The frustrating thing is when you're debating people who firmly believe that generative AI "has no utility"... but also refuse to ever try it themselves.

(Which they might even justify because they've read the transformer paper or whatever. That doesn't help inform you if these things actually have practical applications!)