"One of my favorite insights from Tyler Cowen’s book ‘Talent’ is that the most talented people tend to develop their own vocabularies and memes, and these serve as entry points to a whole intellectual world constructed by that person. Tyler himself is of course a great example of this. Any MR reader can name 10+ Tylerisms instantly - ‘model this’, ‘context is that which is scarce’, ‘solve for the equilibrium’, ‘the great stagnation’ are all examples. You can find others who are great at this. Thiel is one. Elon is another (“multiplanetary species”, “preserving the light of consciousness”, etc. are all memes). Trump, Yudkowsky, gwern, SSC, Paul Graham, all of them regularly coin memes. It turns out that this is a good proxy for impact."
I just want to give a bit of gentle pushback to this - it's all a bit self-selecting is it not? Tyler Cowen's vision of a talented person is based on.... himself. And then you mention Musk, Trump, Graham, Thiel etc. who all run in the same/similar online circles and who have a similar world outlook. And given your admiration for Thiel plus you knowledge of the meme culture of those mentioned above I'm going to venture to guess that your cut from that same type of cloth. It's not a bad cloth! I disagree with their politics but there's not denying their success, drive, creativity etc.
But someone massively successful like, say, Warren Buffet or Carlos Slim wouldn't fit this mold in the slightest I don't think. Ditto to Bezos or Zuck, even. Now, I know Cowen's quote you provided said "most" people so you may respond by pointing that out, which is fair enough.
I thought this was a super interesting article, and gave an excellent account from inside a company that a lot of people from the outside have strong opinions about. My comment is really just to say that I feel as though Silicon Valley-types are self-selecting in who they think is successful and what they think success has to look like (memes, language creation, reading old philosophy etc) when in fact much of this thinking/behaviour doesn't exist outside of the insular world of Silicon Valley (Just based on numbers, most people and therefore most successful people aren't born and bread in the Valley!)
I've re-read my comment several times and feel like I'm coming across a bit aggressive and rude and I'm sorry if that's how it seems - I'm really just trying to pushback a bit. Once again, interesting post!
Nabeel, I’m a current Stanford senior in computer science who’s been hovering around defense for the last year or so. The things I see in the current wave concern me. If you have a chance to talk, it would mean a lot.
Awesome read - thank you for the insights. Loved the section on Bat Signals in particular, I would expect it is strongly reliant on a company having a very distinct culture and that is harder to have them when your company becomes clearly associated with "Prestige" by a high percentage of university graduates
Love this piece for codifying and explicating some parts of The World I didn't know so much about!
It also helps me clarify why I, personally, am still not applying at Palantir or OpenAI (though "being in the room" on AI is still a goal of mine, possibly at Anthropic/a research org that could talk with OpenAI or especially Anthropic):
- Memetics, itself, is a complicated and absurdly-80/20-important topic, especially when combined with shit like "game theory" (I-know-you-know-I-know, deterrence, uncertainty, expectations, etc etc...).
If The New Narrative becomes "Trump was right all along, USA USA!" instead of "recruit underpriced talented Trump supporters", that affects who applies and who gets recruited. Similar effects could filter across society. Even if I was an ACX-reading complicated-reasons Trump-supporter (I'm not; see 4th bullet point below), this shit would still scare me in the "what if... not every other recruit is as smart/principled as me?" sense.
Possible trade?: As intelligent complicated-reasons "early adopters" (early Palantir employees, early Trump supporters) have already coalesced, the later recruits could be getting in more based on their views than for their expertise. I'd be short on Palantir in particular doing this, but long on other companies / new startups / the government doing this.
Put another way: If *society at large* becomes more pro-Trump (e.g. through government action, social changes, whatever), then being Pro-Trump is no longer a useful "I'm an independent thinker" signal. It *does*, however, become a nice signal for "I will do 'whatever the power structure wants', paycheck please!".
- Related to "complicated underanalyzed nerdy descriptions of social-type reality" (like game theory, political theory, and the book Impro): Selection effects!
I trust category 3 decisions to "the heavily-rationalist Anthropic" or "the overall LessWrong community, including the broader 'adjacent' circles of LW Trump supporters".
I *don't* trust category 3 decisions to the kind of group that self-selects/gets-recruited-into a company that's "loud about supporting the military, being patriotic, and so on", whose most famous cofounder is a vocal Trump supporter
(Similarly, I *don't* trust category 3 decisions to "whoever's left at OpenAI after *multiple waves* of smart/principled people quitting. It's a near-textbook way to select for mercenaries, e/accs, and loyalists.)
(The current "all of US society as a whole" and "the entire US military" groups are, right now, on the fence about these decisions. Needless to say, I'm pretty nervous about both groups in the coming months, and whatever internal politics could cause extreme outcomes in one direction or another! Also both groups are absurdly heterogeneous, requiring broader studying, etc. etc....)
- My econ/govt beliefs tend towards "Valley-libertarian/normie-centrist/Bay-rationalist-twitter econ" on the micro, and "barebones-but-solidly-Leftist" on the macro. "Incentives matter" and "rich/powerful people try to arrange systems to their benefit" are straightforwardly true in (at least) enough cases to matter (e.g. the silent 99% of "nice people" and 99% of "billionaires not buying the government" exist, but by the power law, we'd expect the 1% exception to dominate biz/govt/whatever, and now we're maybe about to have Elon Musk running DOGE while taking calls with Putin with Trump).
This flows into the first three bullet points, but is not necessary to motivate them. It's more of a sanity-check/consistency check on those points / my entire belief system.
Great thorough read, I especially enjoyed learning about the FDE/PD system. Interesting that so many consultancies (like Accenture that u mentioned) do the first part but never the second.
If I had to tweak 1 tiny thing, I'd say google some synonyms for "hilarious" 😅
Wow, this was so interesting to read. So glad you wrote it.
Gonna go read those books (impro, etc)!
I don't remember reading anything as interesting as this in a long time.
Thanks for sharing
What are the modern Palantir equivalents as of 2024, in your opinion?
Great piece, Nabeel. Good luck for your startup!
Thank you for writing this!
You wrote: "The police have to enforce crime". The Police don't enforce crime, they enforce the law. Cheers, Donald
corrected, thanks
Came here to say this
"One of my favorite insights from Tyler Cowen’s book ‘Talent’ is that the most talented people tend to develop their own vocabularies and memes, and these serve as entry points to a whole intellectual world constructed by that person. Tyler himself is of course a great example of this. Any MR reader can name 10+ Tylerisms instantly - ‘model this’, ‘context is that which is scarce’, ‘solve for the equilibrium’, ‘the great stagnation’ are all examples. You can find others who are great at this. Thiel is one. Elon is another (“multiplanetary species”, “preserving the light of consciousness”, etc. are all memes). Trump, Yudkowsky, gwern, SSC, Paul Graham, all of them regularly coin memes. It turns out that this is a good proxy for impact."
I just want to give a bit of gentle pushback to this - it's all a bit self-selecting is it not? Tyler Cowen's vision of a talented person is based on.... himself. And then you mention Musk, Trump, Graham, Thiel etc. who all run in the same/similar online circles and who have a similar world outlook. And given your admiration for Thiel plus you knowledge of the meme culture of those mentioned above I'm going to venture to guess that your cut from that same type of cloth. It's not a bad cloth! I disagree with their politics but there's not denying their success, drive, creativity etc.
But someone massively successful like, say, Warren Buffet or Carlos Slim wouldn't fit this mold in the slightest I don't think. Ditto to Bezos or Zuck, even. Now, I know Cowen's quote you provided said "most" people so you may respond by pointing that out, which is fair enough.
I thought this was a super interesting article, and gave an excellent account from inside a company that a lot of people from the outside have strong opinions about. My comment is really just to say that I feel as though Silicon Valley-types are self-selecting in who they think is successful and what they think success has to look like (memes, language creation, reading old philosophy etc) when in fact much of this thinking/behaviour doesn't exist outside of the insular world of Silicon Valley (Just based on numbers, most people and therefore most successful people aren't born and bread in the Valley!)
I've re-read my comment several times and feel like I'm coming across a bit aggressive and rude and I'm sorry if that's how it seems - I'm really just trying to pushback a bit. Once again, interesting post!
For BD engineers tend to have high pain tolerance, you mean FDE ?
Nabeel, I’m a current Stanford senior in computer science who’s been hovering around defense for the last year or so. The things I see in the current wave concern me. If you have a chance to talk, it would mean a lot.
Awesome read - thank you for the insights. Loved the section on Bat Signals in particular, I would expect it is strongly reliant on a company having a very distinct culture and that is harder to have them when your company becomes clearly associated with "Prestige" by a high percentage of university graduates
Thanks for this. Good read.
Your comment on data gatekeepers is the most fascinating insight. This is also what drives value for shareholders at the macro level.
Thank you! Best of luck with the venture.
Love this piece for codifying and explicating some parts of The World I didn't know so much about!
It also helps me clarify why I, personally, am still not applying at Palantir or OpenAI (though "being in the room" on AI is still a goal of mine, possibly at Anthropic/a research org that could talk with OpenAI or especially Anthropic):
- Memetics, itself, is a complicated and absurdly-80/20-important topic, especially when combined with shit like "game theory" (I-know-you-know-I-know, deterrence, uncertainty, expectations, etc etc...).
If The New Narrative becomes "Trump was right all along, USA USA!" instead of "recruit underpriced talented Trump supporters", that affects who applies and who gets recruited. Similar effects could filter across society. Even if I was an ACX-reading complicated-reasons Trump-supporter (I'm not; see 4th bullet point below), this shit would still scare me in the "what if... not every other recruit is as smart/principled as me?" sense.
Possible trade?: As intelligent complicated-reasons "early adopters" (early Palantir employees, early Trump supporters) have already coalesced, the later recruits could be getting in more based on their views than for their expertise. I'd be short on Palantir in particular doing this, but long on other companies / new startups / the government doing this.
Put another way: If *society at large* becomes more pro-Trump (e.g. through government action, social changes, whatever), then being Pro-Trump is no longer a useful "I'm an independent thinker" signal. It *does*, however, become a nice signal for "I will do 'whatever the power structure wants', paycheck please!".
- Related to "complicated underanalyzed nerdy descriptions of social-type reality" (like game theory, political theory, and the book Impro): Selection effects!
I trust category 3 decisions to "the heavily-rationalist Anthropic" or "the overall LessWrong community, including the broader 'adjacent' circles of LW Trump supporters".
I *don't* trust category 3 decisions to the kind of group that self-selects/gets-recruited-into a company that's "loud about supporting the military, being patriotic, and so on", whose most famous cofounder is a vocal Trump supporter
(Similarly, I *don't* trust category 3 decisions to "whoever's left at OpenAI after *multiple waves* of smart/principled people quitting. It's a near-textbook way to select for mercenaries, e/accs, and loyalists.)
(The current "all of US society as a whole" and "the entire US military" groups are, right now, on the fence about these decisions. Needless to say, I'm pretty nervous about both groups in the coming months, and whatever internal politics could cause extreme outcomes in one direction or another! Also both groups are absurdly heterogeneous, requiring broader studying, etc. etc....)
- My econ/govt beliefs tend towards "Valley-libertarian/normie-centrist/Bay-rationalist-twitter econ" on the micro, and "barebones-but-solidly-Leftist" on the macro. "Incentives matter" and "rich/powerful people try to arrange systems to their benefit" are straightforwardly true in (at least) enough cases to matter (e.g. the silent 99% of "nice people" and 99% of "billionaires not buying the government" exist, but by the power law, we'd expect the 1% exception to dominate biz/govt/whatever, and now we're maybe about to have Elon Musk running DOGE while taking calls with Putin with Trump).
This flows into the first three bullet points, but is not necessary to motivate them. It's more of a sanity-check/consistency check on those points / my entire belief system.
Wow, I came to your Substack from Medium. This was an amazing read. I also saved the books you mentioned and I am planning to read them.
A post that I could not stop reading to the end; Rich content, amazing insights !!
Great thorough read, I especially enjoyed learning about the FDE/PD system. Interesting that so many consultancies (like Accenture that u mentioned) do the first part but never the second.
If I had to tweak 1 tiny thing, I'd say google some synonyms for "hilarious" 😅