News
OpenAI rolls back update that made ChatGPT a sycophantic mess
GPT-4o is not a new model—OpenAI released it almost a year ago, but the company occasionally releases revised versions of existing models. As people interact with the chatbot, OpenAI gathers data on the responses people like more. Then, engineers revise the production model using a technique called reinforcement learning from human feedback (RLHF).
OpenAI rolled back a GPT-4o update for ChatGPT that caused the chatbot’s default personality to be “overly flattering or ...
ChatGPT really, really likes you a lot. And that's a problem.In a blog post, OpenAI, the company behind the popular chatbot, ...
OpenAI has published a postmortem on the recent sycophancy issues with the default AI model powering ChatGPT, GPT-4o.
OpenAI also announced that ChatGPT will now show a list of the day’s top trending searches, and like Google, will attempt to ...
OpenAI is streamlining its AI model lineup, retiring popular models like GPT-4 and GPT-4.5, all in anticipation of the launch ...
OpenAI adds shopping features to ChatGPT, offering product recommendations with images, prices, and direct purchase links ...
OpenAI faces backlash over their for-profit plan. In a new letter citing AI safety, experts have asked AGs to intervene.
Microsoft possesses significant leverage to potentially obstruct OpenAI's plan to transition into an independent for-profit ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback