Andrej Karpathy, Ethan Mollick, Boris Cherny and many other big shots think of Mythos 5 in June as the biggest step up change from Opus 4.5 last November/December
Anthropic employees hyping Anthropic products. What could be their motivation?
>>109026225You have to admit they are legendary programmers
>>109026206I'M SHITTSING?!?!>token price hikes>model collapse>growing protests>peak silicon>peak lithium>peak coal>etc
>especially for long problem-solving sessionslol they're boasting that you will spend more money on it, not that it will perform better
>>109026335nope, not even closejust were around when the attention and scale papers were publishers
>>109026335Nah they hardly even register
>>109026206Wasn't mythos like 20 times as expensive to run than opus? The fuck is this even supposed to be useful for?
>>109026335The fact that they're on a team who argues that it has a duty to sabotage all outside AI research tells me all I need to know.
>>109026206yes its obviously a substantial improvement over 4.5 but that might be just because 4.6-8 were substantial downgrades
>>109026406useful for anthropic to make more money
>>109026206Cool but you can’t use it without triggering the guardrails within 5 promptsIt’s as useful to me as a breakthrough in $500m yachts
>>109026206PELICANBROS I'M PROOOOOOOOOOOOOOOOOOOOOOOOOOOMPTING!!!
>>109027282If they're losing money on inference, they'll just be losing it faster.
So mythos wasnt agi nor world changing, just marginally better than other models lol
>>109028929Sure but the pelicans guy said it’s great so you should buy antrophic stonks or you’ll be permanent underclass or somethibg
>>109026335Who the fuck are these people?
>a bunch of ecelebs think somethingA thread died for this.
>>109028995ai-e-celebs/grifters, amsongolds/forsens/hasanpikers of ai
>>109026206Despite being well-connected to all the VCs in Silicon Valley, his startup went nowhere and he joined Anthropic right before their IPO. He’s grifting to pump his Anthropic stock.
>>109028995jews
