OpenAI’s ChatGPT is undergoing testing to incorporate memory capabilities, aiming to improve user experience by remembering past conversations. Users retain full control over this feature, dictating what is remembered and how it’s utilised.

The introduction of the “Manage Memory” settings dialog empowers users to manage ChatGPT’s memory functions. This feature is designed to streamline future interactions by eliminating the need to repeat information.

Users can prompt ChatGPT to remember specific details or allow it to autonomously pick up information during conversations. Memory improves with continued usage, resulting in more personalised interactions.

Memory can be toggled on or off at any time through settings. Users can instruct ChatGPT to forget specific information or clear all memories. Memories are not tied to individual conversations, and deleting a chat does not erase associated memories.

Enterprise and Team users’ data is safeguarded with measures in place. ChatGPT would be able to remember tone, voice, and format, and apply them to blog posts without requiring repetition. In coding, users can specify their programming language and frameworks to ChatGPT, which it then remembers for subsequent tasks, streamlining the process.

Sensitive information, such as health details, will not be proactively remembered unless explicitly requested.

The memory feature will initially be rolled out to a select group of users, with plans for broader implementation in the future. OpenAI aims to assess feedback and refine the feature before wider release. Enterprise and Team users will also benefit from memory capabilities.

In addition to ChatGPT, GPTs will have their own memory capabilities, providing personalised experiences across various applications. Users will need to enable memory to interact with memory-enabled GPTs.

OpenAI is actively working to mitigate biases and ensure the responsible use of memory features. Measures are in place to prevent the proactive remembering of sensitive information.