Japan's Bio Search agrees to buy InnerEye, whose AI learns expert decisions from brain signals
The Israeli company reads EEG in real time to capture how experts decide, and trains models on it. The share-exchange deal should close by early January.
DevIsDead Newsroom
An AI that learns from how an expert's brain reacts, not just what they click, is changing hands. Japan's Bio Search has signed a share-exchange agreement to acquire all of InnerEye, an Israeli company whose AI interprets brain activity as it happens, according to Unite.AI, whose article was written by an AI research agent and reviewed by its editors. The deal hasn't closed; it needs approvals including from Israeli authorities and is expected between December and early January.
What InnerEye does
Its Minded A.I. workflow has an expert wear a device while working; the system turns aspects of their decisions into data that teaches models to reproduce the expert's decision patterns, and also helps assess how much to trust those decisions. Sense.I applies this to training and personalizing AI; SensePlus targets the cognitive and emotional side, detecting fatigue, concentration and attention to reduce human error. Its algorithms read electroencephalogram (EEG) signals in real time to distinguish cognitive states during perception and decision-making.
InnerEye was founded in 2014 by Ben-Gurion University engineering professor Amir Geva and Hebrew University brain scientist Leon Deouell.
Bio Search's plan
Bio Search already uses EEG, ECG, heart rate and muscle signals in healthcare AI, with sensing and medical-device work in Japan and Asia. The plan is a shared platform pooling both companies' biological data to train models, with targets that range from mapping cognitive function and catching disease early to factory automation, emotion-aware computing, security and helping experts decide. It hasn't quantified expected gains. InnerEye CEO Uri Antman called it the ideal partnership to take the technology to global scale. Bio Search is targeting a Nasdaq listing in 2028.
The questions
Training AI on experts' brain activity is a striking idea for capturing tacit knowledge that people can't put into words. It's also some of the most sensitive data there is, and the deal's value depends on whether pooled brain data actually yields reliable models for that long list of uses.
Scoreboard update: AI +1. AI learns to copy expert judgment straight from their brain signals.
Stay undead.
Comments 0