The web and later smartphones have been the real startup breeders of the past 30 years. Biotech and AI require quite a bit of academic knowledge to innovate on and aren't on the same potential of big platforms like web/phones.
In this way I think it's very similar to the web and then mobile booms
Look at websites like CivitAI. Its literally just a front-end + a S3 bucket backend that allows people to share and download large stable-diffusion models (100MB-8GB).
The website founder was not an AI guy, he was a web-dev background and former founder. The website is already a huge success, forcing the incumbent (Huggingface, backed by Amazon) to try and hurriedly copy its features.
The tech and business mindset of 15 years ago is remarkably similar to that of 15 years before that, and 15 years before that, even.
Businesses need technical functionality that enables the core mission of the company, and they need accurate operational insights.
This means "boring" technology. It means reliable systems and processes. It means accurate reports who's methodology can be understood clearly.
I prefer to solve real problems with boring technology, than to solve imaginary problems with exciting technology.