Basil Jarrett | Testing the obvious

NOBODY NEEDED a PhD, a lab coat or expensive testing equipment to figure this one out, but in a “surprising” revelation from the ever-brilliant minds at MIT’s Media Lab this week, it turns out that using AI might be making our brains lazier. Wow! You don’t say. Shocking, right? In the same way that the invention of the TV remote made us too lazy to walk three feet to change the channel, or that Google Maps killed our natural sense of direction or willingness to pull over and ask somebody which way to Albuquerque, it really shouldn’t take expensive brain scanning equipment to convince us that if you outsource your thinking to a chatbot, your brain might just sit the exercise out.

TESTING THE OBVIOUS

MIT tested three groups of people writing essays: those who used only their brains, those who used Google, and those who turned to ChatGPT. Surprise, surprise, the brain-only group had higher engagement, stronger neural activity, and wrote more creative, insightful essays. The ChatGPT group? Not so much. Their work was neat, yes. But also mechanical, repetitive, and, according to English teachers, utterly soulless. By essay number three, most participants had gone full autopilot: copy, paste, tweak a sentence or two, and submit. You don’t need electrodes and EEGs to tell you that’s not how the mind grows.

But here’s the real kicker. When ChatGPT users were asked to revise one of their essays without using AI, they couldn’t remember a thing they had written. Why? Because they didn’t write it. It’s the intellectual equivalent of asking someone to explain a movie they slept through.

THE REAL QUESTION

Now, before you cancel your ChatGPT subscription and wrap tin foil around your head, let’s take a step back. The problem here isn’t AI. It’s how we use it. Technology has always had this effect. Cars made us walk less. Calculators made us forget our times tables. The spellcheck in MS Word gave us amnesia for i-before-e rules. AI is simply the latest tool in a long history of human inventions that trade cognitive effort for convenience. So no, MIT hasn’t discovered anything new. They’ve simply confirmed the obvious: if you stop using your brain, it stops working as well as it should.

The real question then isn’t whether AI is dumbing us down, but rather, how do we stop it from doing so. The first place to start is with our children. If we’re handing our kids tablets at age two, letting Alexa answer every question, and allowing ChatGPT to write their homework while we beam with pride at their “brilliance”, then we’re not just failing as guardians and educators, we’re actively aiding and abetting their intellectual decay. One idea to fix the problem is to ban AI for first drafts. Let children write from scratch, make mistakes and struggle. You know, like how us adults did it. Then, and only then, let them use ChatGPT as a revision tool, not as a crutch.

TEACHERS MUST ADAPT

Teachers too must resist the temptation to throw tech at every problem. Not every assignment needs to be typed, and not every project needs to be “digital”. Writing cursive isn’t archaic. It’s a cognitive workout that reinforces memory and recollection. Greater reading comprehension comes from wrestling with text, from flipping pages, from margin notes and mental debates, not by cutting and pasting. We need to bring back the basics to the classroom even as we adopt to AI. And even as we do make ChatGPT as standard in the classroom as chalk and dusters, we should also encourage students to interrogate it. “What did ChatGPT get wrong?” “How would you say it differently?” “What assumptions did the AI make?” If students can’t answer those questions, they’re not learning.

Certainly, there also needs to be national guidelines on AI use in schools, and we see where Government is looking in that direction. But until then, we need to immediately adapt curricula that teach kids not just how to use ChatGPT, but how to think about what it spits out. We need policy and practices that ensure our education systems integrate AI responsibly, without turning schools into glorified prompt machines.

GYM SCIENCE

Make no mistake: AI is here to stay. And that’s not a bad thing. ChatGPT can be an extraordinary tool for brainstorming, summarising, explaining complex ideas. If we want our children to be thinkers, doers, and problem solvers, and not just prompt whisperers and content curators, we need to teach them to use their brains first, and their tools second.

Regular gym goers will tell you that if you don’t train your legs the way you train your biceps, chest and triceps, you’ll end up looking like 90s cartoon hero, Johnny Bravo. It’s a phenomenon known as muscle atrophy. But just like unused muscles that slowly wither away, so too does the brain when it’s constantly handed a shortcut. Every time you read, write, wrestle with a tough idea, or form your own opinion, you’re lifting cognitive weights. You’re building stamina, creativity, logic. But if ChatGPT is doing all the heavy lifting for you, don’t be surprised when your intellectual frame starts to sag under the weight of real-life challenges.

That’s why the latest “groundbreaking” study out of MIT, claiming that ChatGPT users demonstrate lower brain activity and cognitive performance than their non-AI peers, should’ve surprised absolutely no one. Any gym bro could have told them that.

Major Basil Jarrett is the director of communications at the Major Organised Crime and Anti-Corruption Agency (MOCA) and crisis communications consultant. Follow him on Twitter, Instagram, Threads @IamBasilJarrett and linkedin.com/in/basiljarrett. Send feedback to columns@gleanerjm.com