[zmgsabst]

You’re implicitly assuming only a global maximum will lead to useful AI.

There might be many local maxima that cross the useful AI or even AGI threshold.

[eru]

And we aren't even at a local maximum. There's still plenty of incremental upwards progress to be made.

[lolinder]

I never said anything about usefulness, and it's frustrating that every time I criticize AGI hype people move the goalposts and say "but it'll still be useful!"

I use GitHub Copilot every day. We already have useful "AI". That doesn't mean that the whole thing isn't super overhyped.