1) This Sankey chart is missing some layers, and as a result it's kinda misleading. Some examples:
- I'm no longer at Anthropic, and many others are no longer at OpenAI.
- Tom Brown didn't jump straight from a group dating app to co-founding Anthropic; he spent several years at DeepMind and OpenAI.
- Brian Krausz worked several places in the ~decade between his YC startup and Anthropic. He then left Anthropic but was brought back in with the Stainless acquisition.
- Stefan Stokic ran a non-YC startup for five years before joining OpenAI.
2) I don't think 35 people joining Anthropic or 70 joining OpenAI at some point after doing YC is a substantial concentration at a macro level. That's <1% of YC founders, even if you include those who no longer work at either of the companies.
3) As for why people joined: beyond the compensation, it isn't clear that a startup is more impactful than powerful AI in their domain. Here's an excerpt from an email I sent to one of the Anthropic co-founders when I was considering joining in 2024:
> For context, my YC company builds dev tooling for bioinformatics. [...] I no longer have the conviction that our niche-specific dev tooling is going to have a significant impact on the field as we approach expert-level AI.
Ultimately, the core value of our product was actually supplanted by LLMs; bioinformaticians are much better at writing code and managing infrastructure with LLMs.
I think more and more folks are seeing the impact of what they'd build as a subset of what powerful AI could do, and as a result they end up joining a lab.
Well yeah, but being a CTO at a major org isn't exactly a low-paying gig.
The interesting thing to me is that the belief that Anthropic is going achieve something like AGI appears to have really spread through SV. If not AGI, at least greatly disturb their old cushy SaaS jobs.
Also, I still do believe some people just want to work on "cool" stuff.
edit: not just CTOs and CPOs, I mean John freakin Jumper, who got the 2024 Nobel Prize in Chemistry, left his VP role at DeepMind to have no reports, and be an IC at Anthropic.
> Also, I still do believe some people just want to work on "cool" stuff.
One thing I’ve noticed more and more is all the “cool” jobs require you to already be an established domain expert.
Another things I’ve noticed, and this is probably mostly just my opinions evolving, but all the “cool” stuff that catches my interest these days is not in software. Software is boring.
I’ve been in a bit of creative/intellectual rut the last few years, but before that there were a handful of things that intrigued me in the last 5-10 years.
Most recently, I became interested in the field of photonics/optics, particularly in the field of novel optical communications research.
I’ve had mild interests in things like
Biology and History since I was a child. The latter particularly hooked me a few years back.
Neuroscience is a pretty cool one as well. The human brain is of course one of the most fascinating pieces of hardware in existence and understanding it (to whatever level we currently can) is incredibly interesting to me.
At one point I thought about going into security research, but now that everyone and their mother is headed that direction, it’s become less feasible.
Unfortunately I don’t even have a bachelors degree, and most of the fields are research heavy and not very accessible to hobbyists, so my actual interaction with these things remains relegated to reading papers/pop sci articles/etc.
Don't forget the study of neuroscience is much larger than just the human brain. Many other target organisms are studied e.g. fruit fly and are much more thoroughly understood on a fundamental level.
Respectfully, when people want to understand the human brain they usually intend to study sociology or psychology rather than neuroscience. Although, of course, neuroscience is applicable to human brains as well.
> If not AGI, at least greatly disturb their old cushy SaaS jobs.
I mean, this is happening right now.
I doubt there are few people talking to customers and potential customers that haven't run into the "look what I rolled last week with Claude" only for it to be legitimately impressive. The idea of technical moats is evaporating in front of our eyes.
The biggest AI bubble seems to be forming within the minds of the people working in SV itself, judging by the talk about the inevitability of AGI, the strong language around fomo (permanent underclass and all that) and the close proximity of the whole ecosystem there.
So wether it's belief in AGI or a desire for building cool stuff, as an outsider looking in, it seems to be informed by the same echo chamber of thought.
I think it is more than just money. Having a hand in steering the ship that they think will change the world is pretty attractive. I came out of retirement just to watch the spectacle.
> I came out of retirement just to watch the spectacle.
While I certainly have no hand in steering the ship, agentic building brought me out of hibernation/mild depression. I am neither young, nor rich. However, I still had lots of product ideas, and now I am able to create and test them in days. And yes, I am actually making sales now.
I got into this early at Sonnet 3.5 release time, so I went through initial AI psychosis almost two years ago, but still today, I am more excited to wake up and work than I’ve ever been in my life.
I think most of HN can't logically arrive at this conclusion because they don't exactly buy into the premise that
(1) (digital) AGI is a real concept, and will be achieved imminently
(2) AGI will be world (and possibly more) changing.
(3) Anthropic will be the one to do that, or at least capture a large part of it
If you buy into all of these, which I'm assuming a lot of SV does at this point, then its kind of a no brainer to go to OpenAI or Anthropic. And optics wise it seems like Anthropic is "winning" right now, so thats the lab with the talent flow.
If Anthropic actually creates a singleton, contributing to that is way more impactful than being CTO at some random company (though I may disagree that said singleton would treat equity holders any differently).
Do you have examples? Because besides the other reasons mentioned about huge earning potential and being part of an important moment in history, the not-so-hidden secret is that any management position below C-level and above small team manager is usually extremely shitty. You are responsible for cat herding but rarely get to make the decisions in the first place. There are very few people IMO who desire or excel at "mid-high level" management positions (again, levels from director to anything below CTO) - I've seen people who enjoy and are good at leading small teams, but people at the director level are either there because it's lucrative or with the hope to get to CTO level ASAP, but otherwise it's one of the worst jobs in the "stress to moments of enjoyment" ratio.
So I could imagine tons of people in those roles who would love to be ICs again if there was also a chance at outsized wealth.
I would do that in a heartbeat if I was in that situation.
Having a major leadership role at a big company is kind of a slog. Getting a chance to go back to a fast growing startup as an IC in a way that will only benefit their career (ex-Anthropic or ex-OpenAI makes your resume gold plated) is an easy choice. They can go back to any job they want after this.
I think this might only be shocking to some people who have forgotten that many people in these positions actually enjoy working on hard things and being surrounded by other like minded people. If you’ve only worked in po-dunk startups or other companies where leadership looks like a Game of Thrones style fiefdom building game it’s harder to imagine executives who are smart, capable, and enjoy working hard. It’s pretty common in Silicon Valley despite all of the sneering here.
I disagree. I'd immediately have a ton of questions for most ex-OpenAI and ex-Anthropic people as my default position would be "poor business/economics judgement in role, chose to ride a hype cycle, now sat in front of me for a reason".
I'm not saying it's toxic to have it on your work history, but it's far from gold plated. They'd have work to do for me, and for many, many people in the industry, to prove that I'd want to work with them.
You are not in a position to hire them, let alone at a big interesting company. Forgive my bluntness, but your opinion doesn’t matter to their situation.
When I say their resumes are gold-plated I mean at the companies they want to work at.
I'm not surprised, but what is surprising is we're now in a period in which the mask is slipping and no one seems to care enough to try to put it back on.
For all the hubris and arrogance of the tech scene, and especially the Bay Area's, founders at least pretended to pursue a greater calling than money by leaving cushy gigs to pursue a new venture they were passionate about. Sure, that only lasted a handful of years before it became a gold rush to move to the Bay for YC and pursue the startup-to-acquisition pipeline, but they at least pretended to care.
The naked pursuit of wealth at the cost of potentially tearing apart society is mildly alarming because the scale it's happening at is so big.
It seems like different people? The early tech scene was defined by people who got into tech because they loved it, jobs were paying under six figures. Then it became a gold rush, and the tech scene became defined by the newcomers who were after cash first, tech second. It's not that the old timers gave up on their ideals (well, some did), they were simply supplanted by larger numbers of people following the money.
In my opinion there is no "slipping mask", merely a changing of the guard that always happens when massive piles of generational wealth are at stake.
Nah, we knew back in the 90s that the tech industry had open disdain against American workers (pressuring congress to expand immigration to depress wages) and children (supporting charter schools while attacking children (yes when you attack public schools you're attacking the only people that benefit from it: children)). Not too mention all the horrid, extremely hostile business pursuits we saw from companies like Oracle + MSFT + Apple.
SV has always operated on exploitation of their workers, they just hid it well from the public.
Having been there i have to respectfully disagree. Probably because, as far as i can tell, you are warping their intentions to be what your intentions would be if you did what they did.
I sometimes think, speeding up the Red Queen hypothesis to accelerate Cambrian Explosion by printing money from the top so that high-agency people would develop new products and services to capture more of the newly created money is a good thing?
I just don't understand how to connect this analysis to what's happening. What OpenAI and Anthropic both say, loudly and repeatedly, is that they're working on one of the most impactful technical problems in the history of humanity. The reason they have lots of money is that lots of investors believe they're right.
Perhaps you've seen under their mask, and you know they're scamming the investors. After the NFT bubble it's a plausible hypothesis. But I don't know what it means to say that the mask just isn't there and it's naked pursuit of wealth.
Whatever you think of AI for your own work, the fact that the entire economy is betting everything on it is really concerning. It’s not just the fact that it may not work well economically speaking and would end up with a market crash, or all the negative externalities from AI development, it’s also the opportunity cost we are paying. With so much human capital and resources dedicated to developing and running LLMs all the other business, research opportunities aren’t being explored and invested in
I prefer that scenario to the one we had before where too many smart people were building out social networks and similar services trying to maximize ad impressions.
We are still dealing with the consequences…
I don't know, I feel like llms are the even worse replacement for social networks and maximized ad impressions. Now you don't even need a network or users to produce the addictive content and inject narrative that is so impactful on our brains and lives
I have more meaningful intellectual conversations with LLMs in a given day than in a month on social networks. And the LLMs never try to make me feel angry or stupid, they never ignore my arguments to go on their own soapbox, they almost never feed me misinformation (and never deliberately), and they've yet to tell me to my face that the purpose of their interactions with other people is not to convince or even communicate with those people but to perform for an audience (often one that doesn't feasibly exist).
Some caveats here: my view of social media is biased by the fact that jerks tend to pull me out of the woodwork- I get ragebaited, although I think I avoid the literal bait- so many of my interactions are with jerks. Looking around, I expect I'm not alone, but this may not be the default experience. And the LLMs certainly aren't perfect. They can hallucinate, they're often sycophantic, they're bad at judgment in a number of ways, probably others I'm missing.
Compared to actual people, they suck. Compared to social media as it exists today, they have orders of magnitude to fall before your comparison makes sense to me. And, at least for now, the LLMs are the ones getting better while the social networks are- AFAICT, universally- getting worse.
> I have more meaningful intellectual conversations with LLMs in a given day than in a month on social networks
Social media felt this way in the early days, too.
> they almost never feed me misinformation (and never deliberately)
This is dangerously untrue. LLMs will deliberately tell you the things you want to hear. They even "call you out" in exactly the ways that are most palatable to you. There is ample evidence that LLMs can be biased. The ability to broadly influence individual opinions, without even needing bot accounts on social media is exactly one of the reasons so many people who want power/money are flocking to control these LLMs.
> so many of my interactions are with jerks
This is one of the ways that the LLMs are "scarier" and more powerful than social networks. Real people are more likely to be obviously pushing agendas or otherwised easily dismissed.
> LLMs are the ones getting better
Getting better at what? The things you list are exactly the things that make LLMS subtly powerful and far more dangerous that social media. It's the reach and anonymity of social media without needing to control a feed algorithm, deal with content or people that doesn't fit your agenda, or produce fake content that does.
> I have more meaningful intellectual conversations with LLMs in a given day than in a month on social networks.
This is kind of making the parent's argument for them about the dangers...
> And the LLMs never try to make me feel angry or stupid, they never ignore my arguments to go on their own soapbox, they almost never feed me misinformation (and never deliberately), and they've yet to tell me to my face that the purpose of their interactions with other people is not to convince or even communicate with those people but to perform for an audience (often one that doesn't feasibly exist).
You're basically asking to be coddled, affirmed, validated, and protected. The world has conflict. Not all conflict is bad. Yes, social media weaponizes that conflict for ads, but I think the opposite is more dangerous in creating completely distorted realities. What you're describing sounds worse than the worst nightmare I've ever had, a world in which I'd be constantly disconnected from everyone else and would be disillusioned into thinking I'm not.
> Some caveats here: my view of social media is biased by the fact that jerks tend to pull me out of the woodwork- I get ragebaited, although I think I avoid the literal bait- so many of my interactions are with jerks.
>You're basically asking to be coddled, affirmed, validated, and protected.
Incorrect. My interactions with LLMs simply don't look like this. I don't know where you got this idea from what I said. I have a prompt that begs the thing to call me out, and it currently does so too much, which is exactly what I prefer.
What it doesn't do is rub minor disagreements in my face, jump to conclusions about who I am based on pre-existing beliefs, or tell me to go to therapy when I earnestly admit my humanity.
My guess is these AI companies are counting on the money to come from traditional SAAS style services only. This whole AI would be used to gather enough capital to rewrite most of the existing SAAS services under one umbrella, sort of super APP concept. The SAAS businesses doing the layoffs will soon realize that for this whole time they were just pouring their money into their own demise
> I remember lamenting all through the 2010s that all the high paying jobs were basically ad revenue supported.
That's never been true for financial services, many of which offer very high-paying jobs for high-frequency trading, hedge fund management, market operations, etc. There are many other industries with high value IT jobs that are not ad-supported.
> too many smart people were building out social networks and similar services trying to maximize ad impressions.
Honest question: where should these very smart people work then in your opinion? Nearly all suitable-looking, "smarter" ideas that come to my mind will bring less revenue/profit in expectation than, for example, letting them maximize ad impressions.
I of course wish we lived in a "more intelligent" world where this is not the case. :-(
Call me crazy, but the smartest people I know have rejected maximizing their employer's revenue and profit as their goal. Or even their own company if they go that route. Find a niche, work it, enjoy it, and enjoy the rest of your life, too.
The problem with this is people investing their time/expertise/money/energy into the common good and then have a small number of people ruining the benefit for their own ends, knocks the well-meaning out of the most positive people. The common good only works well when everyone wants the common good.
You seem to have independently discovered the market failure commonly know as the Free-rider problem, for which there are many solutions, and which is not a dispositive preemptive argument against the establishment or management of common good(s).
For example, Elinor Ostrom won a Nobel Prize for her work illuminating design principles illustrated by long enduring common pool resource institutions.
> Ah, then my second wish will be for everyone to want the common good.
People have very different, often contradictory opinions what the "common good" is.
I would even claim that this question is a special case of the millennia-old philosophical problem of defining what "good" actually means (i.e. defining ethics), which has resulted in myriads of different opinions:
> where should these very smart people work then in your opinion?
Anywhere else. Ads are a net negative to society and provide no value. ANY activity outside of pushing ads is a better contribution to society as a whole.
Probably not on systems that rot democracy, freedom, and human expression. Companies like Google and Meta are net negative for humanity and the people that built these companies and continue to work for them are evil in their apathy towards the human cost of their salary.
That will certainly come in time. We're already starting to see/hear the beginnings of it. My hope is that the market maintains competition because the moment one or two corps have monopoly/duopoly they will go full throttle toward that I'm afraid.
I'm not even sure that I do prefer the AI train to the social media / ad train. I certainly would if the direction was "we're throwing ourselves blindly in AI so to gradually free citizens from the need to work", but right now it's "we're throwing ourselves blindly in AI so that we can sacrifice workers for higher corporate profits".
Not to mention how they will be used by terrorists and evil people to create biological weapons and other tools of mass slaughter. Or how it will be/is being used by criminals to target vulnerable people and businesses. Or how coupled with drone warfare the world becomes an unlivable hellhole in the next decade. But Marc Andreesan said only losers are against “tech accelerationiism” on twitter so here we are.
You can, right now, find out how to synthesize VX or sarin, or how to weaponize anthrax. You can and have been able to acquire this knowledge without the use of AI. Knowledge is not the obstacle to proliferation of biological and chemical weapons. It's hard to set up the infrastructure. See Aum Shinrikyo.
If you believe that AI is powerful enough to help the good people (however you define that), why isn't that also coupled with the belief that it is powerful enough to help the bad people (ditto)? Either the capability of this tech is completely overblown or it can help a wide swath of people including those with conflicting ideologies.
Sounds like you're reading into Amodei's propaganda as to why they should control AI. The information to make biological weapons for example is nothing new, it has been on the Internet for decades.
Should we control nuclear weapons? Is nuclear technology of the sort we’re fighting a war over right now a bigger technological advance than AI when it comes to war fighting capabilities? I suppose if you think ai is nothing big, it makes sense to not want to control it. Or, if you are all for widespread nuclear proliferation, then your position would make sense. ELSE: you’re totally illogical.
An LLM can't help you build a nuke. LLMs can't paper over hard engineering challenges like "You need to reinvent the centrifugal separator industry" and "rocket engineering is pretty difficult" and "you need really really good data to simulate things before you can actually build an implosion device" and less engineering challenges like "The CIA keeps murdering anyone who works on this"
There's never been an information access problem with nuclear weapons. The information was so public that most physicist realized everyone was building bombs from the sudden lack of published papers on how you could build bombs.
It's just silly how people seem to believe that if you just have a page worth of text saying something, you can suddenly do that thing.
Even a magical superintelligent AI could not do these things. It's going to tell you the exact same thing that has been known for nearly a century now:
"Hire 100 engineers and a couple physicists and 1000 very skilled machinists and laborers, and spend a billion dollars on plumbing, and then do that a second time if you need to deliver the device using a missile"
Meanwhile a dirty bomb is trivial.
These things don't happen because people who want to cause mass harm are generally poorly resourced.
I mean there are ads in certain chatGPT plans. There have already been multiple studies showing that regular use of genAI makes you less intelligent...
Additionally, it's pretty clear that the owners of this tech (dario, altman, etc..) do not have the best interest of "the people" at heart and only care about accumulating as much power and wealth possible. Not that it's their responsibility to save humanity or something, but I'm not sure this future is much better...
"The smartest people" were not working on ads because building an advertising behemoth is not a smart thing to do.
When there's nothing good being invested in, "the smartest people" are simply underutilized. They do things that they think are good instead.
"The smartest people" have never sought maximum capital accumulation, and have not historically been driven by "highest wages" jobs.
The computer industry formed out of WW2 because a bunch of smart people ran into each other and did what they thought was really cool and had realized was suddenly possible. Von Neuman literally met one of the other guys and discussed computers as a theory at a train station and that meeting was influential on the future industry. They did not try to make billions of dollars. They had fairly average jobs that paid average. Often working for the government itself.
"The smartest people" did not work on targeting ads at people, and while some of them may have been involved with early transformer work, they certainly would be disillusioned with LLMs by now and would not be working at companies burning trillions on them.
This both destroys discourse, and centralizes the means of production into ~4 companies, with everyone else paying them an economic tax in order to be competitive.
None, but for actual communication by lay people, it cost a ton of money for audio and video calls by having to pay their phone service provider a lot, or calling cards, or whatever.
> With so much human capital and resources dedicated to developing and running LLMs all the other business, research opportunities aren’t being explored and invested in
It probably feels this way if you only see headlines like this, but the LLM companies and even extending into their data center buildouts are still only a part of the economy. Investment dollars get spread around far and wide. Investors can’t even get allocations in these companies if they want to because they’re so oversubscribed.
This headline and website are a good example. If you think being a YC founder is rare and reserved for the truly elite, you may not have seen the size of YC batches lately. The number of YC founders is in the tens of thousands. Showing a list of 100 of them who joined a couple big tech companies doesn’t mean anything by itself.
It’s on the order of 1% of YC founders. YC startups don’t work out all the time and their employees and founders go work for other startups. Barely worth noting, if not for a website making it look like a big deal.
I fear another related possibility. With *so* much money on the table, if this new industry shrinks or collapses, it is hard to believe that it will collapse same way dot-com did, and instead investors will seek to protect their capital who knows at what cost. Given the influence they have it is not hard to imagine the impact on policies, etc. I fear that tax payers are not the winners in either case, whether this industry grows or not.
This reminds me of the office rental leases during and after lockdowns where it seemed like the idea that owners might loose money was just not even on the table and we needed people to be back at the office otherwise the value of office buildings would fall and that'd be terrible...
> Whatever you think of AI for your own work, the fact that the entire economy is betting everything on it is really concerning. It’s not just the fact that it may not work well economically speaking and would end up with a market crash, or all the negative externalities from AI development, it’s also the opportunity cost we are paying.
It's worth pointing out that "entire economy is betting everything on [AI]" you're talking about here is in the economy in the "numbers go up sense" (e.g. the stock market). It's not the economy in the providing for the general well being sense (e.g. gives people jobs so they can get housing and food).
Investment from VC and wealth funds is going pretty much only to AI projects. The massive capital expenditure we are currently seeing is going to AI infrastructure (HBM has pretty much no use case outside of LLMs and a few niche industries). It’s real money being allocated to a single bet
I don’t really buy the opportunity cost argument. The bets we are placing look EV-positive to me.
The 90th percentile upside is way, way higher than is priced into the current stocks, the 10th percentile looks to me like 2000 where a bunch of investors loose their shirts and then for the next few decades new companies boom on all the infrastructure that was built.
The “failure mode” for this tech is that we don’t advance much beyond Sol/Fable and the existing models just get cheaper. If there is an AI bust in hindsight it likely will look like the dotcom bust did, which is a small blip on the decades-scale GDP curve.
Opportunity cost arguments are more convincing when there’s some specific opportunity that you’re comparing against.
Otherwise, it’s sort of like saying that people will find a job doing something else, without any thought to what that something else would be. The jobs don’t create themselves.
I also think this is a huge problem in government. A lot of government services could be digitized via simple CRUD API/UIs and instead politicians talk about AI...
Researchers are increasingly using LLMs and other transformer based AI. My friend at a well known research institution asked Anthropic for access to their biology model (not sure if it's Mythos or some other sort of uncensored model) and recently made breakthroughs with it and other AI.
Yes, I don't have their permission to dox them, so it will have to be vague for now. Still, there are many news articles outlining the same point I made, except with real examples.
It’s not just the fact that it may not work well economically speaking
That's not going to happen. Everything is different now, and the only long-term losers will be those who continue to deny it. Even now we still see stupid analogies to parrots, calculators, toddlers, and so on, being advanced here of all places.
and would end up with a market crash
That might happen, because the first attempt a market makes to allocate funding and spending optimally is almost never on target. Great, it'll give the rest of us a chance to buy in.
I know Ed Zitron is a dirty word on HN most days but having recently read some of his (highly editorialized) work… it seems like he’s right and the numbers just aren’t making sense.
Combine that with the fact that “good enough” is a real thing and cheap Chinese models are either there or close depending on your use case and it’s hard to see how this ends well.
Oh definitely. Personally, I think it will get to the point where we have local good enough models.
I’m just incredibly confused how these companies and investors don’t see the writing on the wall UNLESS the point is to just extract as much money as possible and not be left holding the bag.
Not only that but with so much human capital and resources devoted to AI’s explicit goal of replacing human labor there are two outcomes: 1) it succeeds and we have a serious social problem on our hands with large swaths of the population out or a job, or 2) it fails and we have a very large market and economic crisis as the AI bubble collapses.
Maybe there’s a partial success future that isn’t those two, but I’m not hopeful. And I’d take the Ai bubble collapse over AI BigTech achieving its goals
This analysis doesn't actually prove the title, as it only considers founders that have gone to OpenAI or Anthropic, however even if it did it isn't particularly suprising.
Sam Altman was the president of YC, so it's not particularly suprising that he has hired lots of talent from YC - these are people that are pre-qualified because he has seen their work before outside of a job interview (better the devil you know than the devil you don't).
This happens everywhere - person leaves company A to company B and then poaches the best talent from company A. Obviously in this case the talent are founders rather than staff members, but still.
YC has tens of thousands of founders, more are probably at Google and Facebook than Anthropic and OpenAI. I’m not sure a sample of 105 founders shows… anything?
The question I have is why are these companies hiring these people and what does it say about their hiring practices and the amount of capital they’re poorly spending?
It’s exceedingly unlikely that any of the people who were working on YC startups previously have any real professional experience with any of the following: slurm, collectives, NUMA systems, RDMA, compilers, systems programming, general HPC performance estimation or measurement, CUDA or ROCM or any kind of GPGPU/accelerated computing. But that is the core business of both of these companies.
I’m not surprised that these companies are well funded and hiring a lot of people. I’m surprised that they chose to hire the people who were previously making “Uber but for dogs” gimmick apps and not just hollowing out the HPC specialists from national labs.
People ask the same question of why YC funds yet another Uber for dogs or a button on the Touch Bar that cost $10/mo to help join a meeting faster.
They invest in people more than ideas, so you’ve got, at least in many cases, people with good pedigree and skills (age adjusted anyway) building on stupid ideas, but that are eminently employable.
Obviously there are other factors, I’m not really trying to defend anything but just point out that there are legit reasons why someone impressive enough to get into YC would also be impressive enough to get a good job. It’s not like it’s random founders off the street.
This is what always stood out to me. Founders talk about the risks they take as their major legitimizing force. But what risk? They’re the types who can (mostly) skip an interview and go direct into a six figure job - because of the sales and founder networks they develop.
Startup founders with no risk doesnt sound like a recipe for great companies.
YC founders aren't the ones taking risks. The risk comes from pouring your life savings into a company and not getting into any incubators and never getting any VC funding and then having no savings and nothing to show for it.
There is an opportunity cost in working on a startup that may fail and leave you with nothing while your uni peers have been cashing juicy RSUs in FAANG-type companies so the actual "risk" could be "having 0 net worth while living in a high CoL area by the time they reach middle age".
I'm on the list. I had real professional experience with HPC infra for comp bio, but that wasn't my role at Anthropic. There's a lot more to do at the company than manage infrastructure.
Tom Brown, an Anthropic co-founder who was also a YC founder, does lead compute. There were many years between his YC company and Anthropic, including time at OpenAI.
Others on the list went straight from their startup to OpenAI/Anthropic but to a team that fit them well. For example, Stefan went to a team at OpenAI where he works with startups, Lenny went to work on product (early ChatGPT), and Raza's YC startup was evals/observability for LLMs.
Maybe they're sales hires. Hiring influential folks gets them aligned with your business. It means you get their Roladex and tap into their network. They can poach good people. They can convince folks to buy into Claude. If they were able to fund Uber for dogs they could probably pitch AI just fine.
How many times have you heard a C-level person say "I was just meeting with a bunch of my C-level friends and they're all talking about how AI is the future. We should look into this Claude thing"?
They are hiring HPC experts too. It’s a small community and you just don’t hear about it.
These 100 people repress a tiny fraction of the people who work at these labs. And many of the people that work at the labs are building on top of the models, not building the models themselves.
The top % of YC founders are legitimately quite talented, and high agency/entrepreneurial talent goes a long way in these companies, even if its not in deep infra or research
The actual AI part of AI startups are at constant risk of commoditization. Where the companies will survive or fail is how they manage to attract customers and lock them into their offerings so they can't move away to the next competitor.
Someone with experience running a startup that has no moat probably does have some relevant experience in that area.
I actually have experience with that stuff, “old school” deep learning and ML as well. Is that something worth joining the fray with I wonder? Or is it as you point out really only the richly connected “locals” that are recruited?
Not to try to grab any pity, but I have pretty deep experience in the GPU optimization space and have not been able to even get an engineering screen even with referrals.
I have a sinking feeling that AI models have gotten so good internally, and/or the GPU guys in their labs have found a good enough processes, that AI is doing the research behind the doors right now (it should be at least capable, at heuristic optimization problems that are benchmarkable), while the hard part of their business is the political game: convincing other companies to buy into AI, integrations, trying to gather more data where they can.
> The question I have is why are these companies hiring these people and what does it say about their hiring practices and the amount of capital they’re poorly spending?
> I’m not surprised that these companies are well funded and hiring a lot of people. I’m surprised that they chose to hire the people who were previously making “Uber but for dogs” gimmick apps and not just hollowing out the HPC specialists from national labs.
Culture. The Uber for Dogs idiots will gulp down the kool-aid like they’re dying in the dessert.
I mean, what are you suggesting “someone” should do about it? The companies can spend their money however they want, and if they want to hire a bunch of product people from YC startups nobody can stop them.
I’m just surprised they need this many of them. What do they do all day? And who is left to manage their hundreds of thousands of servers?
A lot of the AI game is still integrations and getting entrenched in processes. It has the double whammy that you get to train on their data (not talking about ALL enterprise contracts - but cursor pretty explicitly has this business model. So I'd imagine some enterprise contracts have this.)
I’m just surprised that the people actually keeping the business operating and the tokens flowing don’t resent these product people for sitting around all day coming up with ideas for the same TC package as the people putting out ops fires and bringing new capacity online all day every day. I feel like the actual engineers are probably getting shortchanged.
Things from YC never really were independent companies, I think it's more clear to view all of YC as one company. The separate entity thing is mostly a trick so they can book revenue to themselves as real revenue.
Look, we have 37 other YC companies paying for our product it's so popular and successful. Hits different if you view it as a single company, imagine AT&T promoting their internal customer service software, used by 37 divisions of AT&T it must be good!
Effectively, industry roles are a tiered system which determines access to the cashflow. Going the YC founder route is a much faster and more efficient way to secure a high tier than climbing the SWE ladder. The whole thing resembles a kind of mini class system, with high tiers granting access to generational wealth, and low tiers being better off than the average non-industry Joe. Once you're in the high tier, there's no way to fall anymore, you just move wherever the cash is being pumped in, collecting it.
I’ve seen many people climb the SWE ladder or build a YC company (having done both myself), and trust me the founder route is not the most efficient. You only believe this because you didn’t see the 2 failed startups and 10 year journey grinding 80h weeks almost running out of cash a few times.
Not that I disagree, but it doesn't refute the original argument. It could even support it, class protecting their own from more people figuring out there's no clothes on the emperor.
Nobody in their right mind should start a company up to some equilibrium. As a society we should fund the appetite of experimental founders to put more pressure on quality than what natural conditions might allow.
This. I saw way more people build a few millions of retirement nest egg reliably in 10-20 years, by working on a boring good swe job and manage their investment a bit better. It's a much safer route.
What part of "start a company" requires "grinding 80h weeks"?
Sounds to me as inability to delegate. Or implicit in the [acquire VC, aim for high growth] mindset as if that were the only path to success.
Companies can grow organically. With a positive cash-flow early on & reasonable work/life balance for founders & employees alike. Heck, there's even non-profits.
Yes they can, but unless you strike luck, they take many years, often decades of work to grow to a level where you can retire and enjoy the money.
With FIRE and similar concepts being pushed around everywhere, and the plight of the masses to keep their jobs and homes, I see where that desire for "make money fast" comes from. And from a game theory mindset - it's not wrong. (That is also a bit related to the explosion of gambling and gambling-alike stuff such as "prediction markets"...)
> Heck, there's even non-profits.
Non-profits often enough run on conditions bordering on (self-)exploitation.
> Going the YC founder route is a much faster and more efficient way to secure a high tier than climbing the SWE ladder.
YC has funded over 5,000 companies. If you assume 2-3 co-founders per company, that's more than 10,000 to 15,000 people. The vast majority of these founders aren't producing "generational wealth" outcomes. There's no glamorization of the companies that shut down, the ones that are scraping by, and the ones that get their founders a normal job, but those are the far more likely outcomes, especially in the more recent spray-and-pray batches.
This is not even close to true. I work at a big tech company and we reject applications from YC founders every week. Most of them are broke with failing companies and would love a cushy mid-level engineering or management job at a faang. People are biased by the success of the top few founders but 99% of them have worse outcomes than the average corporate engineer.
This is a very hot take. Being a founder might confer you some career advantage [1] if you're one of the few to make it to series A and you grow a large team, but if you're like the vast majority of founders and never make it past seed (or realistically, to seed), your network is the alpha and omega of any subsequent job search. It's a bit like saying that being an NBA starter is a good way to get a job as a basketball coach.
Point being: don't start a startup if your goal is to get a job. Just get a job.
[1] Note that I'm not arguing about experience -- you can gain a lot of experience as a startup founder, but that experience is rarely directly marketable. Also, most startup founders are completely clueless when they start, so "a lot of experience" is a relative term.
It speaks to me of these fascinating Jungian shadows, where a thing/person has a front claiming to do their thing and in some non-obvious way they facilitate the opposite.
YC mints companies, founders, it disrupts.
Its shadow feeds your disruptive moves into SV's establishment tier, and selectively shuts down efforts, and moves people into its favored buckets of enterprise.
A successful unicorn founder once said a billion dollar company is just made up of a bunch of people who could each have started a $10m company by themselves.
I guess you can extrapolate that and say a trillion dollar company is made of a bunch of people who could have started a billion dollar company.
The site infers that 105 people are working at the two companies. What’s the denominator though?
After 20 or so years of YC, with multiple catches per year, and the insanely high failure rate of start ups, there must be a lot of former or floundering founders.
IMO we are in the era of “Final Companies” and the appeal of being a founder these days just isn’t great compared to the entrepreneurial boom of the 2010s and 2000s. The risk to reward ratio has never been worse.
If Product-Market fit for small companies take 3-4 years and AI can streamroll you by coming from behind, it makes sense to join the big boys rather than trying to finetune your creation in the hope of making something big.
Yup. Capitalism is a dark forest now. If you try to do anything really interesting, big players will take you out fast and take over the market.
Any business that relies on loudly marketing itself to wide audiences is at risk.
The only viable business path now is niche domains with limited visibility, B2B sales. No social media presence, no website, just quiet business deals being done in exchange for goods or services that is all under the radar of megacorps. In tech this is now virtually impossible as all useful services that could be provided have already been provided in some form by now.
There are no low or mid hanging fruit, everything is now very high up.
The sepia-tinted color scheme and the way LLMs overuse rounded corner tiles to the absolute death is part of it. Really what it comes down to, is that these LLMs have a design vocabulary of like 15 traits. None of the websites they generate have every single trait, but they are composed of some subset of 10 such traits, so it's still overwhelmingly obvious the moment you open the page, even though each one is slightly different. Humans might use 1~5 of those traits, but they'll also be mixed in with a much wider set of traits with more individuality to them.
- Orange/beige-ish colors
- Rounded corners
- Cards with a thin border
- A thicker colored border on the left of cards
- Serif font for headings
- Monospace fonts for small text
- Headings that often have an unnecessary subheading / pre-heading
- Little badges, often with a "status indicator" dot on the left
- Obviously LLM-generated text / language
I'm sure I'm missing many but the above are dead giveaways!
The copy is usually a bit of a giveaway too, ”explore your future path to greatness and experience the defining divider between those who can and those who can’t” or as most humans would write the header ”plans”.
Most AI-generated sites have coopted the Inter font, and usually they go for a dark colour scheme, the rounded tiles, a container representing a macOS-styled terminal with code, and, the thing that annoys me the most, divs that fade is as you scroll.
1) This Sankey chart is missing some layers, and as a result it's kinda misleading. Some examples:
- I'm no longer at Anthropic, and many others are no longer at OpenAI.
- Tom Brown didn't jump straight from a group dating app to co-founding Anthropic; he spent several years at DeepMind and OpenAI.
- Brian Krausz worked several places in the ~decade between his YC startup and Anthropic. He then left Anthropic but was brought back in with the Stainless acquisition.
- Stefan Stokic ran a non-YC startup for five years before joining OpenAI.
2) I don't think 35 people joining Anthropic or 70 joining OpenAI at some point after doing YC is a substantial concentration at a macro level. That's <1% of YC founders, even if you include those who no longer work at either of the companies.
3) As for why people joined: beyond the compensation, it isn't clear that a startup is more impactful than powerful AI in their domain. Here's an excerpt from an email I sent to one of the Anthropic co-founders when I was considering joining in 2024:
> For context, my YC company builds dev tooling for bioinformatics. [...] I no longer have the conviction that our niche-specific dev tooling is going to have a significant impact on the field as we approach expert-level AI.
Ultimately, the core value of our product was actually supplanted by LLMs; bioinformaticians are much better at writing code and managing infrastructure with LLMs.
I think more and more folks are seeing the impact of what they'd build as a subset of what powerful AI could do, and as a result they end up joining a lab.
105/13000 is a very small number to focus on. This data doesn't really mean anything.
The interesting thing to me is that the belief that Anthropic is going achieve something like AGI appears to have really spread through SV. If not AGI, at least greatly disturb their old cushy SaaS jobs.
Also, I still do believe some people just want to work on "cool" stuff.
edit: not just CTOs and CPOs, I mean John freakin Jumper, who got the 2024 Nobel Prize in Chemistry, left his VP role at DeepMind to have no reports, and be an IC at Anthropic.
I find this whole trend extremely interesting.
One thing I’ve noticed more and more is all the “cool” jobs require you to already be an established domain expert.
Another things I’ve noticed, and this is probably mostly just my opinions evolving, but all the “cool” stuff that catches my interest these days is not in software. Software is boring.
Most recently, I became interested in the field of photonics/optics, particularly in the field of novel optical communications research.
I’ve had mild interests in things like Biology and History since I was a child. The latter particularly hooked me a few years back.
Neuroscience is a pretty cool one as well. The human brain is of course one of the most fascinating pieces of hardware in existence and understanding it (to whatever level we currently can) is incredibly interesting to me.
At one point I thought about going into security research, but now that everyone and their mother is headed that direction, it’s become less feasible.
Unfortunately I don’t even have a bachelors degree, and most of the fields are research heavy and not very accessible to hobbyists, so my actual interaction with these things remains relegated to reading papers/pop sci articles/etc.
I mean, this is happening right now.
I doubt there are few people talking to customers and potential customers that haven't run into the "look what I rolled last week with Claude" only for it to be legitimately impressive. The idea of technical moats is evaporating in front of our eyes.
So wether it's belief in AGI or a desire for building cool stuff, as an outsider looking in, it seems to be informed by the same echo chamber of thought.
The IC pay at OpenAI alone is comparable to that of an EM at Google. CTOs at most large orgs end up earning the same as an EM at Google.
OpenAI and Anthropic pay is extremely competitive. The last time I saw something similar in SV was Google back in the 2000s.
While I certainly have no hand in steering the ship, agentic building brought me out of hibernation/mild depression. I am neither young, nor rich. However, I still had lots of product ideas, and now I am able to create and test them in days. And yes, I am actually making sales now.
I got into this early at Sonnet 3.5 release time, so I went through initial AI psychosis almost two years ago, but still today, I am more excited to wake up and work than I’ve ever been in my life.
(1) (digital) AGI is a real concept, and will be achieved imminently
(2) AGI will be world (and possibly more) changing.
(3) Anthropic will be the one to do that, or at least capture a large part of it
If you buy into all of these, which I'm assuming a lot of SV does at this point, then its kind of a no brainer to go to OpenAI or Anthropic. And optics wise it seems like Anthropic is "winning" right now, so thats the lab with the talent flow.
If Anthropic actually creates a singleton, contributing to that is way more impactful than being CTO at some random company (though I may disagree that said singleton would treat equity holders any differently).
So I could imagine tons of people in those roles who would love to be ICs again if there was also a chance at outsized wealth.
Having a major leadership role at a big company is kind of a slog. Getting a chance to go back to a fast growing startup as an IC in a way that will only benefit their career (ex-Anthropic or ex-OpenAI makes your resume gold plated) is an easy choice. They can go back to any job they want after this.
I think this might only be shocking to some people who have forgotten that many people in these positions actually enjoy working on hard things and being surrounded by other like minded people. If you’ve only worked in po-dunk startups or other companies where leadership looks like a Game of Thrones style fiefdom building game it’s harder to imagine executives who are smart, capable, and enjoy working hard. It’s pretty common in Silicon Valley despite all of the sneering here.
I'm not saying it's toxic to have it on your work history, but it's far from gold plated. They'd have work to do for me, and for many, many people in the industry, to prove that I'd want to work with them.
When I say their resumes are gold-plated I mean at the companies they want to work at.
For all the hubris and arrogance of the tech scene, and especially the Bay Area's, founders at least pretended to pursue a greater calling than money by leaving cushy gigs to pursue a new venture they were passionate about. Sure, that only lasted a handful of years before it became a gold rush to move to the Bay for YC and pursue the startup-to-acquisition pipeline, but they at least pretended to care.
The naked pursuit of wealth at the cost of potentially tearing apart society is mildly alarming because the scale it's happening at is so big.
In my opinion there is no "slipping mask", merely a changing of the guard that always happens when massive piles of generational wealth are at stake.
SV has always operated on exploitation of their workers, they just hid it well from the public.
Perhaps you've seen under their mask, and you know they're scamming the investors. After the NFT bubble it's a plausible hypothesis. But I don't know what it means to say that the mask just isn't there and it's naked pursuit of wealth.
Some caveats here: my view of social media is biased by the fact that jerks tend to pull me out of the woodwork- I get ragebaited, although I think I avoid the literal bait- so many of my interactions are with jerks. Looking around, I expect I'm not alone, but this may not be the default experience. And the LLMs certainly aren't perfect. They can hallucinate, they're often sycophantic, they're bad at judgment in a number of ways, probably others I'm missing.
Compared to actual people, they suck. Compared to social media as it exists today, they have orders of magnitude to fall before your comparison makes sense to me. And, at least for now, the LLMs are the ones getting better while the social networks are- AFAICT, universally- getting worse.
Social media felt this way in the early days, too.
> they almost never feed me misinformation (and never deliberately)
This is dangerously untrue. LLMs will deliberately tell you the things you want to hear. They even "call you out" in exactly the ways that are most palatable to you. There is ample evidence that LLMs can be biased. The ability to broadly influence individual opinions, without even needing bot accounts on social media is exactly one of the reasons so many people who want power/money are flocking to control these LLMs.
> so many of my interactions are with jerks
This is one of the ways that the LLMs are "scarier" and more powerful than social networks. Real people are more likely to be obviously pushing agendas or otherwised easily dismissed.
> LLMs are the ones getting better
Getting better at what? The things you list are exactly the things that make LLMS subtly powerful and far more dangerous that social media. It's the reach and anonymity of social media without needing to control a feed algorithm, deal with content or people that doesn't fit your agenda, or produce fake content that does.
This is kind of making the parent's argument for them about the dangers...
> And the LLMs never try to make me feel angry or stupid, they never ignore my arguments to go on their own soapbox, they almost never feed me misinformation (and never deliberately), and they've yet to tell me to my face that the purpose of their interactions with other people is not to convince or even communicate with those people but to perform for an audience (often one that doesn't feasibly exist).
You're basically asking to be coddled, affirmed, validated, and protected. The world has conflict. Not all conflict is bad. Yes, social media weaponizes that conflict for ads, but I think the opposite is more dangerous in creating completely distorted realities. What you're describing sounds worse than the worst nightmare I've ever had, a world in which I'd be constantly disconnected from everyone else and would be disillusioned into thinking I'm not.
> Some caveats here: my view of social media is biased by the fact that jerks tend to pull me out of the woodwork- I get ragebaited, although I think I avoid the literal bait- so many of my interactions are with jerks.
This is what therapy is for.
>You're basically asking to be coddled, affirmed, validated, and protected.
Incorrect. My interactions with LLMs simply don't look like this. I don't know where you got this idea from what I said. I have a prompt that begs the thing to call me out, and it currently does so too much, which is exactly what I prefer.
What it doesn't do is rub minor disagreements in my face, jump to conclusions about who I am based on pre-existing beliefs, or tell me to go to therapy when I earnestly admit my humanity.
Now its AI.
Maybe AI doesn't materialize the way we want, but I'll take that bet over business as usual before every time.
That's never been true for financial services, many of which offer very high-paying jobs for high-frequency trading, hedge fund management, market operations, etc. There are many other industries with high value IT jobs that are not ad-supported.
(Edit: typo)
Honest question: where should these very smart people work then in your opinion? Nearly all suitable-looking, "smarter" ideas that come to my mind will bring less revenue/profit in expectation than, for example, letting them maximize ad impressions.
I of course wish we lived in a "more intelligent" world where this is not the case. :-(
Call me crazy, but the smartest people I know have rejected maximizing their employer's revenue and profit as their goal. Or even their own company if they go that route. Find a niche, work it, enjoy it, and enjoy the rest of your life, too.
For example, Elinor Ostrom won a Nobel Prize for her work illuminating design principles illustrated by long enduring common pool resource institutions.
https://en.wikipedia.org/wiki/Elinor_Ostrom#%22Design_princi...
People have very different, often contradictory opinions what the "common good" is.
I would even claim that this question is a special case of the millennia-old philosophical problem of defining what "good" actually means (i.e. defining ethics), which has resulted in myriads of different opinions:
> https://en.wikipedia.org/wiki/Outline_of_ethics
Anywhere else. Ads are a net negative to society and provide no value. ANY activity outside of pushing ads is a better contribution to society as a whole.
Seems like an even bigger opportunity to inject ads, influence thought, and scrape user information.
The people are there because they are betting on AGI, not chatbots that make software engineering go faster.
Whereas ads are almost exclusively negative.
Edit: all that and spellcheck still sucks too.
https://www.nytimes.com/2026/07/10/us/politics/ai-terrorism-...
There's never been an information access problem with nuclear weapons. The information was so public that most physicist realized everyone was building bombs from the sudden lack of published papers on how you could build bombs.
It's just silly how people seem to believe that if you just have a page worth of text saying something, you can suddenly do that thing.
Even a magical superintelligent AI could not do these things. It's going to tell you the exact same thing that has been known for nearly a century now:
"Hire 100 engineers and a couple physicists and 1000 very skilled machinists and laborers, and spend a billion dollars on plumbing, and then do that a second time if you need to deliver the device using a missile"
Meanwhile a dirty bomb is trivial.
These things don't happen because people who want to cause mass harm are generally poorly resourced.
During the AI boom: Attention is all you need
Additionally, it's pretty clear that the owners of this tech (dario, altman, etc..) do not have the best interest of "the people" at heart and only care about accumulating as much power and wealth possible. Not that it's their responsibility to save humanity or something, but I'm not sure this future is much better...
When there's nothing good being invested in, "the smartest people" are simply underutilized. They do things that they think are good instead.
"The smartest people" have never sought maximum capital accumulation, and have not historically been driven by "highest wages" jobs.
The computer industry formed out of WW2 because a bunch of smart people ran into each other and did what they thought was really cool and had realized was suddenly possible. Von Neuman literally met one of the other guys and discussed computers as a theory at a train station and that meeting was influential on the future industry. They did not try to make billions of dollars. They had fairly average jobs that paid average. Often working for the government itself.
"The smartest people" did not work on targeting ads at people, and while some of them may have been involved with early transformer work, they certainly would be disillusioned with LLMs by now and would not be working at companies burning trillions on them.
This both destroys discourse, and centralizes the means of production into ~4 companies, with everyone else paying them an economic tax in order to be competitive.
Facebook Marketplace also provides me with quite a bit utility.
It probably feels this way if you only see headlines like this, but the LLM companies and even extending into their data center buildouts are still only a part of the economy. Investment dollars get spread around far and wide. Investors can’t even get allocations in these companies if they want to because they’re so oversubscribed.
This headline and website are a good example. If you think being a YC founder is rare and reserved for the truly elite, you may not have seen the size of YC batches lately. The number of YC founders is in the tens of thousands. Showing a list of 100 of them who joined a couple big tech companies doesn’t mean anything by itself.
It’s on the order of 1% of YC founders. YC startups don’t work out all the time and their employees and founders go work for other startups. Barely worth noting, if not for a website making it look like a big deal.
It's worth pointing out that "entire economy is betting everything on [AI]" you're talking about here is in the economy in the "numbers go up sense" (e.g. the stock market). It's not the economy in the providing for the general well being sense (e.g. gives people jobs so they can get housing and food).
I wasted like a half hour trying to find decent sources but I guess I don't know how to search the internet anymore.
I think we're betting more than the stock market on AI, even if I can't say for sure how much.
The real economy didn't change at all, but confidence was gone and so was lending.
The 90th percentile upside is way, way higher than is priced into the current stocks, the 10th percentile looks to me like 2000 where a bunch of investors loose their shirts and then for the next few decades new companies boom on all the infrastructure that was built.
The “failure mode” for this tech is that we don’t advance much beyond Sol/Fable and the existing models just get cheaper. If there is an AI bust in hindsight it likely will look like the dotcom bust did, which is a small blip on the decades-scale GDP curve.
Otherwise, it’s sort of like saying that people will find a job doing something else, without any thought to what that something else would be. The jobs don’t create themselves.
AI is the one and only possibility for saving millions (likely billions) of lives from “old age”, cancers, Alzheimer’s, ALS… you name it.
It’s a big bet. The alternative is we all die anyway: sadly after a gradual decline at best or horribly after long undignified suffering at worst.
I’ll take the economic risk.
Says you and everyone else with a financial incentive to trick people into believing this.
We've had text to speech for a while for example but everything is now AI. Does it use any AI features? Not at all ...
the end goal isnt to serve a chatbot to replace google or a coding agent to build crud
This is untrue. Spend is about 2% of GDP.
That's not going to happen. Everything is different now, and the only long-term losers will be those who continue to deny it. Even now we still see stupid analogies to parrots, calculators, toddlers, and so on, being advanced here of all places.
and would end up with a market crash
That might happen, because the first attempt a market makes to allocate funding and spending optimally is almost never on target. Great, it'll give the rest of us a chance to buy in.
Combine that with the fact that “good enough” is a real thing and cheap Chinese models are either there or close depending on your use case and it’s hard to see how this ends well.
I’m just incredibly confused how these companies and investors don’t see the writing on the wall UNLESS the point is to just extract as much money as possible and not be left holding the bag.
Maybe there’s a partial success future that isn’t those two, but I’m not hopeful. And I’d take the Ai bubble collapse over AI BigTech achieving its goals
Sam Altman was the president of YC, so it's not particularly suprising that he has hired lots of talent from YC - these are people that are pre-qualified because he has seen their work before outside of a job interview (better the devil you know than the devil you don't).
This happens everywhere - person leaves company A to company B and then poaches the best talent from company A. Obviously in this case the talent are founders rather than staff members, but still.
It’s exceedingly unlikely that any of the people who were working on YC startups previously have any real professional experience with any of the following: slurm, collectives, NUMA systems, RDMA, compilers, systems programming, general HPC performance estimation or measurement, CUDA or ROCM or any kind of GPGPU/accelerated computing. But that is the core business of both of these companies.
I’m not surprised that these companies are well funded and hiring a lot of people. I’m surprised that they chose to hire the people who were previously making “Uber but for dogs” gimmick apps and not just hollowing out the HPC specialists from national labs.
They invest in people more than ideas, so you’ve got, at least in many cases, people with good pedigree and skills (age adjusted anyway) building on stupid ideas, but that are eminently employable.
Obviously there are other factors, I’m not really trying to defend anything but just point out that there are legit reasons why someone impressive enough to get into YC would also be impressive enough to get a good job. It’s not like it’s random founders off the street.
This is what always stood out to me. Founders talk about the risks they take as their major legitimizing force. But what risk? They’re the types who can (mostly) skip an interview and go direct into a six figure job - because of the sales and founder networks they develop.
Startup founders with no risk doesnt sound like a recipe for great companies.
Tom Brown, an Anthropic co-founder who was also a YC founder, does lead compute. There were many years between his YC company and Anthropic, including time at OpenAI.
Others on the list went straight from their startup to OpenAI/Anthropic but to a team that fit them well. For example, Stefan went to a team at OpenAI where he works with startups, Lenny went to work on product (early ChatGPT), and Raza's YC startup was evals/observability for LLMs.
How many times have you heard a C-level person say "I was just meeting with a bunch of my C-level friends and they're all talking about how AI is the future. We should look into this Claude thing"?
These 100 people repress a tiny fraction of the people who work at these labs. And many of the people that work at the labs are building on top of the models, not building the models themselves.
Someone with experience running a startup that has no moat probably does have some relevant experience in that area.
I figure most of these are working on products, rather that deep frontier research.
They are hiring for sales.
Culture. The Uber for Dogs idiots will gulp down the kool-aid like they’re dying in the dessert.
Is this causing massive problems for us as a society? Of course, but no one seems to want to do anything about it, so here we are.
I’m just surprised they need this many of them. What do they do all day? And who is left to manage their hundreds of thousands of servers?
AI for dogs.
Breeding product ideas? ChatGPT, but for dogs?
> And who is left to manage their hundreds of thousands of servers?
They learned from Elon how to get by with a skeleton crew?
Same as they were doing in their startups: burn mon..., eh, tokens.
> And who is left to manage their hundreds of thousands of servers?
AI, of course.
Why spend to resources to invent AGI when you can just gaslight the whole world by telling everyone that LLMs are already better than people?
Look, we have 37 other YC companies paying for our product it's so popular and successful. Hits different if you view it as a single company, imagine AT&T promoting their internal customer service software, used by 37 divisions of AT&T it must be good!
SELECT * FROM yc_founders WHERE employer IN ('OpenAI', 'Anthropic');
If there are 7k founders [1], the graph only shows 1.5% of the people from YC.
[1]: https://www.ycombinator.com/investors
Jensen Huang himself says nobody in their right mind should start a company https://www.cnbc.com/amp/2024/05/11/jensen-huang-i-didnt-kno...
Not that I disagree, but it doesn't refute the original argument. It could even support it, class protecting their own from more people figuring out there's no clothes on the emperor.
Sounds to me as inability to delegate. Or implicit in the [acquire VC, aim for high growth] mindset as if that were the only path to success.
Companies can grow organically. With a positive cash-flow early on & reasonable work/life balance for founders & employees alike. Heck, there's even non-profits.
Yes they can, but unless you strike luck, they take many years, often decades of work to grow to a level where you can retire and enjoy the money.
With FIRE and similar concepts being pushed around everywhere, and the plight of the masses to keep their jobs and homes, I see where that desire for "make money fast" comes from. And from a game theory mindset - it's not wrong. (That is also a bit related to the explosion of gambling and gambling-alike stuff such as "prediction markets"...)
> Heck, there's even non-profits.
Non-profits often enough run on conditions bordering on (self-)exploitation.
YC has funded over 5,000 companies. If you assume 2-3 co-founders per company, that's more than 10,000 to 15,000 people. The vast majority of these founders aren't producing "generational wealth" outcomes. There's no glamorization of the companies that shut down, the ones that are scraping by, and the ones that get their founders a normal job, but those are the far more likely outcomes, especially in the more recent spray-and-pray batches.
Point being: don't start a startup if your goal is to get a job. Just get a job.
[1] Note that I'm not arguing about experience -- you can gain a lot of experience as a startup founder, but that experience is rarely directly marketable. Also, most startup founders are completely clueless when they start, so "a lot of experience" is a relative term.
YC mints companies, founders, it disrupts.
Its shadow feeds your disruptive moves into SV's establishment tier, and selectively shuts down efforts, and moves people into its favored buckets of enterprise.
Edit : and money. I guess most founders follow the money here.
Would be curious what some of the VP+ people are building inside Anthropic if they joined as engineers
I guess you can extrapolate that and say a trillion dollar company is made of a bunch of people who could have started a billion dollar company.
After 20 or so years of YC, with multiple catches per year, and the insanely high failure rate of start ups, there must be a lot of former or floundering founders.
Any business that relies on loudly marketing itself to wide audiences is at risk.
The only viable business path now is niche domains with limited visibility, B2B sales. No social media presence, no website, just quiet business deals being done in exchange for goods or services that is all under the radar of megacorps. In tech this is now virtually impossible as all useful services that could be provided have already been provided in some form by now.
There are no low or mid hanging fruit, everything is now very high up.
Where are YC founders now who went to openai and anthropic ? OpenAI and Anthropic, mostly
I still can’t put a finger on it. I’ve seen real people use these fonts and layouts yet theirs look original.
Whoever finds an explanation for this solves AGI (/s)
So you're saying the site screams "Claude's style!" to you. Not too different from "I know a WordPress site when I see one".
- Orange/beige-ish colors - Rounded corners - Cards with a thin border - A thicker colored border on the left of cards - Serif font for headings - Monospace fonts for small text - Headings that often have an unnecessary subheading / pre-heading - Little badges, often with a "status indicator" dot on the left - Obviously LLM-generated text / language
I'm sure I'm missing many but the above are dead giveaways!
The copy is usually a bit of a giveaway too, ”explore your future path to greatness and experience the defining divider between those who can and those who can’t” or as most humans would write the header ”plans”.