In 2011 a psychologist at Columbia named Betsy Sparrow sat people down and had them type a run of trivia facts into a computer. An ostrich's eye is bigger than its brain, that sort of thing. She told half the group the computer would save their work. She told the other half it would be wiped. Then she asked everyone to write down as many facts as they could remember.
The people who thought the file was saved remembered far fewer. Their heads had already made the decision without asking them. No point holding something the machine is holding for you.
Sparrow published it in Science and called it the Google effect. The idea underneath was older than Google. Her mentor Daniel Wegner had named it back in 1985, transactive memory, the way a couple who've been together thirty years split the load. One remembers the birthdays. The other remembers where the spare fuses are. Neither carries the whole thing, and together they recall more than either could alone. We've always done this with people we trust. Now we do it with a search bar.
And it costs something you can measure in tissue.
In 2000, before anyone had a maps app, Eleanor Maguire scanned the brains of London cab drivers. To get licensed in London you sit a test they call The Knowledge, memorising twenty-five thousand streets and how they knit together, and it takes years. Maguire found these drivers had visibly more grey matter in the back of the hippocampus, the bit that handles spatial navigation, than ordinary people. The longer they'd driven, the bigger it was. They'd grown the muscle by using it.
Then the sat nav arrived. In 2017 a team at University College London put people in a scanner and had them find their way through a virtual city. When the volunteers worked out their own route, the hippocampus lit up, running the streets and weighing the options. When they followed turn by turn directions from a device, that activity went quiet. The part of the brain that builds the map switches off when something else is holding the map for you. Same trade Sparrow found, written in navigation instead of trivia.
None of this is new panic. Nicholas Carr wrote "Is Google Making Us Stupid?" in The Atlantic in 2008 and got called a luddite for it. Socrates reckoned writing would ruin memory, and in a narrow way he was right. Ask anyone under forty to recite a phone number they didn't have before mobiles. We gave that up on purpose and got something back for it.
So a tool takes a job off you and your brain reallocates. Fine. That's the deal we've been signing for a long time. What changes with AI is the job on the table. Memory was storage. Navigation was a lookup. The thing sitting in front of us now does the reasoning step, the part where you decide what actually matters. That's a different order of muscle to hand over.
I build AI into small businesses for a living. Cafes and trades, the kind of shop where the owner still does the books at 9pm. The thing I watch for is which part of their thinking they hand across. The quoting maths, the roster juggling, the stock reorder, the first draft of a customer email, hand all of that over. It's the fuse box. You were never precious about it and your day gets better without it.
The judgement is the part to keep. The second an owner cant tell whether the AI's answer is right or rubbish, they've given away the muscle that mattered, and it wastes away the same way the cabbie's does when he stops driving. Slow and quiet, and you dont notice till you need it.
Refusing the tool is how you lose, so that's off the table. The whole game is staying deliberate about what lives in your own head and your own hands. Sparrow's volunteers didn't come out stupider. They came out very good at knowing where to look, and that turned into a real skill of its own. Pick the skill you want to be good at in ten years and keep doing it by hand often enough that it stays.