Engineer argues AI fingerprints in code and text are just lower-quality output, harming human communication

ryunuck · x · 2026-09-20

Software engineer ryunuck argues that text should not be recognizable as AI-generated. For code, there is an optimal formatting for any given task, so any fingerprinting is fundamentally lower-quality code — and since formatting produces identical compiled output, the fingerprint is meaningless anyway.

He extends the point to AI text slop: the most legible expression of any idea is one that reads as human, so model outputs that can be distinguished from human writing are by definition not optimized for human communication. He warns this subtly erodes society's fabric of communication, comparable to 2010s cringe culture.

Related event: Engineer Argues Detectable AI Text Is Itself a Sign of Low Quality(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →