Probably they will not! They lost an incredible amount of talent and people are still leaving on a daily basis. Check the handler of @xai , Elon should be worried about this non-stop bleeding. Compute is not an edge, especially if they are on an 11% MFU.
Deeply, I wish they will. Rooting for them.
Thereโs another way to read it, normal people still use these chatbots for many reasons. been hearing and reading that these bots became the fav โpsychotherapistโ for many, because they just tell people what they wanna hear, help them cope, and make them feel right , like, โthey just donโt understand me.โ A huge part of how people use them now is for emotional support, reflection, and self-regulation, not only for nerd stuff or productivity.
@pourfairelevide itโs more a psychological effect. People like to symbolize and make heroes. Saying someone โsolved philosophyโ sounds amazing, itโs less about the truth but rather about their need to create a hero who solved something messy and tough! Even himself admitted he was wrong later.
@noetic_emetic He didn't solve all of philosophy and was delusional to think he did, as he admitted in his latest work. Philosophy is not a single puzzle with a final solution.
@iScienceLuvr The entertainment business is huge and makes a lot of money. Running frontier labs costs a fortune, so imo it makes sense that these companies look for extra projects on the side to bring in more cash
Yang is responsible for one of the rare pretraining research breakthroughs at openai that worked at scale in the last ~year
congrats to MSL and hope my pretraining friends at openai are ready to pick up some major slack!
People who failed often dismiss new attempts. Not because the goal is impossible, but because admitting itโs possible would make their own failure harder to cope with. To calm their cognitive dissonance, they claim โit canโt be done,โ projecting their limits onto everyone else. Real constructive criticism points to a way forward, but destructive criticism is just a defense mechanism, not a real evaluation.
@leptis_magnus@faustocoppi60 It was an 8% gradient with a strong lead-out. I don't know what more he needs. Jonas dropped him on a 5% climb. Almeida is a top tier rider but needs to improve his positioning and explosiveness .