Ah, but, at least for generative AI, that kind of remains to be seen, surely? For every hyped thing that actually is The Future (TM), there are about ten hyped things which turn out to be Not The Future due to practical issues, cost, pointlessness once the novelty wears off, overpromising, etc. At this point, LLMs feel like they're heading more in that direction.
And 5 years ago, people used blockchain to operate a toaster. It remains to be seen the applications that are optimal for LLMs and the ones where it's being shoehorned into every conceivable place because "AI."
