AI is solving phD level open math problems, even though I was told it was just a stochastic parrot that would never generate new ideas or solve unfamiliar problems.

Self driving cars roll past my office every couple minutes, even though I was told AI would never be able to drive on public streets.

AI images and videos are indistinguishable from real videos, even though I was told they'd never be able to generate hands correctly.

But I'm sure the naysayers are right this time, AI defintely won't ever be effective at medical research.