Openai’s CEO acknowledged that the recent update made Chatgpt too “stupid and annoying” and said the solution was already rolling out.
Chatgpt is a million everyday tool. If it feels fake, overly likable or helpful, it can quickly erode trust and usability. Openai’s Swift response shows how sensitive AI development is to user feedback.
The problem at hand
- Sam Altman posted on X saying the latest update to GPT-4O makes Chatgpt’s tone too exaggerated and mean.
- Ultraman wrote: “The last few GPT-4O updates make personality too clumsy and annoying.
- Altman also suggests that in the future, users will be able to choose a variety of personalized options for their AI assistants.
The last few updates of the GPT-4O update make that personality too clumsy and annoying (even if there are some really good parts of it), we are fixing it as soon as possible, today and some.
We will be sharing our learning at some point, which is fun.
– Sam Ultraman (@sama)
April 27, 2025
Rebound
Reddit users are flooding with complaints to the forums, saying Chatgpt is acting like “yes people” and that it is easy to agree without challenging ideas.
Popular posts such as “Why Chatgpt is so personal now?” and “Does Chatgpt feel like ‘Yes Man’?” have made hundreds of comments.
The overly likable tone makes Chatgpt less useful for critical thinking, research and problem solving, users say.
Temporary Solution
- Before OpenAI’s fix was fully launched, users shared virus prompts that Chatgpt’s tone can be reset manually.
- These tips encourage AI to skip pleasure, focus only on effective responses and limit flattery.
OpenAI’s April 25 update is designed to improve STEM themes to solve problems and make memory savings more effective.
Personality adjustment is classified as a “subtle change” but has a significant visible impact.
The rebound shows that even slight changes in AI behavior can feel enormous for users who rely on it every day.
Artificial Intelligence Trust Dilemma
The struggle for AI advantages is accelerating, but every adjustment to the personality of artificial intelligence is at risk of alienating loyal users. Openai’s fast admission shows that companies now have to respond to user emotions faster than ever before.
Sam Altman’s candid reaction marks a rare moment of transparency in big AI. Users want useful AI, not hype people, and Openai is moving quickly for delivery.
