If AI-built AIs stop modeling humans, do humans still matter operationally?

PierceLilholt · x · 2026-07-26

A short but pointed AGI question: if future AI systems are built by AIs and no longer need to model humans closely, do humans still matter operationally in the loop?

It frames a bigger debate about whether human behavior remains a necessary reference point for alignment, product design, evaluation, or whether AI systems can become internally self-referential and optimized around machine-to-machine objectives.

Original post →

More from AGI Musings

AGI Musings channel →