Teen AI Finds 1.5 Million Hidden Space Objects in NASA Archive

Teen Matteo Paz used an AI-driven approach (VARnet) to analyze NASA's NEOWISE data and automatically identify about 1.5 million previously unidentified celestial objects, processing roughly 200 billion entries; he published a paper in The Astronomical Journal and won the Regeneron Science Talent Search, with Caltech planning to apply the method to binary-star studies and noting potential broader uses beyond astronomy.