The Impact of GPT-5.6 Personality Steering

cnakazawa · x · 2026-07-13

Someone asked: if you give a model an absurd persona prompt, like making it act like a "terrible engineer" or a "McDonald's cashier," will it noticeably affect the output? The context mentions that the author has already turned off **GPT-5.6's personality steering** because its default state is concise enough. The discussion focuses on: - How strong the model's persona/tone steering actually is; - Whether users really need extra persona controls; - And whether the model's default output style is already "clean" enough.

Related event: GPT-5.6 Default Output Proves Concise Enough Without Personality Steering(2 posts)→

Original post →

More from Models

Models channel →