yea it's clear no one is actually reading the paper. the study showed the group who used LLMs for the first three sessions then had to do session 4 without them had lower brain connectivity than was recorded for session 3 with all the groups showing some kind of increase from one session to the next. Importantly, this group's brain connectivity didn't reset to the session 1 levels, but somewhere in-between. They were still learning and getting better at the essay writing task. In session 4 they effectively had part of the brain network they were using for the task taken away, so obviously there's a dip in performance. None of this says anyone got dumber. The philosophical concept of the Extended Mind is key here.

imo the most interesting result is that the brains of the group that had done sessions 1-3 without the search engine or LLM aids lit up like christmas trees in session 4 when they were given LLMs to use, and that's what the paper's conclusions really focus on.