Request for Startups – Spring 2025(ycombinator.com)
ycombinator.com
Request for Startups – Spring 2025
https://www.ycombinator.com/rfs#spring-2025-introduction
2 comments
The most fascinating part of this is that YC isn't asking for any non-AI startups. That says a lot about the future of software development. This is the new way.
Compliance & audit turned into one important aspect of my life's work, and I wouldn't have stuck with it unless it was fertile territory for computerization. Which is the kind of thing where I had an early start in school, but didn't do much good when I first started my career since most offices didn't yet have any computerized office machines at all.
Then you should have seen the "belt-tightening of the '80's" :(
Wouldn't have made it through that without it being fertile enough ground without computerization to begin with. That became a niche I always favored, where there is opportunity without a computer, but eventually when introduced properly, puts everything on steroids.
There is a parallel to what I've seen from Manual > Computerized, being somewhat "analogous" with "Merely" Computerized > AI Enhanced.
Caldwell is the furthest outlier, if you boiled it down to "a deep focus on getting the most out of limited hardware resources were critical." And "this includes this kind of maniacal focus from the founders themselves."
I wonder why his "video has been removed by the uploader" now though.
You don't really have to specifically imply AI or even computers at all.
DeepSeek, Google, and Carmack are well-recognized examples where these type of fundamental characteristics can be leveraged in the "computer business", but there's a whole world of other businesses out there.
Beyond the layer where there's still a whole lot of further computer opportunity without including AI.
AI can be pretty "deep", it's possible when you're entrenched in it, it becomes less easy to see outside the trench than others ;)
Then you should have seen the "belt-tightening of the '80's" :(
Wouldn't have made it through that without it being fertile enough ground without computerization to begin with. That became a niche I always favored, where there is opportunity without a computer, but eventually when introduced properly, puts everything on steroids.
There is a parallel to what I've seen from Manual > Computerized, being somewhat "analogous" with "Merely" Computerized > AI Enhanced.
Caldwell is the furthest outlier, if you boiled it down to "a deep focus on getting the most out of limited hardware resources were critical." And "this includes this kind of maniacal focus from the founders themselves."
I wonder why his "video has been removed by the uploader" now though.
You don't really have to specifically imply AI or even computers at all.
DeepSeek, Google, and Carmack are well-recognized examples where these type of fundamental characteristics can be leveraged in the "computer business", but there's a whole world of other businesses out there.
Beyond the layer where there's still a whole lot of further computer opportunity without including AI.
AI can be pretty "deep", it's possible when you're entrenched in it, it becomes less easy to see outside the trench than others ;)
Again, like in the previous years, almost all startups yc want or found are ai-infrastructure-startups targeted at software developers/engineers. They clearly don't believe that products actually making use of AI outside of AI-development or stores itself have a future. Meanwhile Google, Anthropic and OpenAI have all turned from Foundation-Model to Applied-AI-Product-Centric-Companies itself, throwing shit at the wall and see what sticks. And just besides all that sad state, the profession of the software developer slowly dies out, being replaced by autonomous ai agents. A profession that succeeded making itself obsolete by its own genius. Congrats.