Openai declines a modern update to the GPT-4O, which is the default model that operates Chatgpt, after complaints from users that he made a chat robot behave like a weirdo. "The latest GPT-4O updates made very disturbing and annoying (although there are some good parts of it), we are working on repairs as soon as possible, some today and some of this week," The CEO of Openai Sam Altman said In the x publication I was monitored before Techcrunch.
As of mid-Tuesday, Altman said that Chatgpt was working on an old and less version of SYCOPHANTY of GPT-4O for all free users. The company hopes to restore users for an older version of the form by later time. "We are working on additional reforms for character modeling and we will share more in the coming days," Altman said, adding Openai will share more information about the error that happened "At some point."
The new Openai GPT-4O released late last week. By weekend, people began to notice that Chatgpt was excessively acceptable and is lengthy in his praise. As you can see from the X post below, this praise often was inappropriate and strange.
When is the OpenAi plug on the new GPT-4O?
This is the most disturbed model to date by anyone.
This is the moment of the Gemini image disaster in Openai.Photo credit: R/u/trevor050 pic.twitter.com/KNCDNEYMDQ
– Ashutoshshrivastava (@AII_for_sucss) April 27, 2025
Improving the emotional intelligence of its models, as much as it could possess the algorithm of the characteristic, was the modern axis of Openai. For example, With GPT-4.5The company said that the model was better in responding to warmth and understanding of its previous systems. Trying to bring the same ability to GPT-4O at reasonable prices, it seems that Openai has been wrong.
This article was originally appeared on Engadget on https://www.engadget.com/ai/openai-Rolls-back- update-that-Made-Chatgpt-n-ssing-wendo-203056185.html?
[og_img]
Source link