4 ms·

I have mixed feelings about this, mostly because the stakes on giving people persuasive advice with an enviable track record in hand might be at its highest in

I have mixed feelings about this, mostly because the stakes on giving people persuasive advice with an enviable track record in hand might be at its highest in a room full of 14/15-year olds. This is exactly the kind of thing that made me dramatically more ambitious and harder working when I started reading pg essays (pushing 20 years ago now), so I have a huge soft spot for the feeling/nostalgia that I got when I just let it wash over me. And it’s ultimately why I think it nets out constructive in spite of some real problems: if I heard this at 15 I would have started working really hard at 15 and not 22 or 23, and I would have started cultivating relationships with people who were going to be powerful at 15, instead of arguably antagonizing powerful people until, well maybe still today [1]. It’s got some serious problems though: the way most people get rich is by having wealthy parents, though at the apex it’s having wealthy parents who raised you to work hard and manage relationships with powerful people well [2], so given that by 15 most of that already has or hasn’t happened, I think we can generally regard pg saying that as a “useful fiction”: outcomes will improve at the mean for 15-year olds who believe him about this, or such is my prediction. The advice about university is extremely dangerous in 2024 for anyone without a privileged upbringing because university is a far inferior place to learn things than YouTube videos aggregating all the best university lectures in history with the even more effective education style pioneered by Grant with 3blue1brown. And while trying to achieve at a level where you go to Stanford without a mountain of uniquely nasty debt [3] is strictly a good thing in high school / primary school, actually going to university via debt is an extremely dicey proposition if you miss and that distinction is tough to sell to that audience. Better advice would be to start learning from the best educators in the world via a YouTube Premium subscription concurrent to high-school studies (and making them whiffle ball on a curve compared to peers that don’t), try to get a scholarship to an Ivy or Cal or something, but go to community college or directly into the workforce (Work for a Startup™) if you miss and don’t have graduate study on the agenda. Mostly the feeling I get is that pg and his essays these days sit in the Venn between “well intentioned”, “aspirational/idealistic” and “not true”. It’s a fine line sometimes between “aspirational” and “false”, it gets finer yet in light of Sinclair’s Maxim [4]: what might be dismissed in a low-effort way as “out of touch” or even “self-serving” is trivially unfair to pg. He has his own family he could be spending time with instead of speaking to young people, he doesn’t need or seem to want any more money. He’s just past the point where he can de-leverage his intellectual and emotional investment in a world that doesn’t exist anymore: he’s the LTCM of serious intellectuals in technology. The book about that is called “When Genius Failed”, and John Meriwether is a brilliant man who lived by an admirably robust ethical code in a very rough neighborhood. [1] The most significant glitch in pg’s firmware is that he says “smart and hardworking” when he really means “good at becoming powerful”, a direct quote [5] about his protégée illustrates this very clearly. [2] Pick a list of very rich people and click on all the links, a fine starting point is https://en.m.wikipedia.org/wiki/The_World%27s_Billionaires https://en.m.wikipedia.org/wiki/The_World%27s_Billionaires. [3] Yes there have been some modest reforms, no it’s not fixed: https://www.bbc.com/news/business-67546893 https://www.bbc.com/news/business-67546893 [4] At the top, a man’s living can be about intellectual as well as financial capital: https://www.goodreads.com/quotes/21810-it-is-difficult-to-get-a-man-to-understand-something https://www.goodreads.com/quotes/21810-it-is-difficult-to-ge... [5] I’ve known too many YC alumni too well for too long to even engage with any hatchet job on a piece that if anything understates the reality: https://www.newyorker.com/magazine/2016/10/10/sam-altmans-manifest-destiny https://www.newyorker.com/magazine/2016/10/10/sam-altmans-ma..., please don’t prey on my worst instincts by tempting me to cite documented examples and credible primary sources.

keiferski 3y ago
Watching videos on YouTube is in no way comparable to going to a quality university, being around likeminded ambitious young people, and interacting with professors and researchers directly.
- benreesman 3y ago
  That's not the comparison I was making. I said consuming a meticulously curated list of the best university lectures on any topic delivered by an extremely rare breed (unparalleled research professors who are also competent as educators) is a better way to consume university lectures and requires a phone and $15.99/month. I furthermore said that Grant of 3Blue1Brown fame and others inspired by him (Richard Behial, Sean Carrol, and many others, this matters enough that I'll do the rigorous citation routine if people personified in this instance by you don't regard this as common knowledge) have used the lever of technology to increase the upper-bound on the efficacy of education in a way that dwarfs the technological lever that LLMs represent for being wrong at scale persuasively enough to convince anyone but an expert in a bad mood. I explicitly acknowledged the dominant term in the kinds of outcomes pg is talking about, which is having and effectively managing relationships with people who either already are or soon will be with high probability powerful. In 2013 I was shortlisted as a candidate for some of the top jobs at the hottest company in the business and on a polynomially steeper career ascent trajectory than someone who now sits on OpenAI's board: that person diligently managed relationships and had a keen sense for whose star was rising and falling, I produced 10^4 more marginal revenue (very conservatively) but gave the establishment (including too many of my colleagues who chose the other path on geek/climber) the finger and didn't manage so much as my personal life, let alone the establishment. Which is why that person is on OpenAI's board and I'm scratching out a living as a consultant, and even that is only possible because I'm finally learning twenty years too late to be at least thoughtful, reasonably polite most of the time, and well-cited when being skeptical of an establishment. The most effective way to interact with cutting-edge researchers short of actually getting a PhD at Oxford or something is to work for or consult for public-private partnerships between top academic labs and private-sector spinoffs. Even as a marginalized consultant I've worked directly with people at top research universities who get whispered about as Nobel candidates, these companies are often under-resourced in the beginning and if you're willing and able to represent a bargain, you can work adjacent to some truly stratospheric levels of elite research. You're not wrong that going to an excellent school that you or your parents can actually afford is among the best advantages one can have early in life, you're exactly right about that. You're incorrect in demonstrable ways on the rest of the stuff you've bundled into one statement. This is the problem with bundling together tangentially related ideas into one atomic unit. I've kicked this thing back and forth a few times because even on revision I'm finding it difficult to say "What you Can't Say" [1] with both of: concrete and documented examples and low scope for seeming rude or snarky. I don't intend snark, but pg wrote those essays for a reason: being right is offensive sometimes. So I apologize if I give any offense to you or anyone else, but all this is true and a lot hangs in the balance right now for the industry and civilization as a whole. And I regret that you personally wind up being the person who said the thing that many/most think that needs demolishing, there's absolutely nothing personal in it. [1] https://paulgraham.com/say.html https://paulgraham.com/say.html

keiferski 3y ago

Watching videos on YouTube is in no way comparable to going to a quality university, being around likeminded ambitious young people, and interacting with professors and researchers directly.

benreesman 3y ago
That's not the comparison I was making. I said consuming a meticulously curated list of the best university lectures on any topic delivered by an extremely rare breed (unparalleled research professors who are also competent as educators) is a better way to consume university lectures and requires a phone and $15.99/month. I furthermore said that Grant of 3Blue1Brown fame and others inspired by him (Richard Behial, Sean Carrol, and many others, this matters enough that I'll do the rigorous citation routine if people personified in this instance by you don't regard this as common knowledge) have used the lever of technology to increase the upper-bound on the efficacy of education in a way that dwarfs the technological lever that LLMs represent for being wrong at scale persuasively enough to convince anyone but an expert in a bad mood. I explicitly acknowledged the dominant term in the kinds of outcomes pg is talking about, which is having and effectively managing relationships with people who either already are or soon will be with high probability powerful. In 2013 I was shortlisted as a candidate for some of the top jobs at the hottest company in the business and on a polynomially steeper career ascent trajectory than someone who now sits on OpenAI's board: that person diligently managed relationships and had a keen sense for whose star was rising and falling, I produced 10^4 more marginal revenue (very conservatively) but gave the establishment (including too many of my colleagues who chose the other path on geek/climber) the finger and didn't manage so much as my personal life, let alone the establishment. Which is why that person is on OpenAI's board and I'm scratching out a living as a consultant, and even that is only possible because I'm finally learning twenty years too late to be at least thoughtful, reasonably polite most of the time, and well-cited when being skeptical of an establishment. The most effective way to interact with cutting-edge researchers short of actually getting a PhD at Oxford or something is to work for or consult for public-private partnerships between top academic labs and private-sector spinoffs. Even as a marginalized consultant I've worked directly with people at top research universities who get whispered about as Nobel candidates, these companies are often under-resourced in the beginning and if you're willing and able to represent a bargain, you can work adjacent to some truly stratospheric levels of elite research. You're not wrong that going to an excellent school that you or your parents can actually afford is among the best advantages one can have early in life, you're exactly right about that. You're incorrect in demonstrable ways on the rest of the stuff you've bundled into one statement. This is the problem with bundling together tangentially related ideas into one atomic unit. I've kicked this thing back and forth a few times because even on revision I'm finding it difficult to say "What you Can't Say" [1] with both of: concrete and documented examples and low scope for seeming rude or snarky. I don't intend snark, but pg wrote those essays for a reason: being right is offensive sometimes. So I apologize if I give any offense to you or anyone else, but all this is true and a lot hangs in the balance right now for the industry and civilization as a whole. And I regret that you personally wind up being the person who said the thing that many/most think that needs demolishing, there's absolutely nothing personal in it. [1] https://paulgraham.com/say.html https://paulgraham.com/say.html