MIT EEG Study: Heavy LLM Use Reduces Neural Connectivity by up to 55%
An MIT EEG study comparing 54 participants across three conditions—LLM-assisted, search-engine, and no-tool writing—over 4 months found heavy ChatGPT users showed up to 55% reduced neural connectivity. Brain-only writers showed the strongest neural networks. 83% of heavy LLM users couldn't quote essays they had just written with AI assistance. Notably, when brain-only writers later switched to AI, their connectivity increased, suggesting sequential use is healthier than starting with AI.
Why It Matters
This is the first neurological evidence for "cognitive debt" from AI writing assistance, quantifying what had been intuited. The "AI as finishing tool, not starting point" pattern is now empirically supported.