Hacker News Reader

Comment by paxys | Hacker News Reader

People look at the top 5 YC success stories and think every company they fund is of that standard. In reality they "graduate" 500-1000 startups every year. They aren't all winners. In fact I'd wager Pear AI is a lot closer to the norm in terms of quality and competency than, say, Stripe or Airbnb. If you look at their recent batches there is an endless parade of thin ChatGPT wrappers.

View on HN

> In reality they "graduate" 500-1000 startups every year.

I think people think otherwise because they used to be more selective, and haven't noticed just how much their volume has grown over the years. Early on it was a few dozen startups per year, then a hundred or so, and eventually the current state of greenlighting almost two startups every day on average (we're 275 days into the year and YC has racked up 509 companies in this years batches so far). They're less of a startup accelerator and more of a startup shotgun at this point.

Yea I sorta figured YC was more like Harvard in that the brand and selectivity are super important. Isn’t this eating the seed corn?

Harvard admits 2,000 students every year, so it isn't all that different. The reputation is always built/sustained by the outliers rather than the average graduate.

That's still selective compared to the volume of applicants though and the output isn't universally stellar but generally regarded as pretty high. Compared to YC's recent batches with a large number of ChatGPT and other gen AI model wrappers.

Harvard admits 2,000 undergraduates every year, but it's actually a very large school, granting just about 10,000 degrees per year. The annual commencement ceremony in Harvard Yard only goes through the graduates school-by-school; once dismissed all the graduates decamp to school-specific ceremonies all around campus, where they get their names called and walk across the stage to receive the handshake and (empty) envelope. Perhaps more like YC than you think?

The number of undergrad and graduate students in the US is vastly larger than the number of startups.

How many students apply and go to college a year? And how many startups try to get funding a year? These are vastly different numbers.

It’s game theory. AWS would do the same thing by pitting teams against each other internally. Often times, we would have multiple products doing the same exact thing but slightly differentiated. Of course AWS claims to not deprecate services but they would resource the successful service team and PIP everyone out of the unsuccessful team and bring it back to a skeleton crew. If you wonder why the AWS Product offerings are so F'd and inconsistent it's because of them using management techniques like this.

With how terrible AWS is to use this makes complete sense. I will stick with Azure when I can, which is usually.

I’ve never been able to stomach Azure (as much as I like Microsoft’s tools) because it lacks anything like AWS IAM. Without it having actual, pragmatic security let alone simple, cost effective audit documentation seems like so much more of a headache.

By this point and by that I've read, cloud computing preferences are as subjective as liking any other thing.

I've read here and in a lot of other places that AWS IAM is the worst part of AWS. But I don't know what to believe anymore.

This is true for every VC no? And the whole idea behind VCs? They aren't exactly only aiming to fund the startups they think 100% will be successful, as then they wouldn't fund anyone, so instead they spread out the risk to catch any surprise winners.

Does anyone really look at the line-up of funded startups from a VC and think they're all winners?

Yes but there's a difference between a VC investing a billion dollars in one startup and one investing $100K each in a thousand of them. In the first case they will obviously do a ton of due diligence, go over business plans, get board seats, look at code and more. YC on the other hand has a 10-minute chat with the founder and...that's it.

The earlier you invest, the larger the risk and looser the diligence.

> Yes but there's a difference between a VC investing a billion dollars in one startup and one investing $100K each in a thousand of them.

Isn't that just the difference between "seed funding" and "Series-A/B/C/$letter" funding? In the former, you want small amounts spread across as many parties as possible, while in the "Series" funding you do higher amounts but more concentrated, as you have more data to invest more in what you think will be the "winner".

> YC on the other hand has a 10-minute chat with the founder and...that's it.

I don't think that's true, but I've never been through the process myself. I know for a fact that the extensive written application is also part of it, that is reviewed by people before any interview even happens. I'm sure others who actually been through the process can add if anything is missing.

Which part of it do you think is not true? The written application is not extensive at all. It takes maybe 1-hour of focused time. That’s all I had ever spent on the written part of the application and I’ve been interviewed three times as a solo founder. And the interview really is just ten minutes long, though sometimes they go a little over the limit.

For sure- it's just that YC didn't used to operate like that. They have morphed from an interesting higher-touch incubator whose involvement was a strong positive signal into a scattershot VC, but not everyone realizes that so being "YC backed" still carries more prestige than is warranted.

YC does not (cl)aim to be every VC.

Most VCs try to avoid having the portcos actively compete with each other (ie they won't back 2 separate ride hailing apps) b/c they'll end up competing for the same pool of customers

It's meant to be a "smart gamble", not "throw money everywhere and see what does well".

Also the funding should come with a clause to cover this sort of behaviour, if they don't correct it now it will happen again.

Correct what exactly? Did they release this as their final product as the first milestone?

From what I gleaned the company has barely started and the founder recently(?) quit his job. They raised money on an idea and forked another project, changed the branding, and used it as the base to build a prototype

That doesn't mean this is the end product that YC invested in.

Lots of companies created MVPs this way before using funding and their new runway of time to do it properly.

If they do release it as the end product with little effort that’s basically fraud

> Also the funding should come with a clause to cover this sort of behaviour

Lol if you’re not aware, they came up with the gold standard in simple seed investment contracts used by nearly every pre-series-a startup in existence. Adding clauses like “don’t fork open source code” is just pointless and cumbersome legal bs that does nothing but get in the way.

Less "don't fork opensource code", more "don't bring shame upon the family".

This is bad press for YC as much as it is for this poorly thought-out startup.

There is quality/competency, and there is business ethics.

As an incubator, you own the practices of the companies in your portfolio.

It does not take a lot of rotten fruits to ruin the brand.

I just read this same comment from you on another post! Talk about cloning your comment in posts about cloning code.

> If you look at their recent batches there is an endless parade of thin ChatGPT wrappers.

Possibly related post from yesterday:

Y Combinator Traded Prestige for Growth

https://news.ycombinator.com/item?id=41697032

What does this do to society when people with no product, no use, and no hope of profit are consistently rewarded with free money? It practically incentivizes bullshit.

Well this just goes to the core of your view on the role of luck in life. Are there 1,000 startups coming out of YC every year and 5 of them are run by geniuses who single handedly disrupt loads of markets. Or are there 1,000 startups coming out of YC every year full of roughly equally good people 5 of which get extremely lucky and make boatloads of money.

Airbnb just forked hotels, Stripe just forked Visa.

> Airbnb just forked hotels, Stripe just forked Visa.

uhh this is not the same as TFA. this is a very quippy, pg-esque way of excusing the behavior though.

> Stripe just forked Visa

Clicking a button on GitHub is different from existing in the same industry.

Unless I am missing an Apache licensed code base that powers all of Visa…

Reply on news.ycombinator.com