Generativism: Toward a Learning Theory for the Age of Generative Artificial Intelligence

Existing learning theories, including behaviorism, cognitivism, constructivism, and connectivism, are conceptually limited in the context of proliferating generative artificial intelligence (AI) in educational settings. These traditional frameworks predate AI systems capable of generating, synthesizing, and reasoning about knowledge. A new learning theory, 'Generativism,' is proposed, integrating research from distributed cognition, human-AI collaboration, AI literacy, and metacognition, to address these limitations and better account for learning in the generative AI era.

Why it matters

The rapid advancement and integration of generative AI necessitate a re-evaluation of fundamental learning paradigms across all sectors. Organizations must understand how human learning is evolving when supported by AI, as this impacts skill development, knowledge transfer, and operational effectiveness. A revised theoretical framework can guide strategy for workforce development, educational programs, and the design of AI-supported learning environments.

Key insights

• Traditional learning theories (behaviorism, cognitivism, constructivism, connectivism) face significant conceptual limitations as generative AI becomes widespread in education.
• These established frameworks were developed prior to the advent of AI systems capable of knowledge generation, synthesis, and reasoning.
• Generative AI challenges fundamental assumptions embedded within current learning theories.
• The proposed 'Generativism' theory integrates concepts from distributed cognition, extended mind, human-AI collaboration, AI literacy, cognitive offloading, and metacognition.
• Generativism aims to provide a suitable learning theory for the age of generative artificial intelligence.

Source

arXiv — Computers and Society — https://arxiv.org/abs/2606.12441