A Critical Review of “The Illusion Of thinking” by Apple

A recent study by Apple evaluated the reasoning capabilities of Large Reasoning Models (LRMs) through the use of puzzle-based tasks to allow controlled variation in problem complexity. The study concluded that reasoning performance in LRMs collapses beyond a certain complexity threshold. This review provides an explanatory framework for those findings, grounded in the Algorithm Conjecture and the Dynamic Algorithms/Consciousness system introduced in prior work.