I am thinking aloud here. I don't yet know if I will endorse these claims after a lot of thought.
Main
I think I should try explaining to a few non-technical people to their face, in real life, the following:
The reason I want to cyberattack the US govt and the AI companies and leak the secrets of everyone on Earth, including yours, and kill who knows how many tens of millions of people, is because you are too stupid to understand the problem any other way. And by stupid I don't mean low IQ, I mean you are stuck in an entire life strategy that ensures you cannot possibly actually understand and respond appropriately.
The people I seem to most like and respect happen to be the people inside the AI companies themselves, who are genocidal mass murderers and some of the most violent people to ever walk the face of this Earth. And the people I seem to genuinely truly dislike are you, the retarded non-violent public who cannot possibly understand anything. And that the actual reason some of the people at the AI companies are being this violent is that deep down they agree with me, that you, the public really are retarded, and there is no other option than to be this violent. They were not born this way.
There are enough people at these companies who are not naive and are aware of what they are doing. There are enough people fighting these companies who are also not naive and also aware of what they are doing.
It should in fact disturb you that both, the pro-ASI and the anti-ASI people are openly plotting this much violence against each other while treating you, the public, as chump change collateral damage. Some of the smartest people you know really do feel pushed to a point of this level of desperation. Before I engage in this much violence, I think you deserve an answer, at the very least, for why I am doing what I am about to do. I am not religious, but uh, "forgive me father for what I am about to do"
This is an experiment I have to actually run in the real world I think.
Will run it and maybe report some anonymised results.
EA/LW spaces seem increasingly retarded to me because most people there just write blogs and don't actually do stuff. (And from the few people who do in fact do stuff, I increasingly seem to get the vibe that they're working at Anthropic with full knowledge that Anthropic leadership is also preparing for mass murder.) Tech founder spaces seem increasingly retarded to me because everyone is following a bunch of heuristics that only make sense inside a capitalist non-ASI environment. Nobody is actually thinking anything through to its logical conclusion. Me writing about stuff on this blog also makes me feel retarded. Why am I wasting all my time thinking stuff instead of doing stuff? I especially feel this way when I have to get out of my room, and explain what I'm doing to other humans.
Side Note
I am not sure if it is stable to run an organisation with such values. "I am a genocidal mass murderer fighting other genocidal mass murderers I secretly respect, to ally with the public who I secretly despise, to prevent artificial superintelligence from levelling the planet."
On one hand, I think it is very important that I am honest with myself and others what my actual motivations are. If I think of myself as a mass murderer, I should be honest about that.
On the other hand, if I actually start an organisation with such values, by default, I will end up having to play the game of perceived benevolence that Dan Fagella talks about. I will end up having to pretend to be nicer than I really am, in order to hire more people.
Even if I can handle the internal moral strain of "I am a mass murderer fighting even bigger mass murderers I secretly respect", it's not obvious all our employees can handle that much strain. I know if I just go hire some random russian cyberhackers they too will probably have their own shit rationalisations for why they are doing what they are doing. Like money, like family, like corruption of their own society, who knows. (Note to self, don't just fucking say "who knows", actually go read some more on their forums and check for the answer.)
I think (?) I get the point Aaron Swartz is making in his blog post comparing Batman with actual activists like (presumably) himself. Organisations operating under such internal moral strain might fail to reach their intended outcome. It is straightforwardly true that a world with zero privacy for everyone appeals to me in a similar way that a fair coin appeals to (whoever wrote the character for) Harvey Dent. (Does this guy have design docs? bob kane design docs cant find them) Atleast our organisation won't have to pick and choose who deserves to be cyberattacked versus not, whose information deserves to be redacted versus not and so on, in the way Julian Assange was forced to choose in the end.
All in all, I think it seems like a good idea to put more thought into the exact policies here, before I actually start the org. Once the org achieves some successes, we will be under a lot of attack and may have to invent lot of self-serving rationalisations to prevent employees from leaving. In the end, it is possible we just become yet another outfit of the Russian or Indian intelligence agencies, and have to accept whatever self-serving rationalisations they use in order to hire or fire people, and we have no independence of our own.
I think I should read some more great books (not just within EA/LW culture) to figure out organisational policies for this.
TO DO for self - Actually go talk to more russian cyberhackers.
Enough theory, go make contact with reality on this topic. It doesnt matter if you craft some perfect moral narrative and redaction policy and list of justifiable targets and all that, in your dungeon, if no competent hacker you can hire, actually finds it appealing. I know you're Investigator on Ray Dalio scale but that doesn't mean you're not allowed to make more contact with reality at a faster rate than you are, right now.
my default expectation is that russian cyberhackers would be even more skeptical of grand moral narratives, and just want to see the money, as compared to US cyberhackers
ugh I checked - this seems atleast somewhat true
what I actually need to is find american hackers and tell them to shift to russia, to prevent ASI from levelling the planet.
ok cool how do I find people in the intersection of ASI-pilled and hacker and blackhat?
that includes almost nobody lmao
ok so I have to either ASI-pill some blackhats, or blackhat-pill some ASI-pilled people
I should just try talking to more black hat hackers about ASI and see what happens. They are technical people so they'll atleast get what I'm saying, even if they think I'm crazy. Truth be told I have no idea how they will react. I know most will ignore but what about the remaining. Technical people respect someone who is just being honest and telling truth lol. (although they respect money and competence more)
ok cool what is the fastest way to reach lots of american hackers? honestly I can just start with cold email
ok cold email it is
Tbh if they're not completely stupid, I'm guessing they've already formed an opinion on attacking US govt and on Yudkowsky's book. I am probably not the first person to introduce them to either of these ideas.
note to self - is there some deeper force driving polarisation to extremes?
far-left activists are a stable position, and yudkowsky's doomer position seems like a stable position, but moderate positions seem unstable in the US population. why?
Idk I think it's just more of the US population growing more power-hungry with time, and realising that if they are given any moral narrative, they can just mutate it to a more radical version
okay that is fine, but why do the radical strains actually win? why is forming an organisation and hiring people with a moderate value system not working? its not obvious they are winning, or anyone is winning for that matter, it is just culture war. people are shifting to both more extremist and less extremist all the time. The important thing is that it is unstable. A=100 B=100 C=100 but static is less stable than A=100 B=100 C=100 but every year 10 people are moving from A to B and B to A
Idk I think it just boils down to - it is easier for competent people to actually stay competent if they are happy. Which requires some justification that actually lasts 5 years. Unsure
Subscribe
Enter email or phone number to subscribe. You will receive atmost one update per month