Prior AI use carries over: brief AI exposure increases subsequent AI adoption and exacerbates the speedup illusion, while independent exposure does not reduce AI use¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r2 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? Participants exposed to AI in an exposure phase chose AI on 44.5% of subsequent easy tasks versus 27.7% for those who completed tasks independently. → Sunny Yu 2026 q3 i? AI exposure lowered agreement that tasks are faster to complete independently, worsening time calibration; effort calibration showed no significant effect. → Sunny Yu 2026 (2)

Evidence¶

Sunny Yu 2026¶

Sunny Yu, Myra Cheng, Ahmad Jabbar, Ilia Sucholutsky, Katherine M. Collins, Dan Jurafsky, Robert D. Hawkins. (2026). The efficiency-gain illusion: People underestimate the rate of AI use and overestimate its benefits on simple tasks. arXiv. https://arxiv.org/abs/2605.22687 doi:10.48550/arxiv.2605.22687

q3 · i? · causal · r2

Pre-registered Study 3 (N=592) randomly assigned participants to complete two exposure tasks with AI, independently, or no tasks (control), then measured AI adoption on two easy test tasks. The effect held for both easy and difficult exposure variants.

"Participants who completed tasks with AI during the exposure phase chose AI assistance on 44.5% of subsequent tasks, compared to 27.7% among participants who completed the exposure tasks independently (β= 0.54,p <0.001)."

Sunny Yu 2026 (2)¶

Sunny Yu, Myra Cheng, Ahmad Jabbar, Ilia Sucholutsky, Katherine M. Collins, Dan Jurafsky, Robert D. Hawkins. (2026). The efficiency-gain illusion: People underestimate the rate of AI use and overestimate its benefits on simple tasks. arXiv. https://arxiv.org/abs/2605.22687

q3 · i? · causal · r2

Study 3 measured time and effort calibration via 7-point agreement ratings after the test phase. AI exposure worsened time calibration but did not significantly affect effort calibration; qualitative analyses indicated time calibration drove choices.

"those who had previously completed tasks with AI reported lower agreement with the statement “It was faster to complete the tasks independently than using AI”. We reverse coded the value to operationalize the speedup illusion (β=−0.34,p<0.001; Fig. 4). We found no significant effect of previous completion mode on effort calibration (β= 0.08,p= 0.28)."

Discussion¶

Related Claims¶

• AI assistance provides no significant overall time savings and can slow completion on easy task variants, with chat-interface friction (prompting time) dominating on trivial tasks — related
• People use AI on simple tasks at a rate significantly higher than they predict for themselves (self-estimate miscalibration), especially on easy task variants — related
• People significantly underestimate AI-assisted completion times on simple cognitive tasks, even though AI assistance saves little or no actual time (the speedup illusion) — related
• Participants more averse to thinking are more susceptible to the speedup illusion; AI familiarity measures do not predict calibration error — related