Last week I caught myself doing something I didn’t love.
I’d opened up my AI tool to draft a quick reply to a client email. Nothing complicated. Two or three sentences I could have written in two minutes if I’d just sat with it.
But there I was, prompting and tweaking instead…
By the time I had a version I was happy with, I’d spent longer than if I’d just written the thing myself!
And this wasn’t the first time. I’d noticed this pattern building for a while. The reach for the tool had become automatic even for the small stuff. The stuff my brain could absolutely do. The stuff it definitely should do…
So I made myself stop. Because this isn’t really about one email. It’s about a much bigger risk emerging in workplaces right now – capability erosion.
Remember the Uproar?
Cast your mind back to midway through last year, when Kosmyna et al.,’s MIT study on the neural toll of LLM-assisted writing landed and critics climbed onto the highest of soapboxes. “AI is rotting our brains. We’re cooked.” The headlines wrote themselves.
The study had some limitations and was widely critiqued but the instinct underneath the reaction was right. We sensed something was happening. We just didn’t yet have the language for it.
Why Your Brain Wants the Shortcut
Joel Pearson wrote something in his Neuro-Futurist Newsletter recently that has stayed with me.
Critical thinking isn’t a single skill. It’s an orchestra. It requires multiple systems working together and is metabolically expensive. So your brain would really prefer not to bother.
Which means when you’re offered a shiny tool that does the thinking for you, your brain will say yes please, every time. Not because you’re lazy. But because you’re human, and your brain is wired to conserve energy wherever it can.
That’s what makes this so quietly dangerous. The erosion doesn’t feel like erosion, it feels like efficiency.
The Moment of Choice
Some institutions are already responding. Cornell University recently introduced a module teaching students to notice the moment they’re about to let AI do their thinking, and to choose deliberately to do it themselves.
That moment of choice matters. It’s the difference between using a tool and surrendering a skill.
Many organisations are measuring AI adoption. Far fewer are measuring what may be lost alongside it.
And that matters, because the World Economic Forum Future of Jobs Report ranks analytical, creative and systems thinking among the most important skills for workplaces through to 2030.
When routine thinking is repeatedly outsourced, three business risks emerge:
-
Decision quality drops – bad recommendations move faster
-
Capability erodes – writing, reasoning, thinking and judgment decline when unused
-
Oversight costs rise – leaders spend more time checking and correcting.
That’s definitely not efficiency, that’s just delayed costs.
Why This Matters Now
Just look at what happened with Deloitte Australia last year. It took Dr Christopher Rudge from University of Sydney to uncover inaccuracies in a 237-page report delivered to the Australian Department of Employment and Workplace Relations. Deloitte later admitted AI had been used to assist and proper human oversight had failed – the thinking layer was missing.
If one of the world’s largest professional services firms can let that happen, what hope do the rest of us have if we don’t deliberately build our thinking skills?
When calculators arrived in classrooms, the maths didn’t get easier. The questions got harder. The tool raised the bar rather than lowering it.
That’s the opportunity in front of us now. AI should raise the quality of human thinking, not replace it.
The future won’t belong to those who use AI most. It will belong to those who know when not to.
Next time you reach for the tool, pause for five seconds and ask whether your brain could have a go first.
So, where is convenience costing you more than you realise?
