Computer

ChatGPT now “remembers” users, the new potential of AI

When using a AI chatbotone of the main problems concerns the “memory” of this digital interlocutor who, obviously, practically has when opening each new chat reset the user’s identity and everything that concerns him at work (and beyond).

However, OpenAI He made it known that he was working on some sort of additional memory for ChatGPT which will allow this tool to remember the information about who is using it and the conversations occurred (including assigned tasks), so you don’t have to repeat all the commands every time.

A good step forward for artificial intelligence who “with the ability to remember”, gets closer and closer to potential of a human brain.

How ChatGPT memory works

The new feature works in two ways: the first affects the user who can say to chatGPT Of remember specific information, like names, for example, or functions already requested in the past. The second way, however, is much more interesting and allows the chatbot to learn graduallystoring various information as it is used.

Clearly too Everything is fine GPT will have its own memory and will be able to “remember” more things about the user who is using it.

At this point it is important to underline two things: the first is that the ChatGPT memory has a dual value, on the one hand it makes the chatbot more personal it’s more “close” to the useron the other hand it is about a strategically fundamental innovation which makes more intelligent the tool and much more autonomous on several fronts.

By remembering more things, in fact, the program will be able to carry out its functions more immediately and without the user having to remind him from time to time how to carry out a certain action and which parameters to respect.

In short, the possibility of ChatGPT to have “memories” is yet another step forward in a system that is gradually becoming more and more efficient and which, at this rate, will practically be aligned with man and its potential.

For now, however, this “memory” is still in the testing phase but it is easy to imagine its potential and the consequences it could have on the way you interact with chatGPT.

Furthermore, this innovation makes robots even more intelligent and, if we want, aware of their historyan extraordinary eventuality but one that could lead to several complicationsespecially if he was asked “special” things.

Doubts about the new functions

This change could cause quite a few problems to people, especially to users who are most attentive to theirs privacy who would absolutely not want their chatbot to store information about them and the requests made somewhere.

Clearly OpenAI immediately reassured users, underlining that they will still have one tight control over this memory and which, obviously, the system has been trained to forget some information deemed sensitivelike those on health, for example.

You can also explicitly ask chatbot what information does he remember? and, in case, cancellarle removing them from the memory management section.

Furthermore, for those who don’t want it, the new function can be deactivated at any time and the user will go back to using the “old” system without it remembering anything. However, by default the memory will be activatedso when the function arrives globally you just need to remember to “turn it off”

To know more: Artificial Intelligence: what it is and what it can do for us

Leave a Reply

Your email address will not be published. Required fields are marked *