 | A new study from UCLA, MIT, Oxford, and Carnegie Mellon gave 1,222 people AI assistants for cognitive tasks — then pulled the plug midway through. The results: - After ~10 minutes of AI-assisted problem solving, people who lost access to AI performed **worse** than those who never had it - They didn't just get more wrong answers — they **stopped trying altogether** - The effect showed up across math AND reading comprehension - Ran 3 separate experiments (350 → 670 → full cohort). Same result every time. The researchers call it the "boiling frog" effect — each AI interaction feels costless, but your cognitive muscles are quietly atrophying. The UCLA co-author warns this could create "a generation of learners who will not know what they're capable of." Study hasn't been peer-reviewed yet, but the sample size is solid and it's the first causal (not correlational) evidence of AI-induced cognitive decline. The uncomfortable question: if 10 minutes is enough to measurably damage independent performance, what does months of daily use do? Full breakdown → https://synvoya.com/blog/2026-04-20-ai-boiling-frog-cognition-study/ Be honest — have you noticed yourself giving up faster on problems since you started using AI daily? [link] [comments] |
Predictive technology: computational social science
