AI assistance reduces subjective mental effort across all tasks even when it does not reduce completion time, dissociating time and effort¶

Claim · All claims
Evidence · 1 study · causal r3 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? NASA-TLX effort was lower with AI by 0.61 points on average for all tasks, while completion times were mostly unchanged; time and NASA-TLX correlated weakly (r=0.26). → Sunny Yu 2026

Evidence¶

Sunny Yu 2026¶

Sunny Yu, Myra Cheng, Ahmad Jabbar, Ilia Sucholutsky, Katherine M. Collins, Dan Jurafsky, Robert D. Hawkins. (2026). Cognitive offloading and the speedup illusion in human-AI interaction. arXiv. https://arxiv.org/abs/2605.23177 doi:10.48550/arxiv.2605.23177

q3 · i? · causal · r3

Completion-sample participants answered the 5-question NASA-TLX after each task. The article reports "it reduced mental effort for all tasks", a 0.61-point reduction on the 7-point scale, significant for 15 of 24 tasks, while completion time and NASA-TLX showed only a weak positive correlation (r=0.26).

"while AI assistance sped up completion of only a few tasks, it reduced mental effort for all tasks — on average, the NASA-TLX average (higher means more effortful) lowered by 0.61 points on a 7-point scale (SE=0.059,t=−10.38,p<0.001)."

Discussion¶

Learner Variables¶

• Time and Continuity — outcome: instruction changes it

Related Claims¶

• Prompting is a major source of AI-assisted cognitive effort: copying prompts reduces effort but not time, and verbose model responses can make AI-assisted completion slower than predicted — related
• Personalized paths reduced self-reported cognitive load on all six adapted NASA-TLX dimensions with large effect sizes — related
• GATs produced no reliable immediate learning-experience effects on frustration, cognitive load, or situational interest at either institution — related
• Fluent Illusions Mislead Self Assessment — related
• AIfred reduces observed physical-digital context switches by 98% relative to screen-based assistance, while task completion time is longer — related
• AI assistance sped up only difficult tasks and only a few individual tasks, not easy ones — related
• Behavioral interaction metrics are weakly or inconsistently correlated with task performance in AI-assisted programming — related
• No AI participants spent roughly 1.5 times as long on the assessment as AI-Assisted participants, descriptively reported as a marker of offloading — 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
• 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
• The Bloom classification UI shows low perceived workload and high usability in a 50-participant user study — related
• People overestimate how much mental effort AI assistance alleviates (offloading illusion), driven by overestimating the effort of independent completion while AI-assisted effort is accurately estimated — related
• AI overviews in an academic search engine showed mixed effects: summaries may reduce mental demand and frustration for some users, with slightly fewer clicks and query reformulations — related
• Effects of GenAI availability on self-reported understanding are insignificant; effects on interest are significant only under the transient-COVID assumption — related
• AI assistance reduces task completion time in both education groups, the only outcome without an equalizing pattern — reports the opposite
• Interest in AI for feedback in higher education is driven by workload reduction and feedback timeliness amid massification — related
• In a 30-participant user study, proactive paradigms are preferred over user-initiated interaction, with lower workload, fewer window switches, shorter editing time, and higher acceptance — related
• Training on AI interaction strategies slightly increases cognitive load during AI-assisted design — related
• Layered AI feedback was perceived as more encouraging and independence-supporting but imposed higher mental effort — reports the opposite
• Participants with higher subject expertise completed tasks and used the tools more easily than less experienced participants — related
• In simulation, AI-assisted students showed increased scores and reduced solving time, attributed to personalized assistance — related
• GenAI access reduces perceived difficulty of the assisted reasoning task — related
• The AI+Robot condition is associated with lower perceived workload than AI-Only on effort, frustration, temporal demand, physical demand, and Raw TLX — a narrower finding that bears on this claim
• No differences were found on mental effort during instruction or on transfer across illusion-of-control and instruction-method conditions — related
• Proactive feedback produces significant post-intervention increases in Joint Mental Effort of dyads — related
• AI chatbot access lowers teachers' perceived cognitive effort during learning design — a narrower finding that bears on this claim
• Mental effort and study time increase with a topic's element interactivity level, making them indicators of element interactivity — related
• Learners reported higher perceived learning and satisfaction and lower mental effort with the instructor video, despite no performance gains — a perception-performance gap — a narrower finding that bears on this claim
• Self-reported mental effort increased less across tasks in the ChatGPT condition while perceived difficulty rose equally in both groups — related
• Proactive multimodal feedback significantly reduces debugging time on task compared to control — related
• Performance incentives increase time invested in the assisted reasoning task — related
• 30 of 32 participants preferred the AI+Robot condition, describing the robot as making success and failure visible, though a minority found AI-Only useful for focusing on block construction — related
• AIfred users report higher perceived learning support, innovation, satisfaction, and cognitive demand, with no difference in perceived productivity — related
• Users alter expectations and behavior based on perceived AI capabilities even when actual AI performance is unchanged (the placebo effect of AI) — related
• No single control-timing paradigm dominates across subjective measures: eight of ten questionnaire items show no significant differences between paradigms — related
• Motivational interventions can lower perceived cognitive load and improve retention and transfer, indicating load perception is partly independent of task difficulty — related
• Studies collecting both subjective and objective effort measures reveal dissociations between the two types of measures — related
• Recent AI-in-education studies often rely on subjective or performance-during-support measures and weak comparison designs, limiting claims about learning effects — related
• The overcoming-barriers intervention does not reduce perceived time-and-effort cost of retrieval practice at either time point — related