The post-ChatGPT learning-time decline follows an age gradient: largest in high school (31.3% cumulative), small in middle school (9.0%), and undetectable in Grade 5¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? High school topics show a 3.35% per-quarter decline (31.3% cumulative) and middle school a smaller significant 0.86% per-quarter decline (9.0%). → Rismanchian 2026 q2 i? Grade 5 courses show no detectable change in the differential learning-time trend. → Rismanchian 2026 (2)

Evidence¶

Rismanchian 2026¶

Rismanchian, S., Uzun, H., Matayoshi, J., Cosyn, E., & Kurd-Misto, E. (2026). Faster Completion, Less Learning: Generative AI Reduced Study Time on Math Problems and the Knowledge They Build. arXiv preprint. https://arxiv.org/abs/2605.21629 doi:10.48550/arxiv.2605.21629

q2 · i? · causal · r2

Same topic-by-quarter ramp models estimated separately by age group on ALEKS learning interactions. High school "β=− 0.0341, p< 0.001" (31.3% cumulative); middle school "β=− 0.0086, p< 0.01" (9.0% cumulative).

"High school topics show a larger effect of 3.35% per quarter (β=− 0.0341, p< 0.001), corresponding to a 31.3% cumulative decline (CI:−37.3%, −25.4%). Middle school courses show a smaller but statistically significant effect of 0.86% per quarter"

Rismanchian 2026 (2)¶

Rismanchian, S., Uzun, H., Matayoshi, J., Cosyn, E., & Kurd-Misto, E. (2026). Faster Completion, Less Learning: Generative AI Reduced Study Time on Math Problems and the Knowledge They Build. arXiv preprint. https://arxiv.org/abs/2605.21629. doi:10.48550/arxiv.2605.21629

q2 · i? · causal · r2

Grade 5 ramp estimate on ALEKS learning interactions is null: "β=− 0.0012, p= 0.79". The authors attribute the gradient to older students' greater autonomy and unmonitored digital access.

"Grade 5 courses show no detectable change (β=− 0.0012, p= 0.79)."

Discussion¶

Related Claims¶

• College students' learning time on AI-susceptible word problems declines 2.80% per quarter relative to graph-based problems after ChatGPT's release, cumulating to 26.9% over eleven quarters — related
• The post-ChatGPT response-time divergence on AI-susceptible items vanishes entirely under proctored assessment — related
• On randomly assigned proctored retention items, odds of correct response on AI-susceptible items decline 25.0% cumulatively after ChatGPT's release — related
• Workers ages 22 to 25 in the most AI-exposed US occupations experienced a 16% relative employment decline following widespread generative AI adoption — related