ai-making-us-stupid โ Cash et al. (2026) Trends Cog Sci: offloading cognition to AI impedes skill acquisition & accelerates skill decay; risks use-dependent; basic abilities possibly more resilient
llm-cognitive-diagnosis-handwritten-math โ MathCog benchmark: 18 LLMs evaluated on cognitive skill diagnosis from handwritten math; all F1 < 0.5; systematic over-attribution and hallucination of evidence (2025)
ai-engineering-education-balancing-act -- Students attachment to LLM efficiency reflects 'cruel optimism' - benefits depend on critical thinking skills they are still developing