Theoretical Connotation and Cultivation Pathway of 'Human-AI Collaborative Mathematical Literacy' in the Context of Artificial Intelligence

Artificial Intelligence (AI) is profoundly reshaping mathematical practice, revealing limitations in traditional individual-centered mathematical literacy. This paper proposes a new paradigm of 'Human-AI Collaborative Mathematical Literacy' (HACML) based on distributed cognition theory. The necessity stems from four challenges: overcoming biological cognitive limitations, adapting to AI-driven workflow restructuring, mitigating algorithmic ethical risks, and bridging educational gaps. HACML encompasses four core competencies: critical interaction, meta-tool cognition, collaborative decision-making, and ethical judgment, shifting the evaluation unit from individuals to 'human-AI system' collaborative efficacy. Cultivation pathways include: innovating three teaching principles (AI transparency, contextual immersion, reflective practice); designing implementation strategies for four-dimensional capabilities; and transforming teachers into 'cognitive ecology designers'. This provides a systematic framework for mathematics education in the AI era.