btw unlike blockchain/cryptocurrencies that were also massively overhyped, LLMs actually have value & purpose.

I'm totally open to having ethical LLMs that are:
- tailored for specific purposes;
- trained with data that fits the said purpose;
- and used as a supplementary tool

ofc I don't see it coming until AI bubble bursts. but when it bursts, it's important to set that narrative so there won't be more tech bros who would recycle the current approach but serve in with a different pretext.