>>13478351Research associate in ML here, I would never talk about what I'm doing because it is so specific I would doxx myself.
>>13478405what does this have to do with AI you mong, current "AI" is an overhyped decision function
>>13478347unless your training throws NaN loss at you directly before a deadline and you just go to a corner and cry
>>13478302How should it not be hard? We build tools all the time to be better than we are, an AI is still a tool. I can't calculate stuff in my head very fast, a computer is extremely fast, so why shouldn't it be better? Also your image makes no sense
