Agentivism generates six empirically testable propositions linking delegation structure, verification, reconstruction, and reduced-support transfer to learning outcomes¶

Claim · All claims
Evidence · 1 study (2 entries) · theoretical r3 · q1
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q1 i? Learning should be stronger when AI support preserves learner responsibility for problem framing, criteria setting, and justification than when AI support collapses these processes into direct answer delivery. → Lixiang Yan and Dragan Gašević 2026 q1 i? Repeated low-friction delegation without subsequent reconstruction should be associated over time with weaker calibration of one's own competence and greater dependence on external support, even when short-term productivity rises. → Lixiang Yan and Dragan Gašević 2026 (2)

Evidence¶

Lixiang Yan and Dragan Gašević 2026¶

Lixiang Yan and Dragan Gašević. (2026). Agentivism: a learning theory for the age of artificial intelligence. https://arxiv.org/abs/2604.07813 doi:10.48550/arxiv.2604.07813

q1 · i? · theoretical · r3

A theoretical proposition derived from the delegated agency mechanism in the Empirically testable propositions section. The article states that "learning should be stronger when AI support preserves learner responsibility for problem framing, criteria setting, and justification than when AI support collapses these processes into direct answer delivery". This is a prediction, not a reported result.

"A first proposition is that learning should be stronger when AI support preserves learner responsibility for problem framing, criteria setting, and justification than when AI support collapses these processes into direct answer delivery"

Lixiang Yan and Dragan Gašević 2026 (2)¶

Lixiang Yan and Dragan Gašević. (2026). Agentivism: a learning theory for the age of artificial intelligence. https://arxiv.org/abs/2604.07813

q1 · i? · theoretical · r3

The sixth proposition, a longitudinal prediction from the theory. The article states that "repeated low-friction delegation without subsequent reconstruction should be associated over time with weaker calibration of one's own competence and greater dependence on external support, even when short-term productivity rises". The article itself calls this prediction more plausible than fully established.

"repeated low-friction delegation without subsequent reconstruction should be associated over time with weaker calibration of one's own competence and greater dependence on external support, even when short-term productivity rises"

Discussion¶

Related Claims¶

• Falsifiable hypothesis: unrestricted AI use on deep-processing tasks will produce a product–process dissociation—higher-rated assignments but lower unaided delayed transfer — a narrower finding that bears on this claim
• Students did not consistently detect scientific errors in ChatGPT responses, and students with stronger prior knowledge evaluated AI responses more critically — related
• Review reports a field experiment in which GPT access improved supported practice but was followed by poorer unaided test performance, mitigated by a guarded tutor — a narrower finding that bears on this claim
• Product-outcome studies report GenAI-related gains on product-quality measures but do not assess durable, transferable learning — related
• Unstructured ChatGPT use in programming education is linked to overreliance, reduced persistence, unreliable outputs, and academic integrity risks — related
• Immediate task success does not establish learning that persists without AI support — a broader claim this one bears on
• Proposition 1: outcome-aligned ownership performances will predict later individual transfer beyond AI-assisted product quality alone — a narrower finding that bears on this claim
• Proposition 2: mutual evaluation emphasizing reasons and responsive comparison will improve evaluative judgment more than unexplained self-rating — related
• AI-directed problem selection may diminish the metacognitive dimension of self-testing while preserving retrieval practice — a narrower finding that bears on this claim
• Evidence on GenAI in chemistry education is concentrated in short-term outcomes, with durable transfer and retention rarely tested — related