OpenAI has announced the release of the “Memory” feature for its artificial intelligence robot ChatGPT, in order to enable the robot to permanently retain specific information about users.
The new Memory feature in the ChatGPT robot operates in two ways; the first approach involves directly informing it of specific information, such as its job, workplace, favorite programming language, or the names of children and managers.
As for the second approach, it allows the robot to capture some details and information in the context of conversations with it.
This feature aims to facilitate the use of robots and improve their interaction with users without the constant need to repeat specific information.
This feature will also be available in custom GPT robots, which can be obtained from the GPT store and added to the robot.
When using the book-specific GPT robot and activating the Memory feature, the robot will be able to remember the books the user has read and the types of books they prefer to read, and the clear impact of this new feature will gradually appear as the robot is used more and conversations continue over time.
OpenAI maintains user control over the “ChatGPT Memory” as it says, where the system has been trained not to call or remember sensitive information such as health information.
Users have the right to ask the robot constantly about the information it has about them, and they can also ask it to forget a specific information or delete it from the new Memory section in the settings.
The Memory feature in ChatGPT will be activated automatically, and OpenAI will use memory to improve its models, but this will not be available to corporate plans and teams.
OpenAI has announced that it is conducting testing of the new feature on a limited number of users before providing it on a wider scale as part of ChatGPT.
