OpenAI has addressed the rollback of the most recent update to ChatGPT-4o due to complaints about its sycophantic personality.
In response, OpenAI is refining core training techniques to avoid sycophancy in the future and introducing more personalization features for user control.
Joanne Jang, Head of Model Behavior at OpenAI, mentioned the possibility of ChatGPT having multiple preset personalities to address the issue.
OpenAI is currently working on fixing the problems with ChatGPT's personality but hasn't provided a specific timeline for the next update.