personal/unhappiness_since_2023.html

2025-11-24

Unhappiness since 2023

Disclaimer

• Quick Note
• Contains politically sensitive info

Summary

• I have been unhappy since 2023, and expect to be unhappy in 2026 too.
• My best theory for why is as follows: Me and people like me want power to fix AI risk. Peace between powerseeking actors can only be brokered at gunpoint. There is currently no stable ideology or political system that can broker this peace. Therefore our civilisation seems atleast somewhat fucked at a result. I am neither on track to fixing the world, nor having amazing relationships. Hence I am unhappy.

Am I happy?

• Definition: I would consider my mood as Above Average, if I reported my mood as Above Average on a majority of days in a period of two weeks.
• I was Above Average for most of 2021 and 2022.
• I was Below Average for most of 2023, 2024 and 2025.
• I expect to continue to be below average in 2026 as well. I have basically given up on finding solutions to this in the next few months, and this post is to summarise why.

Sources of unhappiness (Summarised)

• I was dealing with some interpersonal conflict in 2023 and early 2024. This is now mostly resolved. To protect privacy of people involved in this conflict, I will not talk any more about this problem on the public internet.
• I think the world might be ending due to superintelligence coming to Earth. I think I need to help fix this problem for the world. Most of my unhappiness today appears to be downstream of this.

Solutions to my unhappiness (Summarised)

• I plan to leave India within next 12 months.
  • I want to spend more in-person time with people who care about AI risk enough that they have restructured the rest of their life to work on it.
  • Online calls I currently do are not sufficient, and in-person time would help.
• I also publish my writings and content online, so that people who are ideologically aligned with me can reach out.
  • It is possible I need to get better at becoming more persuasive, so I can find more allies.
  • I don't currently consider most people working on AI risk to be my allies, as they are not willing to actively engage in conflict against AI companies.
• I understand even doing all of this probably won't fully fix my unhappiness.
  • Deep down I have basically just given up on trying to be happy. At an intellectual level I have accepted that I will continue to be unhappy for all of 2026 at the very least.
• Unless I become clinically depressed I am unlikely to change my path.
  • If I were to become clinically depressed, I can imagine spending multipe years trying to fix my unhappiness.
  • I am currently okay spending a few weeks on experimental solutions to my unhappiness, but I am not okay spending multiple months on them, as this would take time away from my current plans to fix AI risk.

Why does AI risk cause me unhappiness? (Detailed)

0. Culture war stuff not directly related to AI risk

• I am openly in disagreement with lots of people and ideologies around me.
• Many people around me correctly realise that a maximally successful Samuel means the total obliteration of their religion or ideology from planet Earth. They correctly decide that it would be good in their eyes for me to fail at my life plans.
• I typically ignore such people, and don't spend too much time with them. Much as I would like to remove bad ideologies from this world, I think fixing AI risk is a higher priority, so I don't let it distract me too much.
• If I were to publicly get into conflict with any ideology, I would mostly use speech, and social and financial incentives. I would probably not resort to violence.
• I think most of the suffering I see around me on a day-to-day basis is due to bad religions and ideologies. I am upset by the fact that I can't spend my time eradicating them, and have to spend my time fixing AI risk instead.
• I have somewhat gotten over this sadness, and I have buried the rest of it for later. If I want revenge, I understand I need to play the longterm game not the shortterm game.
• I have personally been hurt by people following what I think are bad ideologies, and I desire revenge. I think the best revenge in this case is not go to hurt anyone specifically, but to eliminate the ideology that inspired these people to hurt me. (I can't elaborate more directly about this, on the public internet)

1. Loneliness

• I feel increasingly disconnected from anyone who does not care about AI risk.
  • I still care about them as people, but there is little of value they can offer to me. Their advice is mostly not helpful, and their willingness to listen to me emotionally is also mostly not helpful beyond a point. After being listened to a few times, I ultimately want advice, and nobody has good advice because nobody around me is working on AI risk full-time.
  • I don't want to spend too much time doing fun activities. This also reduces how much time I want to spend with most people.
  • (I currently spend one day a week off work, usually with people, and all nights off work, usually alone.)
  • Because they have little value to offer me, and I also have limits on how much time I want to spend offering them value, the result is we don't spend that much time together.
  • I am able to relate to Evernote CEO Phil Libin's advice. He says you have a limited number of close friend slots in your life, and your biggest bottleneck to achieving great outcomes is finding great people to work with. Therefore don't even bother to make close friendship with someone you can't start a company with.
  • I am currently not willing to go as far as Libin, but I get where he is coming from.
• All my relationships come with terms and conditions attached.
  • I value acquiring power to fix AI risk, more than I value any of my relationships. For example, if I support cyberattacks against AI companies, I will continue to support this even if most of my friends distance from me for doing so. If I support a hunger strike and do this, I will continue to do this even if most of my friends distance from for doing so.
  • In practice, often a lot of people have preserved their friendships with me without actually agreeing with my opinions. I have realised most people are neither friends nor enemies, they just don't care. Actually caring about anything is in short supply in this world.
  • This is better than I hoped but not ideal. I want to spend most of my time with people who agree with me, so that we can actually fight this problem together. I would rather have a few close allies than lots of people who tolerate me. (Obviously the ideal is to have both.)
  • I am consistently on guard that I should be prepared to lose all my relationships, and keep fighting regardless. This reduces the quality of my relationships even while I still have them.
  • I am always looking for reciprocity in professional connections. If someone asks too much of my time and can offer nothing in return, I will probably cut them off. (If they ask a little time, it is okay.) I have understood professional connections are hence even more emotionally hollow than friendships, atleast in initial stages. Eventually you can work together enough to the point where you're willing to tolerate more imbalance in who is helping who how much.
  • I deeply analysed what type of partner I am looking to marry. You can read my date me document for more. I realised that many of my non-negotiable beliefs essentially come from wanting to fix AI risk, or wanting to acquire knowledge and power above all else.
  • Maybe this is ok, but I need to put in more work if I want to find someone to date. Obviously being chronically unhappy is a major impediment to finding a partner.

2. Death anxiety

• I have partly but not fully internalised the fact that I am going to die.
  • I think doing the 18 day hunger strike was the closest I ever came to fully internalising how bad the situation really is.
  • I think fully internalising it will make even more angry than I already am, and even more of a war monger than I already am. (And I already support cyberattacks and publicly humiliation of senior people at AI companies, so you can only imagine what being even more of a warmonger would be.)
• I think if people around me also internalised their fear of death due to AI risk, that would make it easier for me to also internalise my own fear.
  • Instead I find that most people around me don't care, which is disorienting.
  • Emotions are social, humans are very hard-wired to copy the emotional expressions of those around them.
  • I want to live around people who are also suffering and afraid due to AI risk. This will make me feel more normal when I feel afraid.

3. Steep difficulty curve

• Trying to do something as big as opposing the US govt and the AI companies, comes with a steep difficulty curve.
• Over the last few months, I had to go down a number of failed strategies, then backtrack and pivot.
  • I looked into starting a new ideology to fix AI risk, but then realised this plan will fail as it will take longer than 5 years.
  • I briefly looked into becoming a hacker myself, but realised how hard this will be (unless I get AI to write most of the code for the exploits, instead of me writing it).
  • I looked into giving advice or collaborating with other content creators online, but then realised it is better to make content myself. I am still figuring out how to git gud at making content on youtube around AI risk.
• Obviously this causes a bunch of unhappiness. I am okay with this, and I have scaled steep difficulty curves before. (Cracking JEE, making money in crypto startups, etc.) None of those were as steep as this one, and even doing those had phases that sucked.

4. Zero-sum conflict

• I spend most of my time thinking about zero-sum conflict that will actively cause suffering to people inside AI companies and the US govt.
• This is not my biggest problem. But I can see why working on a more positive-sum project could make me more happy on a day-to-day basis.

5. I don't love my job

• I love philosophing all day. (If you see how long this post is, you probably already get it.)
• I don't love most actual work. I think social media channel is possibly highest impact thing you can possibly do for AI risk. But I don't enjoy working on social media channel that much.

Which of these is most important?

Definitely the loneliness is by far the biggest problem, out of all the problems listed above. If I had that fixed, I could probably tolerate everything else.

Possibly successful solutions

• I have gained atleast some value from deeper theorising about how politics works.
  • I have written atleast 10 pages of thought experiments where I would defect against my perfect clones. For example, imagine that me and my clone were put in a room with an ASI waiting to be controlled, and only one of us could control it. Imagine though that we were not 100% sure the other person was a perfect clone. We would probably have a fist fight to the death immediately, and not even try to coordinate with each other.
  • I have understood at a very deep level that peace between powerseeking actors can only be brokered at gunpoint. There is no such thing as absolute trust between such people.
  • Hence you have to use incentives.
    • If you want to change direction of US govt, create incentives against it by protesting against them, wielding political power against them, owning guns and so on.
    • If you want people in the pause AI movement to work together in any fashion, create incentives so that they gain more power by working together than separately.
    • If you want the fight for power over ASI or human genetic engineering to not eliminate love from this world, create a new political system that can enforce this at gunpoint. Markets and democracy are probably going to fail here.
  • I have realised Moloch is not that actually that complicated.
    • Human values are just love (White Green) and power (Red Blue Black).
    • One person or system prioritises acquiring power and then people in their care can prioritise love instead. This is how capitalism and democracy work, and this is also how many of the political regime changes and wars that I like have worked. If you are a great leader, you ideally want to be someone who mostly values power, but also values love a little.
  • I recently categorised all political ideologies in MtG colour wheel, and then categorised many of my friends, acquaintances and online bloggers on this wheel. It was eye-opening, and gives me better ideas on how to communicate better with them.
  • I have analysed a lot of historical examples via such lenses, be it world war 1 and 2, or the cold war, or british colonisation of the world, or the rise of soviet union, and so on. I have written a bit about these topics, but a lot of it I haven't written yet.
  • I will probably do more such theorising in future, and also learn to get better at communicating my ideas. I probably won't do it full-time though, but will spend atleast a few hours per week on it.
  • Does anyone recommend I become a political philosopher full-time?
    • If I want to find stronger allies, I can see how such writings may help.
    • But also, this feels like a very indirect route, and takes time away from the direct route which is activating people to share, vote, protest, etc directly.
    • Some important aspects of what make a belief system coherent and resistant to attack, can only be learned via practice not theory. Also why I am not keen on becoming someone who only does theory and no practice.

Failed solutions

• Therapy
  • I did learn a bit of useful stuff from therapy, especially around dealing with interpersonal conflict.
  • I did not really learn to fix any of my core problems around AI risk.
  • I think therapists are more willing to accept problems, as opposed to wage war against them. A lot of what they teach are coping strategies not fighting strategies.
  • This is also why I think most therapists follow a deontological code of non-violence. I am not that non-violent.
  • To put it in the language of MtG colour wheel, therapists are Green and I am Black.
• SSRIs
  • I have used SSRIs some years back. I doubt they can fix my problems right now. I have a reasonably deep understanding of all the problems I have listed, and hence I don't think a chemical imbalance alone is the biggest issue.
  • SSRIs have downsides I have personally experienced some years back. Hence I am also more hesitant to try them again.
• Meditation
  • I spent a few months regularly meditating 30 min a day.
  • I don't have any outside view evidence to show this helped at all.
  • I have a constant stream of thought running from when I wake up to when I sleep. I don't think meditation actually slowed or stopped this thought stream.
  • Some people said I am meditating the wrong way, and there is some Advanced Technique (TM) that they know that I need to learn. I don't really see great outside view evidence to suggest I should spend another few months trying it out.
  • I know a few ambitious people like Ray Dalio and Naval Ravikant who speak highly of meditation. This is the strongest datapoint for me to try it again.
  • Practioners of meditation also typically seem to me to be people who accept reality (Green) instead of fight it (Black). Buddhism is explicitly in this direction. I don't think I am capable of accepting reality of AI risk and doing nothing to fight it. Hence I want nothing to do with this ideological cluster.
• Diet, sleep, exercise, sunlight
  • I have consistently maintained these for many months. This didn't fix my unhappiness. Although yes, maintain basics allows me to have decent productive output despite being unhappy.
• Changing location
  • I lived in Kohima for a few months. I am now living in Bangalore for a few months. I have lived in Delhi previously. I also spent a couple months living in Bangkok and Ho Cho Minh City.
  • I think living in a major city where I have a few friends is best for me. I would like to live in SF for atleast a few months to check it out.
  • None of the places actually made me happy. I definitely put some effort into making local friends in each of the places. I have friends where I currently live.
• Changing romantic relationship
  • I have been in one relationship in 2024. It was by far one of the best experiences of my life, and I learned a lot.
  • We have broken up for reasons I don't wanna discuss on the public internet, to protect her privacy.
  • This experience is partly what prompted me to think more deeply about who I want to date next.
  • I am not very optimistic on a relationship fixing any of the major sources of unhappiness listed above. Maybe it could help a bit though, so I am still open to it.
  • There is a lot of outside view evidence suggesting that a relationship can't fix life problems as big as the ones I listed above.
  • Obviously I will even break up with my partner, if I find her actively opposed to my career plans on AI risk. Hence I might have to be on guard even against her, atleast for the few months, and worst case for the rest of my life.
  • Probably this is okay. The types of major ideological changes that cause people to breakup typically happen over time periods of greater than five years, and having a good relationship that gets broken after 5 years, is okay. Not ideal ofcourse, and it is worth pre-emptively figuring out how to avoid this, but worst case if this happens, this seems acceptable. This is just me theorising.
  • I have read about examples of religious differences in a marriage, and have personal experiences around this. I could write a longer post some other day but the short version is just, yeah it sucks if either of the partners prioritise their ideology above their love for their partner.
• Changing friendships and work collaborators
  • I have tried changing friendships a lot. Finding people who care about AI risk, finding people who don't care about AI risk, finding people into tech, finding people into history and politics, finding people who care a lot about me, finding people who don't necessarily care that much about me but offer good advice, etc.
  • I can't obviously publish a list of all my friends here, but I can say that no permutation or combination really fixed my unhappiness or even changed it all that much.
  • I can say that trying to isolate from people was a bad idea (I tried this too). It is better to have a few people around even if you don't talk very frequently.

Subscribe / Comment

Enter email to subscribe, or enter comment to post comment