**[OpenAI Personal Assistant Agent Leaked Again: This Time It Might Be Called Dots]** Beating AI Newsflash: OpenAI's next-generation personal agent might be named Dots. Unreleased documents shared by M1Astra and ChrisGPT reveal that it can execute tasks autonomously. Users can contact it via messages, phone calls, Slack, and email, and it can even make purchases on behalf of users with approval. Dots can independently decide when to act directly and when to seek user approval first, with users able to predefine rules. Each Dot operates on an independent VM, with an interface called "Your dot's computer." It can also gain access to the user's computer, which users can revoke at any time and reauthorize through a desktop app. The same batch of documents also mentioned the ability to securely save browser passwords, enabling ChatGPT to automatically log into websites. Each Dot will feature a unique 3D avatar, with the initial release limited to Pro users. A few days ago, the ChatGPT Pro upgrade page briefly displayed "o, your always-on assistant," and the client configuration included `o` and `-o` email suffixes. In previously leaked code, Dots were also linked to internal names like Orbit and Aeon. OpenAI has not officially announced this product yet, so whether `o`, Dots, and Aeon are different products, internal code names, or names for different stages of the same product remains unclear. [Original Link]