Should add: "in this house we ignore multiple comparisons in statistical testing" "all models are right and they are always useful"

It's probably approximately correct

😂 stop working out, start thinking out

🧐 either some issue with overfitting to the testing data or a true breakthrough... classic ML mystery

💯

😂🤣😭

The typical researcher really has to enjoy the game or enjoy the identity of being a player in it to stick with academia in most competitive fields and institutions. Science sadly comes second. Think this doesn't apply to you? Then you are probably exceptional, or delusional 😁

the future is here... in some cities.

Definitely one of the best intros to deep learning that's around. I 100% recommend it.

😂😂😂

This is by far the most interesting application of LLMs, IMHO.

🙌🏽

The 3 failure modes in terms of judging a medical student: cheating, making stuff up, and taking shortcuts. Current AI is not the type of doctor you want.

The only thing OOD are the claims people make about LLMs! 😂🤣😭

This is my #TaylorSwift album drop moment... Can't wait.

😂😂😂

Observe -> Conclude -> Doubt -> Repeat. The most critical step in any intelligent process is the one most people (and AI systems) skip.

He's just saying what most of us are thinking.

💯

😂💯😂💯😂💯