There is a growing body of work on the effects of large language models on the brain, from cognitive offloading, to direct measures of brain activity. While all of this work is interesting, the latter is striking, with electroencephalography (EEG) used to assess cognitive load while writing an essay.
The result? EEG shows LLM use reduces brain connectivity, task engagement, memory recall, and ownership versus search and unaided effort. And the effects persisted after the task was over.
In the following graphic, I put the reduced cognitive activity in context, comparing it to what is observed when people have consumed too much alcohol. In rough terms, ChatGPT suppressed brain activity as much as being twice the legal limit.
Granted, brain activity is a gross measure, and alcohol's effects are broad, but the point holds: There is sharply reduced brain activity when using LLMs, and that has consequences beyond the activity itself. Chat model usage can be like driving your brain drunk.