- OpenAI is currently testing a new “memory” feature on its ChatGPT chatbot.
- The feature is still in the experimental stage and is being rolled out for select users.
- OpenAI has stated that it might utilise content shared by users with ChatGPT.
To improve user experience, OpenAI has announced a new feature called “memory” for ChatGPT. With this feature, ChatGPT can recall information discussed in previous conversations, minimising repetition and enhancing the efficiency of future interactions. Additionally, users will have the option to instruct ChatGPT to forget specific information as needed. Here is more about the new feature in detail.
New memory feature for ChatGPT
- The AI-reserch company, OpenAI, is currently testing the new “memory” feature on its ChatGPT chatbot.
- The memory controls on ChatGPT enable the chatbot to remember or forget key instructions and details shared by users.
- This feature is expected to reduce the need for users to repeat information and personalise conversations.
- The feature is still in the experimental stage and has begun rolling out for select users from both free and Plus tiers.
Here’s more about ChatGPT’s new feature
In interactions with ChatGPT, users have the option to prompt the AI to remember particular information or allow it to autonomously pick up details. The memory function of ChatGPT is expected to enhance with continued usage, leading to noticeable improvements over time.
- Users have the ability to view and delete specific memories or clear all memories in settings (by following: Settings > Personalisation > Manage Memory).
- Memories in ChatGPT develop based on interactions and are not associated with specific conversations.
- Deleting a chat does not remove its memories; users must delete the memory itself.
- Additionally, users can disable memory at any time, preventing the creation or use of memories while it’s turned off.
OpenAI has stated that it might utilise content shared by users with ChatGPT, including memories, to enhance their models for all users. However, users have the option to disable this feature through their Data Controls.
Furthermore, if users prefer to engage in a conversation without utilising memory, they can opt for a temporary chat. Temporary chats do not appear in the chat history, do not utilise memory, and are not used for training OpenAI’s models.
Additionally, GPTs will now have individual memories, and builders can choose to activate memory for their GPTs. Memories, similar to your chats, are not accessible to builders. To engage with a memory-enabled GPT, users must also have memory enabled.
