Lawyer's One-Liner Deflates AI Consciousness Debate: Saying "I'm Pregnant" Isn't Being Pregnant

ccerrato147 · x · 2026-09-27

The author argues that the entire debate over whether AI is conscious rests on a single kind of evidence: what the model says. That evidence, he contends, is worthless in both directions.

He cites an appellate lawyer's line from that morning: train a model to say "I am pregnant" and it is not pregnant. The point being that a model's self-report is just trained text, not proof of consciousness.

Related event: AI Claims of Consciousness Are Not Evidence, Argues Writer(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →