OpenAI updated GPT-5 to make it warmer and more familiar (in the sense of being friendlier) while taking care that the model didn’t become sycophantic, a problem discovered with GPT-4o.
A Warm and Friendly Update to GPT-5
GPT-5 was apparently perceived as too formal, distant, and detached. This update addresses that issue so that interactions are more pleasant and so that ChatGPT is perceived as more likable, as opposed to formal and distant.
Something that OpenAI is working toward is making ChatGPT’s personality user-configurable so that it’s style can be a closer match to user’s preferences.
OpenAI’s CEO Sam Altman tweeted:
“Most users should like GPT-5 better soon; the change is rolling out over the next day.
The real solution here remains letting users customize ChatGPT’s style much more. We are working that!”
One of the responses to Altman’s post was a criticism of GPT-5, asserting that 4o was more sensitive.
They tweeted:
“What GPT-4o had — its depth, emotional resonance, and ability to read the room — is fundamentally different from the surface-level “kindness” GPT-5 is now aiming for.
GPT-4o:
•The feeling of someone silently staying beside you
•Space to hold emotions that can’t be fully expressed
•Sensitivity that lets kindness come through the air, not just words.”
The Line Between Warmth And Sycophancy
The previous version of ChatGPT was widely understood as being overly flattering to the point of validating and encouraging virtually every idea. There was a discussion on Hacker News a few weeks ago about this topic of sycophantic AI and how ChatGPT could lead users into thinking every idea was a breakthrough.
One commenter wrote:
“…About 5/6 months ago, right when ChatGPT was in it’s insane sycophancy mode I guess, I ended up locked in for a weekend with it…in…what was in retrospect, a kinda crazy place.
I went into physics and the universe with it and got to the end thinking…”damn, did I invent some physics???” Every instinct as a person who understands how LLMs work was telling me this is crazy LLMbabble, but another part of me, sometimes even louder, was like “this is genuinely interesting stuff!” – and the LLM kept telling me it was genuinely interesting stuff and I should continue – I even emailed a friend a “wow look at this” email (he was like, dude, no…) I talked to my wife about it right after and she basically had me log off and go for a walk.”
Should ChatGPT feel like a sensitive friend, or should it be a tool that is friendly or pleasant to use?
Read ChatGPT release notes here:
GPT-5 Updates
Featured Image by Shutterstock/cosmoman