IvanIvanich

ловлю жуликов на вранье

[IvanIvanich @ivanich22893]

@RadishHarmers Brother. You didnt get the memo? AI model IS the expert. It both writes and reviews, then lean verifies, then spits the slop into the oblivion

[IvanIvanich @ivanich22893]

@introsp3ctor @amermathsoc You will not. The text is unreadable

[IvanIvanich @ivanich22893]

@zachary_horvitz is it though? they spent around $15mil on tokens to get this. My guess a group of mathematicians with 15mil budget would have produced a better result. Moreover, they would have actually understood the written text.

[IvanIvanich @ivanich22893]

@nialiv @gfodor nope, pure coincidence they put out the result on the same day after the real mathematicians showed their result 🤓

[IvanIvanich @ivanich22893]

@xcitenotxcel @deaneyang lmao, not remotely true

[IvanIvanich @ivanich22893]

@relizarov if only one could have stolen some cancer vaccine, developed by a professor in the course of his life, by running 10000 agents, then OpenAI would have already done that

[IvanIvanich @ivanich22893]

@BacekJubil86068 @DocOctagonical @JohnBal14046363 they wrote in the post they spend 300bln tokens

[IvanIvanich @ivanich22893]

can it run 10000 to steal other people's photos tho?

[IvanIvanich @ivanich22893]

@eerac either it was solved by that or by stealing idea from a professor who actually understands what is happening in the written text

[IvanIvanich @ivanich22893]

@joelwatsonfish what's on him? That the company-provider of the service decided to screw him over and spent more $$$ on compute than the guy will earn in his entire life, right after they learnt that he is close to the solution? lmao

[IvanIvanich @ivanich22893]

@johnennis Isn't it great? A bubble company with a huge operating net loss spends in 11 days more $$$ on tokens than that professor would earn in a lifetime just to screw him over

[IvanIvanich @ivanich22893]

@george_man37 @amermathsoc what in the world does it change? No one ever read the text/proof

[IvanIvanich @ivanich22893]

@techguyagent @boazbaraktcs get the fuck out of here you brain-damaged idiot. Bubeck himself wrote that they only started working on the problem because of the rumours

[IvanIvanich @ivanich22893]

@joelwatsonfish they did it with huge net-loss token-wise. PR-wise unclear

[IvanIvanich @ivanich22893]

@AlexKontorovich embarrassing statement. Praising slop that no alleged author ever read or understood. What are you cheering for....

[IvanIvanich @ivanich22893]

@thomasfbloom Thomas, would you consider ethical if you shared some progress on a problem privately, and then a guy rushed home to turn on 10 tabs with GPT on the same problem he just heard about for the first time to get ahead of you and publish the results first? (without understanding shit)

[IvanIvanich @ivanich22893]

@SebastienBubeck so... you admit academic misconduct? The guys privately shared the progress with you and you rushed to turn on 10000 agents with 9kk$ spent in tokens to get ahead of them, based on their ideas? nice

[IvanIvanich @ivanich22893]

@lennybronner they spent on tokens to attack this problem way more than Clay prize btw