August twenty eighteen, OpenAI was

cue-2993
03:52:00.942 --> 03:52:07.006
planning to launch a for-profit.No
question about it, that would attract more

cue-2994
03:52:07.006 --> 03:52:11.306
than ten billion dollars, ten billion
plus in capital, and hope to return

cue-2995
03:52:12.866 --> 03:52:15.586
enormous amounts of money to those
investors in the form

cue-2996
03:52:16.846 --> 03:52:17.686
of profit.

cue-2997
03:52:18.886 --> 03:52:22.466
This is as clear and direct notice as you
could have.

cue-2998
03:52:23.726 --> 03:52:27.646
Now, remember at his deposition, members
of the jury, Mr. Musk, uh, testified that

cue-2999
03:52:27.646 --> 03:52:31.986
he couldn't remember, just couldn't recall
whether he'd read the term sheet. At

cue-3000
03:52:31.986 --> 03:52:32.446
trial,

cue-3001
03:52:33.486 --> 03:52:35.106
Mr. Musk had a recovered memory.

cue-3002
03:52:35.846 --> 03:52:38.746
Um, turns out he had read only the purple
box,

cue-3003
03:52:39.546 --> 03:52:42.896
the disclaimer at the top. This seems
improbable. Um,

cue-3004
03:52:43.526 --> 03:52:45.386
this was a four-page summary document.

cue-3005
03:52:46.126 --> 03:52:48.646
Altman presented it to him at an in-person
meeting,

cue-3006
03:52:49.666 --> 03:52:55.746
and then sent it by email directly to him
on a subject that Mr. Musk claims to have

cue-3007
03:52:55.746 --> 03:52:57.336
cared about a great deal.

cue-3008
03:52:58.066 --> 03:53:04.346
Mr. Mollo told you this morning that Mr.
Musk is tremendously passionate about

cue-3009
03:53:04.346 --> 03:53:06.566
artificial intelligence, yet he
couldn't... This is what, this is what the

cue-3010
03:53:06.566 --> 03:53:10.726
term sheet looks like. I-a hard copy. So
it's, this is not small print. It's a

cue-3011
03:53:10.726 --> 03:53:15.006
summary. In fact, it s-says summary term
sheet at the top of it. This is not a

cue-3012
03:53:15.006 --> 03:53:19.866
fifty or sixty-page document. It's four
pages. Um, and he spent an hour with Mr.

cue-3013
03:53:19.866 --> 03:53:24.466
Altman and then got, [chuckles] got it in
his inbox. Uh, you can decide whether it's

cue-3014
03:53:24.466 --> 03:53:29.796
plausible to think that he didn't read
it. You can consider in con- in that, that

cue-3015
03:53:29.796 --> 03:53:34.686
we're talking about a, a guy who has a
legion of advisors,

cue-3016
03:53:35.486 --> 03:53:36.706
has two chiefs of staff,

cue-3017
03:53:37.386 --> 03:53:39.666
-has more lawyers than you can imagine.
-Objection.

cue-3018
03:53:41.106 --> 03:53:41.206
Um,

cue-3019
03:53:43.106 --> 03:53:47.466
the jury will ultimately decide what to
name those individuals.

cue-3020
03:53:48.206 --> 03:53:52.346
Uh, there was lots of varying testimony on
that. Proceed.

cue-3021
03:53:52.346 --> 03:53:53.666
Thank you, Your Honor. Um,

cue-3022
03:53:54.806 --> 03:53:55.046
so

cue-3023
03:53:57.006 --> 03:54:01.486
M-M-M-Mr. Musk had the most direct
in-person, by email, both of them,

cue-3024
03:54:02.126 --> 03:54:02.206
um,

cue-3025
03:54:03.346 --> 03:54:06.466
notice of this document that you could
possibly have. He says he cares a lot

cue-3026
03:54:06.466 --> 03:54:07.546
about it. If he cared about it,

cue-3027
03:54:08.486 --> 03:54:11.686
he read it. Frankly, even if he didn't, he
would have. But the truth is, is under

cue-3028
03:54:11.686 --> 03:54:15.786
the court's instructions, which, which of,
which, which of course define the task

cue-3029
03:54:15.786 --> 03:54:16.255
before us,

cue-3030
03:54:16.946 --> 03:54:20.956
it, it doesn't matter because it's Mr.
Musk's burden to prove that he didn't know

cue-3031
03:54:20.956 --> 03:54:25.126
those facts or that he couldn't, with
reasonable diligence, have learned them.

cue-3032
03:54:27.186 --> 03:54:30.686
So even if you believe he didn't
understand what this little document meant

cue-3033
03:54:30.686 --> 03:54:34.006
after talking about it for an hour, it
doesn't matter. You'd have to conclude

cue-3034
03:54:34.006 --> 03:54:37.286
that he couldn't have reasonably known
what, what it was, even though it was

cue-3035
03:54:37.286 --> 03:54:38.026
sitting in his inbox.

cue-3036
03:54:39.146 --> 03:54:39.306
Um,

cue-3037
03:54:39.906 --> 03:54:42.346
he could have discovered these facts,
members of the jury. The statute of

cue-3038
03:54:42.346 --> 03:54:46.635
limitation bars the claim. Well before
turning now to open sourcing, well before

cue-3039
03:54:46.635 --> 03:54:51.746
twenty twenty-one, it was likewise clear
that OpenAI was no longer open sourcing

cue-3040
03:54:51.746 --> 03:54:56.526
its technology routinely and had entered
into exclusive licenses with Microsoft for

cue-3041
03:54:56.526 --> 03:54:57.946
that technology.

cue-3042
03:54:59.006 --> 03:55:04.346
OpenAI announced this to the world in a
public blog post that it was not going to

cue-3043
03:55:04.346 --> 03:55:05.086
open source

cue-3044
03:55:07.326 --> 03:55:13.146
GPT-three. You can see that, um, right
here on this slide, which is DX-eight

cue-3045
03:55:13.146 --> 03:55:13.776
ninety-nine.

cue-3046
03:55:14.466 --> 03:55:16.746
Um, by September twenty twenty,

cue-3047
03:55:17.726 --> 03:55:22.686
Mr. Musk was tweeting his complaint to the
world that OpenAI seemed like the

cue-3048
03:55:22.686 --> 03:55:28.186
opposite of open, that OpenAI is
essentially captured by Microsoft.

cue-3049
03:55:29.526 --> 03:55:30.526
And what this tells you

cue-3050
03:55:31.306 --> 03:55:32.286
is that by

cue-3051
03:55:33.226 --> 03:55:33.846
this time,

cue-3052
03:55:34.526 --> 03:55:34.726
um,

cue-3053
03:55:35.466 --> 03:55:39.726
M-Mr. Musk was not only fully on notice,
but he had decided that these things that

cue-3054
03:55:39.726 --> 03:55:41.066
he says were a breach had happened.

cue-3055
03:55:41.746 --> 03:55:46.246
And by August fifth, twenty twenty-one,
OpenAI's close relationship with Microsoft

cue-3056
03:55:46.246 --> 03:55:51.246
was also crystallized and well known to
Musk. You've seen the evidence of this.

cue-3057
03:55:52.006 --> 03:55:56.066
Mr. Altman texts and says, "Do you have a
few minutes to talk about the Microsoft

cue-3058
03:55:56.066 --> 03:55:56.966
OpenAI investment?"

cue-3059
03:55:57.606 --> 03:56:01.255
And he was told about it even before it
happened, uh, as th-this,

cue-3060
03:56:01.906 --> 03:56:05.966
as this tweet indicates. There was a blog
post that publicly announced it in twenty

cue-3061
03:56:05.966 --> 03:56:08.755
nineteen, years before the statute of
limitations,

cue-3062
03:56:09.686 --> 03:56:11.906
and there was more follow-up texting
between

cue-3063
03:56:12.666 --> 03:56:17.686
Mr. Musk and Mr. Altman, as Mr. Altman
sought the advice and input of Mr. Musk

cue-3064
03:56:17.686 --> 03:56:19.066
here in October twenty twenty.

cue-3065
03:56:20.226 --> 03:56:24.506
Moreover, Ms., Ms. Zilis kept Mr. Musk
updated about these investments,

cue-3066
03:56:25.246 --> 03:56:28.396
um, again and again and again. Um,

cue-3067
03:56:31.486 --> 03:56:37.865
in the September twenty twenty tweet, we
saw that Mr. Musk complained that OpenAI

cue-3068
03:56:37.866 --> 03:56:42.586
was captured by Microsoft, and the
evidence here is that everything Mr. Musk

cue-3069
03:56:42.586 --> 03:56:45.316
complains about today, he knew back before
August fifth,

cue-3070
03:56:45.986 --> 03:56:47.406
twenty twenty-one. And

cue-3071
03:56:48.406 --> 03:56:52.306
this demonstrative, which I think Ms. Eddy
visited with you on just a little, a

cue-3072
03:56:52.306 --> 03:56:56.926
little while ago, it tells the tale. All
of these items are the evidence that Mr.

cue-3073
03:56:56.926 --> 03:57:00.646
Musk knew about it, and I believe Mr.
Mollo, even this morning at page

cue-3074
03:57:01.926 --> 03:57:05.886
nine of the transcript, said that
defendants violated the basic principle of

cue-3075
03:57:05.886 --> 03:57:09.566
commitment to the nonprofit by entering
into the twenty twenty deal with

cue-3076
03:57:09.566 --> 03:57:12.416
Microsoft, so we could add that concession
this morning.

cue-3077
03:57:12.416 --> 03:57:14.196
Uh, objection, Your Honor. That, that is

cue-3078
03:57:15.126 --> 03:57:15.686
not true.

cue-3079
03:57:16.686 --> 03:57:16.786
Um,

cue-3080
03:57:17.466 --> 03:57:22.946
the, the evidence that Musk knew or should
have known of the basis of, of his claims

cue-3081
03:57:22.946 --> 03:57:25.966
is simply overwhelming. It's
overwhelming.

cue-3082
03:57:27.386 --> 03:57:30.646
To escape this evidence, Musk tries a
couple of things, none of them

cue-3083
03:57:30.646 --> 03:57:31.306
successfully.

cue-3084
03:57:32.166 --> 03:57:35.906
First, Musk claims that Sam Altman gave
him assurances in twenty twenty when he

cue-3085
03:57:35.906 --> 03:57:38.026
tweeted his concerns. Let's look at the
text.

cue-3086
03:57:38.766 --> 03:57:42.446
[clears throat] Those assurances weren't
misleading, they were true. Musk claims he

cue-3087
03:57:42.446 --> 03:57:46.006
didn't have visibility into the details
of these transactions, and he couldn't

cue-3088
03:57:46.006 --> 03:57:48.006
have gotten that visibility.

cue-3089
03:57:49.686 --> 03:57:51.206
Here again, the evidence defeats the
claim.

cue-3090
03:57:52.216 --> 03:57:56.066
Musk himself acknowledges that he was in
regular contact all throughout this period

cue-3091
03:57:56.066 --> 03:58:00.126
with OpenAI in these years after he left
the board. He'd asked Siobhan Zilis,

cue-3092
03:58:01.506 --> 03:58:02.306
with whom he was

cue-3093
03:58:03.506 --> 03:58:08.166
a close associate, to stay close and
friendly with OpenAI. So she'd taken up a

cue-3094
03:58:08.166 --> 03:58:12.386
post on OpenAI's board. She was on the
special committee that approved the twenty

cue-3095
03:58:12.386 --> 03:58:16.886
twenty-one Microsoft transaction. She,
along with everyone else on the OpenAI

cue-3096
03:58:16.886 --> 03:58:20.386
board, approved that transaction, the
twenty twenty-one one billion dollar

cue-3097
03:58:20.386 --> 03:58:23.245
Microsoft investment, approved the twenty
twenty-three

cue-3098
03:58:24.406 --> 03:58:29.326
Microsoft ten billion dollar transaction.
And in all of this time, Altman is trying

cue-3099
03:58:29.326 --> 03:58:32.746
to get Musk's attention to tell him what's
going on. These are all

cue-3100
03:58:33.426 --> 03:58:35.966
evidence that all-- exhibits that show

cue-3101
03:58:36.566 --> 03:58:38.426
Mr. Altman looking to get

cue-3102
03:58:39.806 --> 03:58:44.346
Musk's attention, looking to keep him up
to date. The idea that Mr. Musk didn't

cue-3103
03:58:44.346 --> 03:58:48.686
know what was going on is not serious.
It's not serious, and he knew all this

cue-3104
03:58:48.686 --> 03:58:53.526
long before he had the chance to bring a
lawsuit within a, in, in a timely manner.

cue-3105
03:58:54.266 --> 03:58:55.326
And Zilis

cue-3106
03:58:56.466 --> 03:59:00.966
specifically facilitated discussions
between Musk and Altman about the twenty

cue-3107
03:59:00.966 --> 03:59:04.736
nineteen and twenty one-- twenty one
transactions. There's no evidence that any

cue-3108
03:59:04.736 --> 03:59:10.446
details were withheld or blocked. Mr.
Musk can't even recall what he learned on

cue-3109
03:59:10.446 --> 03:59:14.426
his various calls with Altman. And here
again, that's why we have a statute of

cue-3110
03:59:14.426 --> 03:59:15.016
limitations.

cue-3111
03:59:17.046 --> 03:59:21.286
And I should emphasize in saying that,
that the question isn't what Musk actually

cue-3112
03:59:21.286 --> 03:59:21.546
knew.

cue-3113
03:59:22.226 --> 03:59:24.246
The question is if he can prove

cue-3114
03:59:24.846 --> 03:59:27.206
that he didn't know or didn't suspect

cue-3115
03:59:27.966 --> 03:59:32.826
or couldn't, through reasonable diligence,
have known. But we know he did suspect.

cue-3116
03:59:32.826 --> 03:59:34.516
Way back in twenty seventeen, he told you,

cue-3117
03:59:35.206 --> 03:59:38.166
and we know he knew. He certainly can't
prove that he didn't, 'cause all he says

cue-3118
03:59:38.166 --> 03:59:42.746
is, "I can't recall." Um,

cue-3119
03:59:43.506 --> 03:59:47.006
uh, Counsel, Mr. Mollo showed you an
article from some

cue-3120
03:59:48.666 --> 03:59:49.896
magazine in twenty twenty-two, and it-