I'm just reeling about how much is going on in that example (which I took outside my hotel): identification of objects, classification of abnormality, correct invocation and inference about gravity, correct diagnosis, correct prediction of consequences. All without supervised training to do this.

I agree with this:

"I also think it’s basically good when people are more exposed to probabilities and they become more normalized ....And while there’s some notion that the public doesn’t “get” probabilities ... I think that’s a significant oversimplification"

Somewhere between dozens and 100 names for the dish: en.wikipedia.org/wiki/Egg_in_...

Bullseye. Mid Atlantic USA

If they're using models that have been pre-tuned (aligned) for desirable properties then this is to be expected. Wonder if they took this into account?

Lots of athletes are playable in games and have written books. But I think you mean primarily authors.

It's a clear violation of the Mariah Carey Accord

If we extend the x axis enough we also have to worry about harm from wasting your life looking in the bathroom mirror. Two minutes per day is already 12 hours per year.

Benefits as a function of seconds in my mind

Wonder if it's a good proxy for patience.

I've been asking myself the same question