It's a risky investment strategy to put so much emphasis on AI startups. Not only do you have the already volatile nature of early-stage software companies (which YC is of course used to), but this is a bet on whether machine learning, chiefly LLMs, are going to continue to outperform other technologies and become sustainable to run.

There's no question in my mind that 'Open'AI is subsidizing the vast majority of LLM research and use today. If the efficacy of LLaMA and its derivatives actually start to approach GPT4+ in any meaningful way, there is quickly going to be a shortage of suitable compute that will completely dwarf the now-subsided Bitcoin mining craze. Plus, untainted training data is going to be harder to find amongst the text contaminated with mountains of early LLM drivel.

As a technology expert, I couldn't in all honesty say that I would want to have money in the YC fund right now. But if it pays off, it could be the biggest software windfall since social networking took off at the beginning of the 2010s.

> We've now tried every point on the spectrum: fully remote, hybrid and fully in-person. So now we don't have to worry if we're being luddites: in-person YC just really is the best.

How was this determined to be best?

(Obviously, they haven't controlled for variables like the switch to 4 smaller batches, and the high percentage of startups all doing one exciting thing (AI). And do they realize the costs. And is it best for some people, and not for others.)

The way Twitter has changed in the last year is so I can no longer read threads since I don’t have an account.
Would have liked to read why in person is better.

I imagine the biggest reason is it raises the sense of commitment for everyone involved.

Was this resurrected? Algolia had this as posted 19 hours ago [0], and I swear I remember reading some of these comments last night.


I will never, and I repeat, never, return to working in an office. My quality of life since going fully remote is through the roof, and my employer is every bit as happy with my work now as they ever were.
I’m a huge fan of remote first companies, but when Jared speaks, I listen. I’d love to know more of any specific lessons learned or observations made that lead to this conclusion
YC is a one-trick-pony. If you have a simple product, it can guide you to get users, investors and to scale it up.

But don’t expect anything else. It would not help, if you are outside of that scenario. It easily can screw you, by using any product ideas that you’ve refined and sharing these freely with a relevant startup in the batch.

> It's a beautiful space with an incredible history - it's where the US built battleships for WWI and WWII.

Pretty sure this plays fast and loose with the history of American naval warfare. The only battleship built in San Francisco, that I can think of, is the 19th-century relic USS Ohio.

I was in the YC batch before this one (W23), which was the last remote one since started in COVID. The thread is spot on!
Well then, i suppose our overlords have decided we must be in person. But since they will “free” us with ai soon what’s the point?
This is off topic, but it’s crazy to me that the current incarnation of Twitter still seems to be the communication platform of choice for a good portion of the tech community. Presumably Jared still considers it to be the best way to share this information despite the fact that the app is so user hostile. Even using nitter to bypass these restrictions, we’re still breaking a paragraph into separate posts, as if sharing text longer than a sentence on the intent is a something we need to handle in a piecemeal fashion.

At a time when I can use ML algorithms to create beautiful images from a simple description, Twitter feels positively prehistoric. And yet the people investing in the bleeding edge continue using smoke signals to reach their audience. It’s remarkable.

Twitter is such a garbage way to convey this kind of information.
@dang, please change the root link to the nitter thread.
> Before covid, founders often asked us to run YC remotely so that they would't have to move to SF to participate. We never did…[we feel justified in our decision]

Completely ignoring the founders who still have to move, apparently?

You are being luddites.

COVID having forced you to be remote is not enough to validate this constant preaching from SV leadership (and only leadership - no one without millions in equity ever lobbies for this) that nothing can replace in person.

The fact that Paul Graham compares it to communism is fitting for so many reasons.

Wow, even the bleeding edge, move fast and break things, disruptive crowd has reaffirmed one tradition: "in-person is the best." I think this settles it. WFH is less productive.
I wondered what was the last time you could say that "half the YC batch is working on an X startup", for any value of X.

When did this pseudo math talk creep into normal speech?

It reeks of charlatans pretending to be technical when there is no need to be technical.