Let’s not kid ourselves: today we’re all being pushed to use AI to write, analyse data, sketch out strategies, synthesise ideas, write code — faster and faster, with less and less actual thinking. Those who do it are probably productive and keeping their jobs; for everyone else, oblivion — maybe, for now.
And yet, silent as a killer lurking in the shadows, there’s a risk that AI is eroding the fundamental skills that make us effective. It’s called Skill Atrophy — the gradual loss of judgment and pragmatic execution that comes from over-delegating to GPT models. In plain terms: the slow death of active thinking.
When we offload to AI, yes, we get through our professional workload more easily. But that doesn’t exempt us from thinking critically, synthesising, choosing the final form in which we communicate our ideas — regardless of whether a machine suggested them in the first place.
So if our brain must remain in the production loop, we’re looking at three possible scenarios:
-
In the first scenario, we become the bottleneck for the machines — slaves to processes designed for 5–7 interactions a day, not the hundreds that AI’s speed inevitably demands. Here, our human limitations are what slow the machine down, and we’re its servants either way.
-
In the second scenario, we get comfortable with the LLM’s ready-made answers just to keep up with the pace AI sets. We produce more interactions per day, but we lose the instinct to ask questions, to challenge the machine’s output. Compared to the first scenario, we’re just differently enslaved.
-
In the third scenario — desirable, difficult, and I fear impossible for many — we find a middle ground: neither a bottleneck nor a biological clipboard passing GPT output straight into our Office suite.
The skills at risk, incidentally, are not technical ones. It’s hard for a programmer to forget how to programme. The skills at risk are cognitive and executive.
Specifically, what we stand to lose:
-
Critical thinking — the ability to analyse and challenge what’s been proposed, by constructing and using arguments, not prompts.
-
Decision-making capacity — the faculty to weigh options and evaluate their consequences. Fast interaction with GPT promotes, alas, fast thinking — much to the detriment of slow thinking and everything it’s worth.
-
Strategic vision — independently imagining optimistic and pessimistic scenarios, estimating the probability of each, and building primary and contingency plans.
-
Relational ability — the capacity to communicate and negotiate directly with the people in front of you, without a machine as intermediary. For instance: if you’re having GPT write your emails or LinkedIn posts, and your contacts can tell — well, you clearly have a problem.
If we want to prevent Skill Atrophy, we need to change how we use AI.
First, we need to use it as a sparring partner, not a crutch — writing prompts that explicitly ask the machine to reason with us, to help us sharpen our own perspective. In practice: use interactive prompts and ask the machine to challenge you, not to think for you.
Second, we need to keep making decisions without AI’s help. Do we really need an AI to tell us how to build a marketing plan, evaluate a candidate’s CV, or frame a result that didn’t land perfectly? If we’ve accumulated any real experience, we should already know the answer to those questions is it depends — and it depends mostly on contextual information that only we are equipped to judge.
Finally, whenever we interact with AI, we should always ask ourselves: What’s the rationale behind this solution? What’s missing? Do I have alternatives? We can’t always take shortcuts. Sometimes we need to embrace slow thinking — and so be it.
AI is a remarkable tool. But I don’t think most people are using it correctly. Many believe they’re becoming more effective. In reality, they’re probably just anaesthetising their own intelligence.
So here’s my question: five years from now, who will actually be competitive — the person who knows how to use GPT, or the person who still knows how to think without it?
Worth discussing?
#ArtificialIntelligence #Skills #ProfessionalGrowth #Leadership #FutureOfWork #Innovation #CriticalThinking #Mindset #DigitalTransformation #ContinuousLearning