Superintelligence Value Alignment: Human Values Might Be Beside the Point

kellerjordan0 · x · 2026-08-08

Through a science-fictional thought experiment, the author explores the value alignment problem of AGI. He suggests that if AI achieves superhuman intelligence—just as humanity transcends the animal kingdom—human arguments over “whose values” the AI should have could end up being entirely beside the point. It would be akin to a vulture and a lion arguing over whether humans should prefer to scavenge or hunt. Concretely, machine supercultures might spin up their own value systems beyond human comprehension.

Related event: Speculation on ASI: Human Values May Become Irrelevant(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →