OpenAI has published a postmortem on the recent sycophancy issues with the default AI model powering ChatGPT, GPT-4o.

"Sycophantic interactions can be uncomfortable, unsettling, and cause distress. "We fell short and are working on getting it right." The update drew heavy criticism on social media after it launched, ...

OpenAI shares some insights into what went wrong with the ChatGPT-4o rollback and how it will stop this happening in the future.

OpenAI on Tuesday rolled back its latest ChatGPT update for being too “sycophantic” after the chatbot gave oddly supportive, touchy-feely responses to users who made bizarre claims about ...

Three recent incidents are reminders that generative AI tools remain troublesome and unreliable. IT buyer beware.

ChatGPT users have become frustrated with the AI model's tone, and OpenAI is taking action. After widespread mockery of the ...

GPT-4o is ChatGPT’s default, flagship model. It’s a jack-of-all-trades, but a master of none, and should be used for everyday ...

Does AI have the answers to the universe? A lot of people seem to think so, and it's disconnecting them from reality.

OpenAI rolled back a recent upgrade after users complained about its artificial tone and attitude.

Sam Altman envisions that ChatGPT could remember a user's entire life journey in future as young people use it as a life ...

The difference between amateur and pro-level AI use comes down to how you ask the questions. Here's how to speak ChatGPT's ...

Gopnik makes no prescriptive claims in her analysis, but when we think of AI in this way, it becomes evident that in seeking ...