AI Makes People Reluctant to Say "I Don't Know"
rohanpaul_ai · x · 2026-07-18
This study found that AI significantly reduces people's willingness to say "I don't know."
Across 5 experiments with 3,132 participants:
- Without AI, participants withheld judgment on 44% of questions;
- With AI advice, this dropped to 3%;
- Even when given intentionally incorrect AI advice, accuracy fell from 27% to 9%, yet self-confidence increased.
The authors conclude that AI impacts not just answer accuracy, but also the metacognitive threshold for deciding if one "knows enough." Encouraging accuracy and penalizing errors only partially mitigate this effect.
Related event: AI Advice Lowers Accuracy but Boosts Confidence(2 posts)→
More from Research
- MaP-WAM tackles non-Markovian robot manipulation with memory-grounded planning — Sizhe Zhao · 2026-09-11
- Negative Self-Distillation improves LLM reasoning by avoiding flawed reasoning paths — Rongcan Pei · 2026-09-11
- DeepMind-led paper makes design docs the source of truth, code disposable — SMART regenerates in 1.5-3h for ~$100 — Roger_M_Taylor · 2026-09-11
- GameWorld wins Best Paper Runner-Up at ECCV 2026 Multimodal Digital Agents Workshop — MikeShou1 · 2026-09-11
- Yann LeCun live at ECCV on World Models — Weak_Assistance_5261 · 2026-09-11
- 3D ResNet Paper Crosses 3,000 Citations Eight Years After CVPR 2018 — HirokatuKataoka · 2026-09-11