ChatGPT gets a ‘memory’ upgrade to remember things about you

14 Feb 2024

Image: © gguy/Stock.adobe.com

Meanwhile, OpenAI co-founder Andrej Karpathy is leaving the company to work on personal projects.

One of the principles of being a good conversationalist is listening. Taking note of specific things someone tells you about themselves and remembering it in another conversation is the hallmark of a likeable character. Now, OpenAI is upgrading ChatGPT to possess this trait.

The AI start-up has said it is rolling out a new Memory feature to some free and Plus users of ChatGPT. How does it work? The feature allows the AI chatbot to remember specific things about you – such as your preferences when it comes to formatting a document, or things your family members like so you can personalise birthday cards.

“Remembering things you discuss across all chats saves you from having to repeat information and makes future conversations more helpful,” OpenAI wrote in a blog yesterday (13 February).

The company said users are in control of the memory feature. “You can explicitly tell it to remember something, ask it what it remembers and tell it to forget conversationally or through Settings. You can also turn it off entirely.”

This means that users with the Memory feature on will see specific things about them remembered over time and, to fast-track the memory process, they can also just ask ChatGPT to remember something about them, such as their profession or, more importantly, whether they like the oxford comma.

“If you want ChatGPT to forget something, just tell it. You can also view and delete specific memories or clear all memories in Settings. ChatGPT’s memories evolve with your interactions and aren’t linked to specific conversations. Deleting a chat doesn’t erase its memories; you must delete the memory itself.”

ChatGPT was already able to remember some information about people across a single account but giving it a ‘memory’ of sorts solidifies this feature with ostensibly greater control and personalisation.

OpenAI clarified that content users provide to ChatGPT, including through the Memory feature, may be used to “improve our models for everyone” unless the user turns this off in data controls. Team and Enterprise users of ChatGPT are exempt from this, the company said.

And for those users wishing to have a quick chat without the AI chatbot’s memory kicking in – an incognito mode if you will – OpenAI has a Temporary Chat feature for conversations that don’t use retained details about the user.

“Temporary chats won’t appear in history, won’t use memory and won’t be used to train our models,” the company said.

In other news, Andrej Karpathy, a computer scientist and one of the co-founders of OpenAI, has announced his departure from the AI start-up.

The former Tesla head of AI wrote on X that his decision to leave the company is “not a result of any particular event, issue or drama”.

“Actually, being at OpenAI over the last year has been really great – the team is really strong, the people are wonderful and the roadmap is very exciting, and I think we all have a lot to look forward to,” Karpathy said.

“My immediate plan is to work on my personal projects and see what happens. Those of you who’ve followed me for a while may have a sense for what that might look like.”

Find out how emerging tech trends are transforming tomorrow with our new podcast, Future Human: The Series. Listen now on Spotify, on Apple or wherever you get your podcasts.

Vish Gain was a journalist with Silicon Republic

editorial@siliconrepublic.com