IMHO this post misses the fact that YC becoming a prestige institution is itself a sort of failure mode. You don't want to attract founders who figure YC is a low-risk alternative to grad school that will look good on their resume.
It's tough to avoid that outcome while still conferring positive signal to VCs/potential employees, though.
I'm sure YC/Garry see something in the PearAI founders' ability to market themselves, but I find the whole debacle a bit embarrassing for YC and I know some of my YC batchmates quietly do as well.
For the uninitiated like myself, PearAi just took the source code of continue.dev (not fork, they copy pasted) and did some clunky work on it. That was their entry to YC.
for additional uninitiatedness: one of the founders is "Frying Pan", a popular youtuber. There has been previous discussion on the fact that the cost to build software is approaching 0. If that is given, maybe "taste" is all that matters. Funding a ~productless popular youtuber is a great way to test if "taste" and "brand" is better to invest in than tech in the years to come.
“Taste” is something that’s developed through repetitive exposure to differentiated items in a particular set, combined with extremely high abstract analytical abilities, and that’s something completely different from having marketing or personal branding skills.
I think you’re right that that’s the evaluation happening, but it’s totally misguided. If you’re indexing for differentiating levels of taste I would be very wary of empty vessel young influencers. Taste is built over years and years and imo requires a certain disdain for the crowd. Look at Linus Torvalds as a pinnacle of taste in code for example.
> There has been previous discussion on the fact that the cost to build software is approaching 0.
People have been claiming that since COBOL came out (actually, probably before; I bet some people claimed it about _assembler_), so, er, yeah, will believe that one when I see it.
"did some clunky work on it. That was their entry to YC"
For more context, YC doesn't judge your code, never has. It was never a code quality competition. Orthogonally, they do judge the results (user metrics).
Did they? YC regularly funded competitors. This was always true. Fundamentally, YC is betting on founders. They're optimizing for founders who can move fast enough to find a moat before they die, that's it.
They fund competitors because most YC companies usually pivot out anyway.
> You don't want to attract founders who figure YC is a low-risk alternative to grad school
Of course YC would want that (in the short- to mid-term).
The only thing YC has to do is produce a portfolio of companies that looks good enough that other VCs invest into that. This is completely disconnected to building viable businesses, as they just don't have to be the ones that are left holding the bag, and as an accelerator they are in the best position to do that.
The easiest way to fill that pipeline is to pair current hype XYZ with Harvard (or other ivy league) undergrads (or high-level ex-FANG people). As long as their ROI stays above a certain threshold, that's the main way to scale up YC.
> The only thing YC has to do is produce a portfolio of companies that looks good enough that other VCs invest into that. This is completely disconnected to building viable businesses, as they just don't have to be the ones that are left holding the bag, and as an accelerator they are in the best position to do that.
That's really short term thinking.
It might work for a class or two, but eventually VCs will realize that they're getting bad returns from their investments, and YC won't be nearly as attractive as it is today.
For long term success, YC needs to pick companies that will eventually become successful. Particularly the big, standout successes.
> The easiest way to fill that pipeline is to pair current hype XYZ with Harvard (or other ivy league) undergrads (or high-level ex-FANG people). As long as their ROI stays above a certain threshold, that's the main way to scale up YC.
If you think that's the path to good long-term ROI, I have a startup to sell you.
> but eventually VCs will realize that they're getting bad returns from their investments
I'm not saying that they are necessarily bad returns. It's just that for many reasons there is a strong opportunity for a disconnect between viable business models and seed-investments. E.g. exit event horizons are currently so long[0] that it becomes hard to correlate exit success to seed-funding (for better or worse).
> If you think that's the path to good long-term ROI, I have a startup to sell you.
Oh, I don't disagree with you. But from the actions of YCombinator it seem like either:
- They don't see this as a risk to their long-term ROI (due to some factors we are not seeing here)
- They don't have proper means of self-assessing their selection quality and think they are scaling well while they don't
- The situation is not as bad as the article and some of the comments here make it look like, and everything is fine with YC
Saw one recently that literally forked VSCode and Cursor and called it a company with some really shady practices. Not even sure what YC was thinking with that one, but it indeed falls into the category of comical vaporware.
How did something like this get funded? They must think there will be a follow through to liquidity event, but no clue how. Maybe YC is playing into the bigger fool theory that someone else will come along and pay more so YC can extricate their equity.
Sure, that would be a theoretical failure mode. But that's not really what's happening right now, is it?
YC doesn't look to have a problem of people joining just to get the stamp on the resume and then "half-assing" it after they get into YC. I think that's something that YC is still quite actively selecting against. As long as they are selecting companies that make it to a series ~C (which most founders will stick around for as long as they are on an good-enough upward-presenting) YC can (partially) liquidate at good enough fund performance.
A high-quality early stage team that self-selects out of follow-up rounds may be a decent outcome for some VCs. This means early liquidity in all of the "positive" events. If the founders were high quality, spinning an acquihire out can still recoup some of the loss.
The challenge would come where the founders are not serious, and instead are viewing YC as a stepping stone to a level up position in a big tech/large firm. While I'm sure everyone has this idea to some extent as a fallback, you need people to be committed to making their business work.
> The only thing YC has to do is produce a portfolio of companies that looks good enough that other VCs invest into that.
This is completely incorrect. They need liquidity events. Simply getting to follow on funding without ever making it to an exit is a negative outcome for YC.
While an exit (= aquisition, IPO and similar) is obviously always the optimal end-goal, every round of fundraising is a potential liquidity event for all existing stakeholders.
It's very common to have partial liquidation from roughly Series B-C onwards on the side of founders (e.g. wanting to keep up lifestyle with your C-level peers; removing personal financals as stress factor) and earlier investors (e.g. their funds entering the liquidation period of their lifecycle).
Just looked up the etymology of prestige and it’s interesting.
It comes from Latin praestigium ("delusion, illusion"), then 1500s French prestige meaning “deceit, imposture, illusion”. In the 1800s it started to mean “an illusion as to one's personal merit or importance, a flattering illusion”.
I would have wrongly guessed it originally meant “good reputation” (same as the article author meant it, I assume) and that the association with bullshit/fakery is just a modern twist from people using the word with cynical irony. But bullshit/fakery was in fact the core meaning.
Without being aware of the etymology I've still had a lifetime feeling since childhood that it is very fragile for some reason that is hard to pinpoint or bring into focus very easily.
Really is about the same feeling as when you know something is hype or BS and for that reason more subject to collapse like a house of cards.
A very straightforward way that this manifests: when a VC funds a developer tooling company, all their other portfolio companies are strongly encouraged to use it. Built in customers! The test is how much revenue you can actually bring in from outside the VC bubble.
The "something" that they see is that they have a 1% chance of success. YC is an investment strategy. They noticed that equity which is 99% sure to be worthless is heavily under priced and bought a ton of it. That bet paid off handsomely.
If your batchmates are seeing embarrassment from who else is at the top of such a funnel I don't think much of their judgment. Investors provide capital, not prestige.
YC was a child of its time though, right? Are you asking what could YC have done differently in the context of its history, or are you asking what a new accelerator started today would look like?
I ask because I’m not sure that now is the time for a new startup accelerator to succeed, and we have no way to predict the circumstances that are required for success without couching it in some major changes to externalities.
Great question; the latter, because as you mention, YC was a product of a moment in time and that moment has passed, but during that time horizon, they were very successful (imho).
Edit: YC says "Build something people want." and so I'm going to riff off of that in a bit of a meta way: "Support experiments worth conducting." The accelerator bit comes in once you've reached product market fit and need fuel for the rocket ship, but until then, you're just running an economic science experiment.
It's interesting to consider what the evolution YC 2.0 could be. I do think we're going to be seeing more innovative organizational models that can be facilitated by adept use of AI.
For example, a network organization that hires individuals and small teams. The organization works on various projects and product streams, which which can be spun out as new businesses.
This is a flexible model that allows for many different outcomes and journeys for the people. Less of a startup factory and more of an enterprise garden.
I like the idea of gateways as a system for managing ideas and new business developments. Regular gateways every 6 or 12 months that assess projects for continuation, funding and further development. People can be involved in several projects.
I see this type of organization structure as a kind of 'hyper network'. By using AI to monitor and report on network activity it should be possible to have effective management oversight, direction and communication ... beneficent controlled creative chaos.
completely data driven and pseudonymous. Everyone enters what they are building, traction, progress so far, team etc and internal team votes based on the data progress without looking at founder profile.
make it fair instead of funding the same golden-spoon cliquey gang over and over again.
YC pretty openly and deliberately tries to convince prospective and actual Big Tech employees who are curious or on the fence to quit and do a startup already. There's a great Startup School video about this [0]. I think the critics are right about "doing a startup" and chasing an exit having become an equally normal, precedented, socially-supported path as joining a big company and chasing promotions. But I'd be surprised if YC itself would characterize that as failure.
HN is a lot more jaded towards startups and founders’ games these days. Back in late 2019 there was this thread about a pre-YC Garry Tan video where the tone of the discussion was fiercely against working for startups, saying it was better to join FAANG or start your own company instead:
It's tough to avoid that outcome while still conferring positive signal to VCs/potential employees, though.
I'm sure YC/Garry see something in the PearAI founders' ability to market themselves, but I find the whole debacle a bit embarrassing for YC and I know some of my YC batchmates quietly do as well.