R0057/2026-04-01/C031/H1

Research	R0057 — RLHF Yes-Men Claims v3
Run	2026-04-01
Claim	C031
Hypothesis	H1

Statement

Users apply zero critical thinking to 40% of AI tasks

Status

Current: Supported

Supporting Evidence

Evidence	Summary
SRC01-E01	40% of AI-assisted tasks had zero self-reported critical thinking; 319 knowledge workers surveyed

Contradicting Evidence

Evidence	Summary
—	No contradicting evidence found

Reasoning

Microsoft Research/CMU surveyed 319 knowledge workers who shared 936 first-hand examples of GenAI use. For 40% of tasks, participants reported using no critical thinking whatsoever. Higher confidence in GenAI correlated with less critical thinking. Published at CHI 2025.

Relationship to Other Hypotheses

H1 represents full accuracy. H2 allows for partial correctness. H3 is eliminated by the evidence.