About whether current models learn ARC-AGI-3 game rules like humans do.
Has this happened?
Yes 23 (22%)Not sure 61 (59%)No 20 (19%)
Votes cast 1–2 October 2026: 100,590 votes from 9,694 people.
Hacker News has set AI a lot of challenges over the years. Which ones has it met?
About whether current models learn ARC-AGI-3 game rules like humans do.
Votes cast 1–2 October 2026: 100,590 votes from 9,694 people.
Arc-Agi 3 is coming soon, I'm very excited for that because it's a true test of multimodality, spatial reasoning, and goal planning. I think there was some preliminary post somewhere that did show that current models basically try to brute-force their way through and don't actually "learn the rules of the game" as efficiently as humans do.