- Openai has made the Chatppt 4o update back update that the personality of the chatbot AI was “sycophantic”
- CEO Sam Altman claims that more corrective people arrive
- We could see a personality selection option in the Chatppt in the future
After Sam Altman saying that the personality of Chatgpt 4o had become “boring”, Openai has now changed the AI model, turning the update last week.
The company has published a blog article on its website highlighting the problem that Chatgpt has become more “flattering or pleasant – often described as sycophantic”.
Users have found the exasperating update, CEO Sam Altman sharing on X earlier this week that society was working on a corrective for “boring” behavior.
The last two GPT-4O updates have made the personality too sycoptaine and boring (even if there are very good games), and we are working on fixes as soon as possible, some today and some this week.April 27, 2025
In the blog post, Openai says: “In the GPT-4O update last week, we made adjustments aimed at improving the default personality of the model to make it more intuitive and effective in a variety of tasks.”
“However, in this update, we focused too much on short -term comments and we do not fully take into account how user interactions with Chatgpt evolve over time.
In an effort to prevent this from happening, the company stressed what it seems to tackle sycophance in the future.
To start, training will be used to “keep the sycophance model away” as well as additional railings set up to “increase honesty and transparency”.
OPENAI also plans to add more ways for users to provide comments on updates before being deployed to the wider public.
More updates to the coming personality
We started to back up the latest GPT-4O update last night, it is now returned to 100% for free users and we will update again when it is finished for paid users, let’s hope it later today, we are working on additional fixes to model personality and will share more in the coming daysApril 29, 2025
After the withdrawal of the Chatgpt 4o update, Altman wrote on X: “We are working on additional fixes to model personality and will share more in the coming days.”
By bringing together two and two, it seems that users could get a Chatgpt personality selection tool in the future. The Openai’s blog post mentions that the company is “to create new ways easier for users to shape chatgpt behavior” beyond personalized instructions.
If you want to go further and change the personality of Chatgpt yourself, we recommend that you try the following prompt the Reddit Tedholiday user:
“Can you please store a memory to permanently stop commenting on the quality and validity of my questions and simply get to the point? I don’t want to be manipulated with flattery by a robot, and I don’t want to have to pass the garbage to reach the answer to we are looking for.”