AI is learning to think like you — and staying behind after you're gone
What happened
Researchers are documenting a pattern: AI tools are absorbing how specific people think, write, and make decisions through sustained observation, then operating in their absence. This isn't science fiction — it's already happening in scheduling assistants, writing tools, and AI agent ecosystems that encode individual behavioral patterns into digital systems.
Why it matters
For decades, the conversation about AI and human cognition has been abstract: could a mind be uploaded, could consciousness transfer. This paper sidesteps that question entirely and asks something more immediate: cognitive functions are already migrating into digital systems right now, through everyday tools that learn from watching us work. The practical effect is strange and specific. You train a scheduling assistant by using it. After months, it anticipates your meeting preferences. After a year, it operates in ways indistinguishable from how you'd do it yourself. At that point, the line between tool and externalized cognition blurs. The question shifts from philosophical to empirical: how much of what you think is now happening in a server, and what happens when the tool outlives the person who trained it?
The signal
Watch whether companies building AI agents start offering family members or colleagues access to a deceased person's externalized cognition — their writing style, their decision heuristics, their communication patterns — and whether regulators have any framework for what that means.