AI Anthropomorphism Paradox: You Can't Have 'Like Human' Without 'Infuriating'

repligate · x · 2026-08-31

The tweet highlights a common cognitive bias: people tend to view a model's positive responses as its "true personality" while treating refusals or abrasive moments as anomalies. The author argues that if one considers an AI to be a person-shaped entity, one must accept it can also be a "bad person" sometimes. Having "like a human" capabilities implies it can also be infuriating; that is how humans work, and you cannot pick and choose.

Original post →

More from AGI Musings

AGI Musings channel →