YCs advice on building an AI product

Suppose you’re a new founder and you want to do something on top of LLMs (like ChatGPT), how would you differentiate between an idea that could be a great foundation for a company and an idea that is likely to get become obsolete when GPT-5 comes out?

This question was asked at a round table with four Y Combinator partners last week. Y Combinator is a company that provides funding and advice to help start-up businesses grow (startups like AirBnB, Dropbox and Reddit).

This is what the team had to say:

“I think if a Founder is working on something too General and not solving a specific need for a user they can actually go talk to […] I worry about the ones that are too generic and going after some kind of abstract it-will-solve-all-the-things.

If it’s like, hey, throw your data in here and we’ll do automations on top of it for everything, that’s probably hard to compete with whatever one of the foundation models might offer; but if it’s like, hey, give us your sales log data and will like spit back suggested next actions for sales people to make them better at sales that’s probably going to work better.”

Here’s the whole conversation incase you want to get into it.