Not like this... I thought OpenAI and Deepmind have solved Navier Stokes :(
it's always very funny when you see people go:>oh we're x number of years away from agiand in the next breath they say:>we just need n number of breakthroughs to make it happenlike these known unknowns will arrive on some unknown, but entirely bounded, schedule that will make the agi forecast real
>take Machine Learning course 5 years ago>very first definition of ML says that it can only model the patterns in its training set>2 years later everyone collectively forgets this simple fact
>>107505125>attend machine learning presentations from big-ish firms in the industry>everyone says machine learning is mostly when you ran out of ideas>2 years later everyone loses their shit over a fucking shatbot
>>107504673anyone else attend neurips? I was working my company booth
