When it comes to AI and critical thinking is it a case of use it and lost it?
Something happened to me recently that I can't stop thinking about. I was asked to write a piece for a business publication using proper business language, authoritative tone, structured argument. I've been doing this kind of writing for 25 years and yet I froze.
My idea was solid and my initial pitch was strong, but it was the act of actually getting the words on the page that stopped me. How do I write this? Does this sound right? Where do I even start? I was second-guessing every sentence before I'd written it.
And this paralysis surprised me.
Part of it, I think, is due in part to how our use of language has shifted. Being human is now valued over being businesslike particularly on platforms such as LinkedIn with the rise of the personal brand. Informal has beaten formal on almost every platform you can think of. I've actively been part of that shift, finding my own voice and helping others find theirs. So going back to something more structured and traditionally authoritative felt really alien to me.
But I think there's something else far bigger going on here.
AI has quietly taken over a chunk of the mental work I used to do without thinking. The thinking, the structuring, the internal "does this flow?" check. When I had to do all of it from scratch, with no assistance, I realised some of that muscle had genuinely weakened. And yes it was an uncomfortable thing for me to sit with.
And don’t get me wrong this isn’t just something applicable to me. A study published earlier this year looked at 666 people across different age groups in the UK and found a significant negative correlation between how often people used AI tools and how well they performed on critical thinking assessments. The more we hand our thinking to AI, the less we practise it ourselves, and the weaker that capability becomes. Younger participants had the highest AI dependency and the lowest critical thinking scores which is a worry.
MIT's Media Lab went a step further. They split participants into three groups: one writing with ChatGPT, one using Google, one working unaided. Brain scans showed the AI group had the lowest neural engagement across the board, with weaker connectivity, lower memory retention and less sense of ownership over their own work. Although, this was a very small study of just 54 university students in Boston, what stayed with me was that the cognitive slowdown didn't stop when they stopped using ChatGPT. Even after the study ended, their brain activity stayed sluggish. This says to me that once you outsource your thinking, your brain doesn't immediately bounce back to where it once was.
I've been sitting with all of this, and then I heard something on a podcast this week that pushed it further.
A woman in tech in the US was talking honestly about her life. She's someone used to operating at pace, getting things done, bossing it. Then she had a child, which to be honest, anyone who's been through it knows completely shakes up your world. You're no longer on your own agenda, you’re now on theirs. The 5am gym visits, the uninterrupted focus blocks and working to god knows when, all of that goes out of the window
She was talking about how she'd leaned into AI tools to automate as much of the cognitive load as possible. parent I know it's easy to drop the ball myself so I completely got it, you're running on less sleep and less bandwidth, you reach for whatever gives you capacity back.
But I sat there and had two thoughts at once. The first was practical as someone who has worked with teams designing data centres, I know every AI query is drawing an enormous amount of energy from data centres and none of this is sustainable. That's not a small cost when you multiply it across millions of people doing the same thing. The second was more uncomfortable. If we hand over enough of our thinking, our structuring, our sense-making, what are we actually left with?
Harvard academics have said it plainly: letting AI write your first draft doesn't just change how you write, it changes how you think. Research with knowledge workers has found people are already self-reporting lower cognitive effort, and what researchers are now calling cognitive debt. We're borrowing against our own capabilities.
This matters more in some sectors than others. In the built environment, architects, engineers and contractors spend careers learning how to translate deeply technical thinking into something a client can follow without losing the nuance. That craft wins work and if it gets quietly outsourced to AI, they're not just losing a writing skill. They're eroding one of the few things that actually differentiates them from competitors with identical technical credentials.
I love AI. I use it myself daily and I help clients think through how to use it well. I'm not standing here waving a flag against it. But I think we're moving fast enough that we're not asking some of the right questions. What happens when we optimise so much of the mental load away that we've also optimised away the skills that made us good at what we do? And who's actually thinking through those consequences? Right now, it feels more like the tech bros building the products than governments. And that, dare I say it when you look at many of them as individuals, is a problem.
When it comes to AI and critical thinking making sure we all continue to use critical thinking and remain cognisant about where our own thoughts are and AI starts is super important. Use it or lose it isn't just something people say about the gym. It applies to how we think, how we write, how we structure an argument without a prompt box in front of us.
I found that out writing a business article and It's probably something that’s worth us all paying attention to.
Thanks for reading
If you want to get clear on what's actually working in your marketing and what's just noise, that's exactly what I do. I work with AEC Directors and marketing leads to deliver strategic marketing that helps you compete. Email ayo@abbasmarketing.com
