← Back to context

Comment by lynndotpy

20 hours ago

Two weeks ago, I heard rumblings that UGC, RL=L, and a matrix multiplication lower bound (even lower than the one here, but so unbelievably lower I think it was a typo), and a few others were about to be proven by someone at OpenAI. I find myself saying "big if true" a lot lately.

Even those on their own were enough to make your head spin. But seeing about 100x that? Geeze.