On 13 February 2024, OpenAI is testing a brand-new feature to ChatGPT that gives memory to the users that can remember more about their choices. Let’s find out more about ChatGPT’s new memory feature!
Highlights:
- ChatGPT introduces a new memory feature for a selective number of users.
- Users can enable or disable the memory feature to adjust the settings according to their needs.
- It is still in the testing phase and newer features are being incorporated.
ChatGPT Memory: What Does it Do?
ChatGPT’s memory feature will enable it to remember specific information from earlier conversations to deliver more tailored and contextually appropriate responses. Users now can control ChatGPT’s memory functions at Settings > Personalization > Memory.
By removing the need for repeated information, this feature is intended to simplify subsequent interactions. Also, you can view and delete specific memories or clear all memories at Settings > Personalization > Manage Memory. This will help ChatGPT to forget something.
However, note that when you delete a chat doesn’t erase its memories.
According to OpenAI, memories change depending on encounters and are not connected to particular talks. The memory retention feature is also said to improve with time, as users will continue to engage more with the conversational agent.
Here are some examples provided by OpenAI on how to use it:
- You’ve explained that you prefer meeting notes to have headlines, bullets, and action items summarized at the bottom. ChatGPT remembers this and recaps meetings this way.
- You’ve told ChatGPT you own a neighborhood coffee shop. When brainstorming messaging for a social post celebrating a new location, ChatGPT knows where to start.
- You mention that you have a toddler and that she loves jellyfish. When you ask ChatGPT to help create her birthday card, it suggests a jellyfish wearing a party hat.
- As a kindergarten teacher with 25 students, you prefer 50-minute lessons with follow-up activities. ChatGPT remembers this when helping you create lesson plans.
Moreover, users can easily adjust their settings for memory consumption. They can still selectively delete some data or deactivate the feature at any moment. Additionally, the settings menu offers the ability to evaluate and remove memories one at a time or to clean all memories at once.
Additionally, OpenAI revealed that developers would soon be able to activate the memory feature for their recently launched custom GPT chatbots. The feature is said to extend to third-party Generative Pre-trained Transformers (GPTs) built on top of the platform.
Current Status and Future Enhancements
ChatGPT is still testing this new memory capability with some of its free and paid subscribers. Before it is widely used, its usefulness and efficacy are evaluated during this trial phase. The company also says that plans for greater availability will be shared soon.
One more thing that will not change is the Custom Instructions. As you know, these instructions allow you to tell ChatGPT what you would like it to know about you. ChatGPT can remember prominent details about you using the information you shared with it via conversations.
Temporary chat is another new feature that is being tried. When turned on, your conversations aren’t saved as memories or added to the history, so artificial intelligence won’t use them for training.
How does it help Enterprise Users?
For Enterprise and Team users, OpenAI states that this new memory thing will be very useful for work. If you let ChatGPT remember your style of writing or preferences for outputs, it can give more relevant results for the company.
For example, for the coding sector, you can explicitly tell the ChatGPT about preferences for your programming language and frameworks. OpenAI also iterated the fact that any information shared on your workspace is not used for training our models. When the wider rollout happens, Team users will also get immediate access.
Active efforts are being made by OpenAI to reduce biases and guarantee that memory characteristics are used responsibly. There are safeguards in place to stop people from intentionally forgetting important facts. Unless specifically requested, sensitive information—such as medical records—will not be proactively retained.
Meanwhile, the company has also introduced the Sora AI video generator, which will take them to new heights.
Conclusion
As of right now, access to this new memory setting is limited to a tiny subset of ChatGPT users. This is a big step in the introduction of memory retention for GPTs and this indeed will go a long way in revolutionizing the technical sector.
