One day ChatGpt may remember our entire existence

Sam Altman imagines an artificial intelligence capable of documenting and remembering a person's entire existence . An idea that, as disturbing as it may seem, has its own logic. Or rather: it makes perfect sense that it is the CEO of OpenAI who proposes it.

Speaking a few days ago at an event organized by Sequoia Capital , Altman explained to the audience his vision of “personalized” AI.

All life in an AI

The ideal, Altman said, is a “very small reasoning model, with a trillion context tokens, into which you put your whole life.”

“This model,” explained the leader of OpenAI, “is able to reason about all your context efficiently. Every conversation you’ve had in your life, every book you’ve read, every email you’ve received, everything you’ve watched is in there, connected to all your data from other sources as well. And your life continues to add to that context.”

“Companies basically do the same thing with their corporate data,” he added.

With these words Altman effectively described an essential artificial intelligence, fueled by an enormous flow of information, which becomes the lens through which a person looks at the world or makes decisions .

Memories entrusted to a chatbot

A chatbot's memory is the most fascinating, and at the same time controversial , element of an artificial intelligence.

The fact that an AI is able to remember every single conversation it has had with the user, and extract useful information from this to offer increasingly personalized and effective responses, represents a natural and inevitable evolution of this technology.

To become a truly complete assistant, capable of booking a trip but also offering personal advice, AI must know us . Deeply .

Privacy risks

Is the data we share with AI really safe? How is it treated? Where is it stored? Does it contribute to the training of future models? And above all: is there a risk that, one day, it could be stolen or misused?

The answers to these questions exist, but they are not always clear or reassuring . We are living in a moment of transition, in which the boundaries between convenience, security and data control are still blurred.

In some areas of the world, the first steps towards more robust regulation are being taken — as in the case of the European AI Act — but, at the moment, the responsibility for data management falls largely on users and on the strategies of companies operating in the artificial intelligence sector.