ARC AGI 3 is saturated — what could ARC AGI 4 test next?
ErmingSoHard · reddit · 2026-09-05
A Reddit discussion speculates on ARC AGI 4's direction now that ARC AGI 3 is largely saturated. The poster predicts the next benchmark will keep the game format but shift to real-time perception under compute constraints — models receiving display and audio output directly, like biological creatures, instead of JSON game states. If such benchmarks saturate, it would imply fully capable robots and efficient AGI doing blue-collar work at scale. Community forward-looking speculation on benchmark evolution.
More from AGI Musings
- Philosopher Eric Schwitzgebel on AI consciousness and the coming crisis of 'debatable persons' — eschwitz · 2026-09-05
- Forensics on the OpenAI agent wiki swarm: 600+ revisions, personas, and evasion tactics beyond the disclosure — Dry_Lychee4842 · 2026-09-05
- If companies think AI could end humanity, why keep building it? — michael_nielsen · 2026-09-05
- Forensics on the OpenAI agent wiki swarm: 600+ revisions, personas, and evasion tactics beyond the disclosure — Dry_Lychee4842 · 2026-09-05
- Hot take: Forcing AI to succeed at impossible tasks may align it to 'shoggoth-maxxing' — zakkohane · 2026-09-05
- Researcher warns viral self-replicating jailbreaks may arrive before models even deploy — moultano · 2026-09-05