If you read The Lean Startup, you'll have the answer to your question. AI startups don't generally provide AI, they solve some problem using AI. Now, if the onl

If you read The Lean Startup, you'll have the answer to your question. AI startups don't generally provide AI, they solve some problem using AI. Now, if the only advantage your AI startup has is that your solution costs less at scale, then you're right, lean isn't going to work well for you. But if your solution is significantly better even at small scale than those available in the "saturated" market, then the lean approach is to get your solution out there in some kind of minimal form, to some small group of customers. You just need to prove that there are people who will pay a price for your product/service that you can imagine making money eventually, even if you have to do some things manually to start that you will automate in software later. The book gives examples of this. Once you've proven some level of market fit with a lean approach, you're much better positioned to get investment. Lean doesn't mean completely bootstrapped. The problem that lean seeks to avoid is burning through investor cash making mistakes in getting market fit that you should have made before investment. By the time you get investor money, you should be ready to make the next level of mistakes. :-)