OpenAI recently revealed some big news. They announced plans to change how they name their models. The next model might not be called GPT-5. It could have a new name and a different kind of architecture. This means it may work in new ways that we haven't seen before.
The research team worked on something called "super alignment." This team had trouble getting enough computer power from OpenAI. They only received 20% of the resources they needed. Because of this, the team couldn't do all the work they wanted to do.
There is more news about the researchers from this team. Some of them are joining a company called Anthropic AI. They will continue their work on super alignment there. Anthropic AI focuses on making AI systems that are safe and reliable. They want to make sure these systems can learn and make decisions in the right way.
Jan Leike, who worked at OpenAI, will join Anthropic AI. He is excited to work on new projects there. He talked about working on "scalable oversight" and "automated alignment research." These are things that help AI systems learn better and stay safe.
In another piece of news, Microsoft is facing an investigation. They have a new feature called Recoil AI. This feature tracks everything you do on your computer. It watches where your mouse moves and what you click on. This raises questions about privacy. People are concerned about how this data will be used and who will have access to it.
These updates show how fast AI is changing. OpenAI's new model and the move of top researchers to Anthropic AI are both big steps. They show how important it is to keep improving and making AI safe for everyone.