According to @milesdeutscher, an MIT experiment scanned brain activity via EEG in 54 adults across LLM-assisted, search-assisted, and brain-only writing tasks over multiple sessions, source: @milesdeutscher citing an MIT study. The LLM group exhibited the weakest neural connectivity, with activity falling from 79 to 42 (a 47% decrease) while completing tasks 60% faster, highlighting a productivity-versus-cognition trade-off, source: @milesdeutscher citing an MIT study. The study reported that 83% of ChatGPT users could not recall what they had written minutes earlier, and researchers described the effect as cognitive debt, source: @milesdeutscher citing an MIT study. Design details included three cohorts of 18 participants each, EEG-monitored 20-minute essay sessions across 3–4 visits, a Session 4 crossover (LLM-to-brain and brain-to-LLM), and post-task interviews on quoting, ownership, satisfaction, and ethics, source: @milesdeutscher citing an MIT study. For traders, these findings question assumptions around blanket LLM productivity and may sharpen scrutiny on AI monetization narratives tied to AI-exposed equities and AI-linked crypto tokens such as FET (ASI) and RNDR, source: @milesdeutscher citing an MIT study. Watch for shifts in corporate AI-usage commentary, user-engagement metrics, and narrative momentum around AI tokens following the study’s visibility, using the study’s results as the catalyst context, source: @milesdeutscher citing an MIT study.