AI-driven digital resurrection and cognitive impacts

The emergence of thanatobots, or “griefbots,” involves the use of generative artificial intelligence to simulate the personalities and speech patterns of deceased individuals. These systems utilize digital footprints, such as emails, voice recordings, and social media posts, to train models that predict how a person might respond to specific inquiries.

While these bots can provide emotionally convincing interactions, they do not possess consciousness or actual memories. Instead, they rely on data architecture to calculate response probabilities, which can lead to the generation of factually false responses that attribute opinions or memories to the deceased.

As the technology expands, companies in regions such as Beijing are developing interactive digital versions of the dead to allow families to engage in real-time conversations. This development has sparked significant ethical and philosophical debates regarding the distinction between symbolic continuity and real presence, as well as concerns over data ownership, economic exploitation, and the long-term control of digital avatars.

Recent developments highlight how the broader proliferation of generative AI and external digital memory is impacting human cognition. Research suggests that as information becomes more easily retrievable online, individuals may be less likely to encode it into their own memory, potentially hindering pattern recognition and complex reasoning. This shift coincides with observations that the Flynn effect—the historical rise in IQ scores—has stalled or reversed in several countries.