The form of AI interaction is about to undergo a new transformation. According to multiple sources, OpenAI plans to officially launch a new always-on AI assistant codenamed "O" at the upcoming OpenAI DevDay.

From current hints, the project's internal codename is Aeon, and related configuration clues have already surfaced in the ChatGPT configuration files. Unlike traditional models that require users to initiate conversations, this always-on assistant stands out by running continuously beyond regular chats, and may even have its own digital identity.

However, many core details about this highly anticipated mysterious assistant remain unclear. For example, what complex task scenarios it supports, what level of system permissions it has, whether it will fully support scheduled tasks and long-term memory functions in the future, and the final official release date, all remain to be revealed by the official during the event.

Notably, there were previous rumors in the industry that OpenAI was secretly developing an innovative hardware device with a doughnut-like appearance. If this hardware project later becomes related to the always-on software agent codenamed "O," this unique single-letter codename would undoubtedly bring fascinating tech puns to the market.