@ash_twtz Benchmark looks disappointing I don't think it is better than opus lol
Originale@ash_twtz Benchmark looks disappointing I don't think it is better than opus lol
Originale@ash_twtz one impressive result is exciting, but agi is a much bigger claim. i'd want to see consistent performance across different tasks before calling it that.
Originale@Pranto39 Yes bro that was just sarcastic but it was surprising that it got it right coz some models fail on it even with thinking mode enabled
Originale@ash_twtz lol
Originale@ash_twtz 6 fingers😂
Originale@ash_twtz Haha
Originale@ash_twtz try 7 fingers
Originale@ash_twtz L
Originale@ash_twtz Finally?
Originale@ash_twtz We went from “AI can’t count fingers” to “AI can use the entire computer. Character development 😭
Originale@ash_twtz Bros tests are built diff
Originale@ash_twtz Finally we have AGI
Originale@ash_twtz Crazy test
Originale@ash_twtz 😂
Originale@ash_twtz Soo interesting
Originale@ash_twtz 6 finger test back in action
Originale@ash_twtz your definition of AGI is actually Terra?
Originalehttps://x.com/ash_twtz/status/2114593070234952019
Originale@ash_twtz 😂😂
Originale@ash_twtz Yeah, it is absolutely AGI. Time: 2 minutes.
Originale@ash_twtz 6 fingers 😂
Originale@ash_twtz Even with thinking effort set to none it still got the finger count right this is wild
Originale@ash_twtz 6 fingers lol
Originale@ash_twtz Insane
Originale@ash_twtz hhhh
Originale@ash_twtz Good ine
Originale@ash_twtz Finally
Originale