Sam Altman told OpenAI employees last Monday that he was declaring a “code red” to improve ChatGPT and ward off threats from Google and other AI competitors, according to an internal memo viewed by The Information.
As a result, the company plans to delay progress with certain products including AI agents, which automate shopping and health tasks, Pulse, which generates personalized reports for users to read each morning, and advertising, which it has yet to publicly admit that it’s working on.
Altman didn’t specifically mention what he felt was wrong with ChatGPT, but he didn’t really have to. We all use it, and we know.
He simply said that, “We are at a critical time for ChatGPT” and directed more employees to focus on personalizing the chatbot for the 800M people who use it and letting each of those people customize the way it interacts with them.
Other priorities include:
- Improving Imagegen, its image generation tool
- Improving “model behavior” so that people prefer interacting with ChatGPT over other models.
- Boosting ChatGPT’s speed and reliability
- Minimizing “overrefusals”, which is when ChatGPT refuses to answer a question or perform a task
OpenAI was originally planning to launch GPT-5.2 later in December, but The Verge reports that pressure from competitors has moved the release forward, which is now earmarked for Dec 9th.
Robert Cyran summed up the situation nicely in a Reuters commentary piece:
“OpenAI clearly would benefit from greater focus… OpenAI is trying to do too much at once with technology that still requires a great deal of development and funding… The only thing bigger than the company’s attention deficit is its appetite for capital.”
