I don't know how AI regulation will work, but wouldn't it be feasible to restrict regulation to usage in specific domains? Like how you can regulate software us

I don't know how AI regulation will work, but wouldn't it be feasible to restrict regulation to usage in specific domains? Like how you can regulate software used in finance, health care and others but you can't just restrict "software" in the general case. You can't prevent me from making a good health care suite, it's just that I might not be able to operate it for profit and thus not expose the general public to it - which I find sensible. I would find it amazingly depressing if capable AI like GPT4(+) in general was blocked to the public, like it were heavy assault vehicles or nuclear devices. Just block specific instances of it. Just block WhateverShadyStartupDecidingIfYouGetAJob using GPT for example, I'm fine with that. If I can replicate their work at home, leave me be. Regarding this tech as if it were explosive devices is way too heavy handed for my taste.