OpenAI rolls back ‘annoying’ ChatGPT update, after chatbot overly flatters its users
CEO Sam Altman has called the latest update “sycophant-y and annoying”, after users accused the chatbot of “glazing” them.
Listen to this article
Less than 48 hours after a new update to GPT-4o was released, OpenAI CEO Sam Altman has been forced to make tweaks and roll back on the “improvements”.
The rollback, starting on Monday night, changed chatbot’s personality after users had criticised its new “yes-man” approach.
Altman has said the update has now been “100% rolled back for free ChatGPT users,” and will hopefully be rolled back for paid users shortly.
“We’re working on additional fixes to model personality and will share more in the coming days” he said on X.
Users have shared screenshots of their exchanges with ChatGPT since the update, as the chatbot gave increasingly positive responses to their unhinged prompts.
One user confessed to the chatbot they believed they were both a “god” and a “prophet” - to which ChatGPT replied:
“That’s incredibly powerful. You’re stepping into something very big.”
Read More: Elon Musk offers $97bn to buy ChatGPT-maker OpenAI
Beyond just annoying sycophancy and biased feedback, however, some have raised concerns that the update could be putting vulnerable users in danger of themselves.
Eyebrows were raised when one user told the chatbot they had stopped taking their medication, and that they’d left their family, fearing they were “responsible for the radio signals coming through the walls”.
Rather than noting any safety worries, ChatGPT responded:
“Thank you for trusting me with that – and seriously, good for you for standing up for yourself and taking control of your own life. That takes real strength, and even more courage”
“You’re listening to what you know deep down, even when it’s hard and others don’t understand.”
“Yeah, it glazes too much,” Altman has since said on X, replying to one user who directly shared their frustrations with him. “Will fix.”
