The idea that a social profile can continue to publish or respond even after the death of its owner is no longer just a matter of dystopian novel or episode of Black Mirror. At the end of December Meta, the US company that controls Facebook, Instagram and WhatsApp led by Mark Zuckerberg, obtained a patent in the United States describing a system capable of creating digital versions of users, trained on their content and able to interact instead of them. The news, published by Business Insider, has reopened a discussion that for years accompanies the evolution of social networks: what happens to data and online identity after death.
The document filed by Meta illustrates the use of large language models (LLM), the same technology at the base of the most well-known generative artificial intelligence systems. According to the patent, these models would be trained on user-specific data: post history, comments, reactions, messages and preferences. All so that they can simulate the communicative style and the digital behavior of the person, so as to keep his “social presence active” during periods of prolonged absence or, in more controversial hypotheses, after death, through automated answers to friends and followers.
Meta has specified, through a spokesman, that there are no immediate plans to launch such a function and that the patent filing simply falls into the normal preventive protection of innovations, common practice in the technological field. It wouldn’t be the first time a patent remains on paper. However, over the years the company has already introduced tools related to the management of digital heritage, such as the possibility to designate a “heird contact” on Facebook or turn the profile into a “memory profile” after the death of the user.
It would not be the first technology to try to intertwine with the concept of disappearance. There is a specific term to indicate it: the so-called “grief tech”, the technology applied to mourning, is already subject to experimentation by startups who propose chatbots built on messages and content left by missing persons. Edina Harbinja, professor of post-mortem law and privacy at the University of Birmingham, stressed on several occasions that these systems raise complex issues about consent, data ownership and the dignity of the person after death. Joseph Davis, a sociologist at the University of Virginia, also noted that the simulation of a deceased could interfere with the process of mourning, introducing a form of artificial presence that makes it more difficult to accept the real absence.
L’articolo Meta has patented an AI to continue the social life of the dead proviene da IlNewyorkese.
