Using LLMs daily makes your own reasoning feel weaker — two studies agree, causation unproven
Dangerous_Ad7101 · reddit · 2026-08-19
The poster shares that with daily LLM use, they offload the effortful parts — problem decomposition, argument building, phrasing — and find working without a model harder than it should be.
Two cited studies point the same way: MIT Media Lab (Kosmyna et al. 2025) found reduced EEG connectivity, worse recall of one's own text, and lower sense of ownership during LLM-assisted essay writing; Gerlich (2025, Societies) found a negative correlation between frequent AI use and critical thinking scores, mediated by cognitive offloading. Neither proves long-term causal damage.
The thread asks the community which tasks they deliberately avoid offloading and which routines actually preserve their own thinking performance.
Related event: Does Heavy LLM Use Weaken Reasoning? Studies Lend Support(2 posts)→
More from AGI Musings
- Tech workers think they're safer from AI than everyone else — they're more at risk, argues dev — thisdudelikesAI · 2026-10-17
- Fields Medalist Villani: math went from least AI-affected science to most affected in one year — robleclerc · 2026-10-17
- Prediction: human surgeons performing life-saving surgery illegal in 15 years — imjustnewatai · 2026-10-17
- Dev scrapes all 43,013 ICLR 2027 submissions and builds a citation-ranking site with Codex and Claude — yshan2u · 2026-10-17
- Scobleizer visits TheHumanoidHub: robots for devs this year, everyone else after 2030 — Scobleizer · 2026-10-17
- The Future of Media: People With Direction vs. People Generating Noise — AIandDesign · 2026-10-17