Home | Search

2026-06-09

I am looking for a cofounder to cyberattack the US govt and US AI companies - Notes to self

Disclaimer

• Quick Note
• Contains politically sensitive info

unprocessed

Note to self

• This document needs more clear separation between my hardline stances and negotiable stances.
  • Hardline stance could be one where I would take large hit to my self-esteem if I did that, and would rather abandon the group and watch the world end.
    • Possibly an example: Leaking info about rape victims of politicians/billionaires, where the victims want us not to leak said info.
    • Quite bluntly I am not sure if I have any hardline stances whatsoever. I seem to generally operate on harm minimisation, so I would start by leaking less info. But if that is failing, I am ready to leak progressively more info, even if that means millions of people get killed in the worst case. Need to think more here.
    • Yeah I think my actual hardline stance is something like - always start by leaking less info not more, and only progressively ramp up if we are say, 2 years away from ASI. Since I can't unilaterally decide this threshold, there needs to be some clear way to decide this threshold. When do we realise that leaking less is not working and we have to leak more?
    • Honestly if the situation is that desperate, I would want the anti-ASI political leaders to tell me what leaks would actually help versus not. It's possible they have way better sense of this than me. If no such leaders exist by then, we are probably fucked regardless of what we do.
  • Hardline stance could be one where the longterm reputation of the group is at stake.
    • Possibly an example: Leaking info about a whistleblower such that it jeopardises long-term trust that other potential whistleblowers have in us.
    • The only people whose trust I truly need to earn are that of whistleblowers we work with. Everyone else, be it our hackers or donors or whoever, they can all go fuck go themselves. Think more here. Why do I think this?
  • Negotiable stance could be one where I have a preference, but if most of the other competent people on our team have a different stance, I may be willing to compromise. I will have limited leverage even within my own team because I am not an expert cyberhacker or an expert fundraiser, atleast at first.
    • Example: I think it will be easier to hire hackers with an anti-war or anti-surveillance stance, than an anti-ASI stance. I might be open to working with them as long as we also leak the anti-ASI info.

Notes to self

• This document is not final until you have gone through atleast 20 of your other writeups, and don't think you have missed anything important

Note to self

• Because you've now introduced a bunch of personal stuff (on motivation and morality) into the doc, maybe you need to demarcate it appropriately so people don't think you're unnecessarily mixing personal and professional life. Some people might have distaste for this, idk. Honestly I would just encourage a (potential) cofounder to also write a similar document about themselves. Whether they show it to me or the public is a separate question, but I think it is useful to write.

Note to self

• Include more examples, atleast those based on public info, and remove them if people ask to be removed. Ask for forgiveness, not permission.

Note to self

• Maybe consider collecting actual stats on the PR/backfire effects - Update - This is Important

Note to self

• Add another meme or two from meme directory, makes it easier to read for others

Note to self

• Sales pitch to hackers
  • My guess is that the median hacker is someone who is primarily in it for the intellectual game / puzzle, someone who gets vaguely annoyed that the topic of morality is being brought up in the first place, someone who dislikes taking orders from anyone else, and someone who respects actual technical competence over someone who just says or writes words.
  • This is obviously a bad cofounder profile but an okay employee profile, and why employees and cofounder are not the same.
  • IMPORTANT - My current writeup seems mainly aimed at myself, and loosely aimed at other EA/LW people, with its neurotic-level obsession on morality and precision and just wordcel behaviour (writing words not writing code). If I have to aim a writeup at hackers it'll have to be a completely different writeup. For instance, my writeup does not contain a lot of technical details on any particular type of attack, nor does it contain details on how to get funding. My writeup is also good for the future historians who can actually judge the effects of my work.

Note to self

• Tighten the editing for the whole intro section, by a lot
  • Consider shifting the section on "why mission statements are fake" somewhere else, it breaks flow. But also, it needs to be near the top so maybe it is fine there?
  • write paragraph long summary (include list of assumptions along with likelihood of being true)
  • The call to action for a potential cofounder needs to be a LOT more tightly edited and appealing to them
  • I still hate how chaotic the introductory section of the post is. This is what is most important for busy readers and potential cofounders. Respect people's time, have a clear call to action.

Note to self

• My plan seems surprisingly justice/fairness-oriented, given that I am someone who openly claims to not care much about justice/fairness but thinks balance of power decides everything.
  • "They screwed us first, therefore we need to screw them over in return" is a very common rhetorical tactic when creating propaganda, and my post too deliberately uses this, atleast when talking about those building ASI and threatening my life, and those in US intelligence who have always been unaccountable. I want to screw over these people, and I am using "they screwed us over first" as a rationalisation.

Note to self

• At some point I should actually go talk to some lawyers about probability of being extradited out of India for hacking, get an actual number on this.
  • Talk to lawyers on what are the risks involved for other people contacting you, either for a brief period of a day or a week, or to actually work together for many months

Note to self

• maybe my document needs a section on my views on technological progress, and also how this connects to low trust versus high trust society?
  • see also: duncan sabien truth or dare
  • Like, it is straightforwardly true that the AI company people have made an extremely violent low trust move by trying to build ASI, and I am planning a violent low trust move in response.
  • It is also true that the holy shit the public is stupid and getting them to agree to any sort of differential technological progress may be hard. Whichever anti-ASI leader seizes power might just shut down technological progress for good, ending humanity's centuries long project since Isaac Newton, and moving us to literal dark ages levels of low trust zero-sum society. This is horrifying, but the fact that I am considering biting even this bullet shows how horrifying racing to ASI is, as well.
  • See also: Everyone higher trust than me is a naive idiot, everyone lower trust than me is scammer or something. I actually loved this insight tbh, I don't know who wrote it.

Note to self

• "I'm unsure if we should attack politicians who are claiming to be anti-ASI but not actually anti-ASI"
  • This section is important and needs to be completed

Note to self

• Maybe the same document should contain more details on technical methods and on funding
  • Technical methods and funding are more likely to update with time as I get more info. Morality and theory of change are more likely to remain more stable.

Note to self

• "Second order propaganda" is a terrible name, think of a better name

processed

Note to self

• Insert "thanos I used the stones to destroy the stones" meme somewhere, it seems topically relevant to using AI to cyberattack AI companies.
• Done

Note to self

• Read atleast one book by Emerson before you go ahead with this plan
• update - done - idk what the fuss is about Emerson, I bounced off of it

Note to self

• I am frustrated with how much time I am spending writing this giant ass manifesto, more than half of which has made no contact with reality, instead of doing actual work that makes contact with reality.
• I don't know, I think LW norms of being too particular about morality have rubbed off on me. I have internalised all their nonsense, and I also genuinely had my own moral uncertainties. This seems like my final attempt to quash everything, so I never have to look at it again. I have wasted the last year arguing on lesswrong and being unhappy due to self-esteem issues. I don't want to waste another few months of my life for the same reason.
• A bunch of this could end up ad-hoc rationalisation anyway.

Note to self

• Holy fuck making a table of contents makes me again realise how schizo this entire writeup reads. I need to organise the sections and sub-sections even more tightly. There are currently too many sections in the document.
• Update - reorged it, the order is still a bit of a mess though, need to reread it to ensure flow is not breaking.

Subscribe

Enter email or phone number to subscribe. You will receive atmost one update per month

Comment

Enter comment