Trevor Paglen Warns AI Brain Decoding Erodes Inner Privacy

Artist Trevor Paglen describes a two-day fMRI experiment at Cornell University in which an AI system called NeuroGen predicted his brain responses and generated images designed to trigger specific brain activity, leaving him feeling that his mind had been read. He links this to a research line running from Jack Gallant's blurry 2011 video reconstruction to Alex Huth's LLM-based decoding of brain activity into text. Paglen also notes that Meta paid $500 million to $1 billion for a wristband startup that turns brain signals into computer commands. Rafael Yuste, who used lasers to make mice see phantom images, co-founded the Neurorights Foundation to press for legal safeguards around brain data. The account is Paglen's own description of the experiment and the field's history; it does not establish that the system read his thoughts independently of the fMRI setup.

Generated from reports · updated 12 h ago

Not enough continuous observations yet to draw a trend.