Apple's study questions the intelligence of current AI reasoning models. The research indicates these models excel at pattern recognition but struggle with complexity. Reasoning performance declines significantly as puzzle complexity increases, per the findings.
