Apple researchers trash AI 'reasoning' models on the eve of WWDC
"Through extensive experimentation across diverse puzzles, we show that frontier LRMs face a complete accuracy collapse beyond certain complexities." — Apple Machine Learning
Philip Elmer-DeWitt has been covering Apple since 1983 — mostly for Time Magazine (28 years), later for Fortune (9 years), where he wrote a daily blog called Apple 2.0. [Read more.]