Given that we have had PhD-level agentic reasoning AI models since last autumn I was expecting to see an explosion of new startups by now, but everything looks exactly the same. What could explain this?

Maybe someone could share some insights from the last YC batch or any other accelerator?