61–64 of 64 posts
Earlier quoted context omitted.
(If it helps anyone else, I'm completely dismissive of the value of crypto).
The best thing to come out of crypto hype is the attention to Markel trees. That probably played a part I. The development of CRDTs, which are straight up magic.
Any other founders feel like YC has lost its luster? Not only are the number of AI companies in the portfolio ridiculous, but the large class sizes seem so much less personal. Not to bag on anyone who’s in YC right now, but it seems to be going the way of many ivy leagues - still meaningful, but very much not the same proxy for extraordinary it once was.
YC's huge advantage is that they have an internal network of startups who support each other with YC money as customers, which then get pushed to VCs as revenue growth numbers. VCs have learnt to get in on the gravy train early, because the brand-name still carries value downstream. For no-name startups and no-name founders with no fancy universities (who are increasingly being ignored), this is a huge disadvantage.
If you want to look at YC's actual success rate, take a look at any of their public market companies' performance. Hint, none of them have outperformed their IPO price. In fact, many of them only found their ways into exits through the SPAC frenzy of 2021.
Earlier quoted context omitted.
> AI slop companies There is so much market up for grabs. Most companies will fail, but the ones that succeed can win really big. And the surface area is so much larger than attempting a non-AI company since the advantages are potentially huge. > ChatGPT wrappers. Create value. Doesn't matter how you do. Just create value.
I like your first point, hadn't thought of it that way; makes it seem just a bit more reasonable. Regarding "create value" though, that's where I think a lot of the negative sentiment comes in: the vast majority of these new AI companies offer nothing at all of value to anyone. Outside of a very small few even the ones with an attractive premise fail to live up to the experience they advertise.
ChatGPT isn't particularly useful.
A wrapper around ChatGPT which extracts specific entries from a PDF bill? That's incredibly useful.
The only niche so far that seems to have caught on for LLMs is code autocomplete, but that's already recording real revenues.
Great advice! The other thing I'd add is to get at least 4-5 YC alum to review your application in a Google doc and do practice interviews with them if you make it to the next round. Also, if you have friends who've done YC before, get them to recommend you!