minus-squareUnrepentantAlgebra@lemmy.worldtoTechnology@lemmy.world•Announcing ARC-AGI-3 - A benchmark that tests if AI can explore, learn, and adapt in unfamiliar situations. Humans score 100%. Frontier AI scores 0.26%.linkfedilinkEnglisharrow-up0·4 days ago If human scores were included, they would be at 100%, at the cost of approximately $250 Wait, why did it cost real humans $250 to pass the test? linkfedilink
Wait, why did it cost real humans $250 to pass the test?