If AGI is defined by outcomes, we may already be there
sebnadeau · reddit · 2026-09-09
The author argues the AGI debate is asking the wrong question: AGI has never had an agreed definition, and the definition determines which questions matter.
The argument
- The 2022-2023 framing: "general-purpose intelligence matching or exceeding humans on intellectual tasks."
- Whether LLMs genuinely understand the world is a separate question—like saying airplanes don't fly because they don't do it like birds.
- The real question is what counts as an intellectual task. Taste, long-term judgment, and some causal judgment resist reduction into discrete, scorable tasks—there's no objective function for what makes a game mechanic good.
- If those abilities can eventually be measured and humans still win, AGI hasn't arrived. But if they can't be objectively measured in the first place, failing at them shouldn't disqualify a system.
Conclusion: under an outcome-based definition, we may already be there.
More from AGI Musings
- AGI Fellowship kicks off at Johns Hopkins Bloomberg Center — ghadfield · 2026-09-09
- Railroads in the 1830s: the only real historical parallel to AGI — Traditional-Chip8339 · 2026-09-09
- Pedro Domingos: 'AlphaGo didn't beat Lee Sedol — the humans who built it did' — pmddomingos · 2026-09-09
- Mechanize's Tamay Besiroglu: why labs achieve the same breakthroughs at the same time — tamaybes · 2026-09-09
- Connor Leahy: the plan for superintelligence is build it first, figure out control later — victor_explore · 2026-09-09
- Researcher: good vs bad AI futures hinge on DL training's alignment generalization — QuintinPope5 · 2026-09-09