What does AI do to the brain of young people?

What does AI do to the brain of young people? That question does occupy me at times, and I'm clearly not the only one. 😊

We know that the brain of children and young people is still very much developing, so what they do shapes their neural wiring. In the media a fair number of bold claims are made about AI's influence, but long-term research into that effect simply doesn't exist yet.

There are, however, indications and studies that can already say something about AI and brain development. It's important to look closely at what kind of AI use was studied and how long young people use it. Jonathan Haidt distinguishes, for screen use in general, between fragmenting time (short, fragmented attention) and storytelling time (longer, meaningful focus). That distinction, I think, also helps with AI.

MIT research showed that students who used ChatGPT directly during a writing assignment showed less brain activity in areas for planning and reasoning. Students who first thought it through themselves and only then used AI performed better and retained more cognitive activity. A small but carefully conducted experiment.

Wilfred Rubens emphasises the difference between support and replacement. If AI takes over the real thinking work, learning gains decrease. If AI helps with organising or understanding, thinking can actually deepen.

Erik Scherder is critical (he calls AI an abbreviation for "diminishing intelligence" 😅). But if you look closely at his examples, they're mostly about screen behaviour that clearly falls into fragmenting time. That makes his concerns less directly applicable to AI use that leans more toward storytelling time, such as collaborating, analysing or in-depth writing.

So the influence on the brain seems to depend on exactly what young people do with AI. But in my view, firm conclusions can only be drawn once real long-term research arrives.