human performance is comparable at 85%

o3 just scored 87.5% on arc-agi

𝓵𝓲𝓿𝓲𝓷𝓰 𝓽𝓱𝓮 𝓮𝓷𝓭𝓵𝓮𝓼𝓼 𝓼𝓾𝓶𝓶𝓮𝓻 • building @buoy 🛟 • skeet.build • pfista.com

https://arcprize.org/blog/oai-o3-pub-breakthrough

helping build farcaster | surfing @ venice, ca | eating @ gjusta / gjelina | “lucky me, lucky mud”

a huge leap forward

and they have no issue claiming this trajectory continues

today i'm back in the "i'm out of a job" camp

kinda crazy that 32% was the high beforehand, it blew it out of the water

looking forward to the chatgpt ultimate plan at $600/m