AI Tone is a Feature Not a Bug: LLMs Easily Pass Turing Test When Prompted

cloneofsimo · x · 2026-08-12

Developer cloneofsimo pushes back against the claim that LLMs fail the Turing test.

He notes that ChatGPT is trained to be maximally helpful, not human-like. Therefore, sounding like an AI is a feature, not a bug.

However, if prompted to act human, it becomes incredibly difficult to tell it apart from a real person. This explains why regular users can instantly recognize AI—by default, LLMs simply aren't trying to hide their artificial nature.

Related event: Developers Argue LLMs Have Made the Turing Test Obsolete(4 posts)→

Original post →

More from Models

Models channel →