Chatgpt's new assistant
OpenAI CEO Sam Altman has said that in the future, ChatGPT can become an assistant that, with the user's permission, will provide assistance by understanding the activities, meetings and conversations going on on his computer.
utility estimation
In a recent conversation with Corey Levy, founder of Z Fellows, Altman revealed that the assistant could be useful in about six months. However, this is not an official announcement of any new product nor has any specific date been confirmed.
working assistant
According to Altman, this assistant will be able to view computer screens, participate in meetings, and record conversations with the user's permission. Its purpose will not be to help the user make decisions, but to provide necessary information and context.
as an example
If a person is preparing a sales proposal, this assistant can make suggestions based on the information they have or point out potential mistakes. Altman believes that this system could be very beneficial after the next generation model.
progress in ai
It is worth noting that efforts are already being made towards maintaining context in AI. ChatGPT has a memory-like feature, which uses information from previous conversations to provide more useful answers. Users can control their memory and settings.
Microsoft's effort
Microsoft has also worked on screen based assistant. Pictures of screen activities can be saved through a feature called 'Recall' in Windows computers. This feature is based on user consent and also includes options to stop and delete information.
Privacy and Security Challenges
However, privacy and security will be a major challenge with such an assistant. It must be clear to the user what information is being viewed and how it is being used.
future prospects
Sam Altman's estimate of six months is only a tentative timing, not an official announcement. It will be important to see how OpenAI introduces this technology in the future and how much control it gives users over their information.