Sabi says EEG can tune AI-generated faces to personal taste

Sabi co-founder and CEO Rahul Chhabra said in an October 29 post that Sabi used EEG-derived arousal signals to steer AI-generated images toward one person's aesthetic preferences, a method he called reinforcement learning with neural feedback (RLNF), showing a face changing as its viewer watched. Chhabra said the approach personalizes generated images without explicit ratings, though the accessible research text offers limited detail for assessing the results. The claim matters because it would make taste a new feedback channel for generative models while raising consent and interpretation questions around neural data.

Sabi says EEG can tune AI-generated faces to personal taste Sabi CEO Rahul Chhabra says EEG-derived arousal signals can personalize generated images without explicit ratings. Accessible research text offers limited detail for assessing the results. By RuntimeWire Staff https://runtimewire.com/author/runtimewire-staff · Published Primary source: Aligned News — AI Intelligence https://x.com/rahulchhabra07/status/2115918744308658325 Why it matters Sabi is testing whether brain signals can personalize AI outputs without repeated ratings. That could make taste a new feedback channel, while putting consent and interpretation of neural data at the center of the product. Rahul Chhabra @rahulchhabra07 https://x.com/rahulchhabra07/status/2115918744308658325?ref=runtimewire , co-founder and CEO of Sabi https://sabi.com/?ref=runtimewire , says Sabi has used brain signals to steer AI-generated images toward one person's aesthetic preferences. In a post published on October 29th, Chhabra introduced the method as reinforcement learning with neural feedback, or RLNF, and showed a face changing as its viewer watched. When you purchase through links in our articles, we may earn a small commission https://runtimewire.com/affiliate-monetization-standards . This doesn’t affect our editorial independence.