Is AI Making Us Stupid? What the Calculator Panic Gets Right — and Wrong

Photo: Markus Spiske · Pexels

Around 370 BC, Socrates warned that a dangerous new technology would ruin human memory. People would stop exercising their minds, he argued, because they could simply look things up. The technology in question? Writing. 📜

Twenty‑four centuries later, we had the same panic about pocket calculators, then about Google, and now — louder than ever — about AI. So it’s worth asking the question honestly: is AI going to make us stupid, or is it just the calculator all over again — a tool that didn’t dull us, but freed our brains for work that actually matters?

The honest answer, as usual, is more interesting than either headline.

🧮 The calculator precedent: the panic that didn’t come true

When calculators entered classrooms in the 1970s, the fear was visceral: a generation that couldn’t do arithmetic. Educators predicted mathematical illiteracy at scale.

It didn’t happen. The most comprehensive research on the topic — a meta‑analysis of 79 classroom studies by Hembree and Dessart — found that calculator use, alongside traditional instruction, didn’t harm students’ maths skills and in most cases improved both achievement and attitude towards the subject.

Why? Because the calculator removed the grunt work, not the thinking. Long division by hand is mostly mechanical procedure — low‑value cognitive labour. Once machines absorbed it, curricula moved up the stack: more algebra, more statistics, more problem‑solving, earlier. We stopped training humans to be slow, error‑prone calculators and started training them to do things calculators can’t.

That’s the optimistic template: offload the mechanical, reinvest the brainpower. And it’s largely the story of every cognitive tool we’ve adopted, from the abacus to the spreadsheet.

📱 We’ve been offloading for a while

The calculator wasn’t a one‑off. Research by Sparrow, Liu and Wegner — the famous “Google effect” study published in Science — showed that when people know information will be available online, they remember the information itself less, but get better at remembering where to find it. Nobody memorises phone numbers anymore; nobody navigates by memory. Our brains adapted from storing to indexing.

Nicholas Carr asked “Is Google Making Us Stupid?” in The Atlantic back in 2008. Nearly two decades on, the fair verdict is: it changed how we think more than how well we think. Different, not dumber.

So far, so reassuring. But here’s where the calculator analogy starts to wobble.

⚠️ Why AI is not quite the calculator

A calculator can’t write your essay, form your argument, or make your decision. AI can — and that’s precisely the new risk the research is starting to map.

An MIT Media Lab study made headlines with EEG scans of people writing essays with and without ChatGPT. Participants who relied on the AI showed weaker brain connectivity during the task, remembered less of what “they” had written, struggled to quote their own essays, and felt less ownership of the work. The researchers called the effect “cognitive debt” — convenience now, paid for with engagement later. (Worth noting: it was a small study of 54 people writing essays in a lab, not proof of permanent brain damage — despite what the viral headlines claimed.)

Meanwhile, a Microsoft Research and Carnegie Mellon survey of 319 knowledge workers found a telling pattern: the more people trusted the AI, the less they thought critically about its output. The reverse was also true — people confident in their own skills engaged more critically with AI, using it as a sparring partner rather than an oracle.

The calculator never tempted anyone to stop thinking. AI does — politely, fluently, and on demand.

Read both studies carefully, though, and neither says “AI makes you stupid.” They say something more precise: passive use disengages the brain; active use doesn’t. The MIT participants who wrote first and used AI afterwards showed healthy engagement. The knowledge workers who interrogated AI output kept their critical thinking intact.

⚖️ The real variable isn’t the tool — it’s the driver’s seat

This is the heart of it. The same tool produces opposite outcomes depending on which of these describes you:

• Use that sharpens you: drafting your own thinking first, then using AI to challenge it; asking it to argue against you; verifying claims; using it to learn faster and attempt harder problems than you could alone.
• Use that dulls you: outsourcing the first draft and the judgement; accepting outputs unread; letting it decide instead of inform; using it to avoid ever being confused — because confusion is where learning happens.

The calculator forced this discipline on us, because it could only do one narrow thing. AI doesn’t force anything. It will happily do as much of your thinking as you hand over. The boundary is now yours to draw — and that’s genuinely new.

✅ How to make AI your calculator, not your crutch

A few habits we’ve found keep you in the driver’s seat — at work and everywhere else:

1. Think first, prompt second. Sketch your own answer or outline before asking AI. You’ll spot instantly whether its output is better or just faster.
2. Interrogate, don’t accept. Ask it what’s wrong with its own answer. Ask for the counter‑argument. The Microsoft study is clear: scepticism is the skill that survives.
3. Keep deliberate practice alive. Pilots still fly manual approaches; keep writing, calculating and reasoning unassisted where the skill matters to you.
4. Delegate the mechanical, keep the judgement. Meeting notes, reformatting, boilerplate — hand it over gladly. Strategy, taste and decisions — those stay human.

We’ve written before about what this means for content and credibility — see our piece on E‑E‑A‑T in the era when AI writes everything. 🌐

CWA Europe uses AI the way we’d recommend anyone use it — as leverage for strategy, never as a substitute for it. Talk to us about your digital strategy.