The Cognitive Gym
Exploring how AI is reshaping the way we think — and why cultivating a “cognitive gym” might be the ultimate competitive advantage in the machine age.
Once upon a time, we built agriculture to amplify resilience.
More recently, we built factories to amplify muscle.
Then we built computers to amplify logic.
Now we are building machines to amplify thought itself. In doing so, we are quietly dismantling the need to think. The need to speak. The need to collaborate.
As physical labour declined, we built gyms to keep our bodies strong.
As cognitive labour declines, what will we build to keep our minds alive?
The think-on, think-off switch
Thinking used to be demanded by the world. From deadlines to decisions and debates, opportunities to think out loud and in silence were aplenty.
Now? Thinking is a box you may tick or not.
AI chatbots draft your emails. AI video generation tools craft marketing campaigns. AI reasoning models create your research plan and deliverables.
Efficiency has increased, and will continue to do so. That’s a great thing for businesses.
Efficiency has replaced effort. That’s a terrible thing for businesses and for people in businesses.
Without effort, cognition softens.
And if it softens too much, there are sprawling ripple effects:
Emotional intelligence decreases
Creativity goes down
Critical thinking collapses
According to the 2024 Microsoft Work Trend Index, 64% of employees struggle to find time to think deeply at work — even as productivity tools proliferate.
That’s not even the worst part. Too heavy a reliance on chatbots is messing with your brain circuitry: it hides the learning curve and infuses a ‘degrowth’ mindset. Science is starting to prove it: neuroscientists at the University of Zurich recently showed that delegating reasoning tasks to algorithms dulls the brain’s anterior cingulate cortex, the area that governs error detection and learning.
The question is, what kind of workplace do we want, and how do we want to interact with technology within it?
The brain as a weapon of mass elevation
To identify a positive way forward, let’s take a step back.
By and large, the industrial age created the culture of fitness. Today, the most fortunate of us have the possibility to invest in our health. We can train the body for resistance, balance, and resilience.
The intelligent age needs to create a culture of cognitive fitness: places, rituals, and tools designed to train the mind for depth, synthesis, and reflection.
Just like the harm GenAI causes needs to be fiercely combatted, the value it adds needs to be evaluated through the right human lens.
Cognitive gyms can serve that purpose. They are not about regulation and mitigation. They’re about elevation.
Cognitive gyms ensure we stay proactive to keep AI in check through three goals:
Introducing deliberate mental distance between an AI output — a GenAI output, to be specific — and our own thoughts.
Reinstating analytical friction — asking “why” before accepting convenience.
Training metacognition — learning to think about how we think when we use AI.
These cognitive workouts need an enabler: a structured, coherent, and even harmonised AI playbook across the corporate world. And we can already write its first chapter:
LLMs are probability machines
Probability equates uncertainty
Uncertainty means unreliability
As an employee, as a colleague, as a leader, do you really want to be unreliable?
If the answer is no, you should head to the cognitive gym to future-proof and stress-test yourself. While you’re there, carry out some thinking training. Here are two simple examples:
Pick out a press article about something you don’t know in a country you don’t know, summarise it, explain it.
During a collaboration sprint, run an AI-generated idea past your team — then ask everyone to challenge one hidden assumption.
Possibilities are infinite, and there is no prescribed format. The goal is to take the brain out of its routine. It’s not necessarily exciting, just like a gym workout. It’s not delivering results overnight, just like a gym workout. It’s an investment in the ability to handle pressure, ‘storytell’ effectively, communicate precisely, lead enthusiastically.
No need to be an oracle: tomorrow’s advantage won’t belong to the most informed.
As I mentioned here, knowledge and information hoarding will soon be a thing of the past.
The advantage will belong to those who can integrate, interpret, and imagine. In AI-powered workplace arenas, I hope to see emerge cognitive athletes of the machine age.
Born to be alive
Those among us who run have certainly experienced the runner’s high, the euphoric state caused by endorphin and endocannabinoid release during intense effort that mirrors the neural reward of cognitive strain overcome.
The same logic prevails with the mind and the thinker’s high:
A delay preventing you from catching your plane? It forces you to think of another path.
A key detail missing for this board presentation? It prompts you to make do without it and reshape your story.
A glitch in your payment system? It pushes you to think creatively to keep orders coming in.
Just like runners push through despite a mind begging them to stop, the cognitive effort comes with rewards: it keeps our brain — our unique thinking muscle! — engaged. Our neurons stretch to connect the unfamiliar.
More on this can actually be found in a 2023 study from the University of Western Ontario, in which they found that participants solving difficult reasoning problems without AI assistance experienced a similar dopamine release pattern to endurance athletes.
Cognitive offloading and the illusion of effort
We have quite a dilemma ahead of us: the barriers to producing any kind of cognitive effort are falling one after the other. For some in the workplace, it is seen at face value as a productivity lift.
The reality is more pernicious. Insidiously, the thinking frontier is shifting into the synthetic hands of technology tools. With GenAI, this process has been dubbed cognitive offloading, but it’s unfortunately more subtle.
Offloading implies a deliberate action. What I am increasingly observing is an unconscious action of knowledge workers who assume an AI chatbot will always take care of the heavy thinking. If the AI assistant is doing the most value-adding activity, one can legitimately wonder who the real assistant is.
We see this happening in real workplaces. At one consulting firm, analysts now start every client engagement by prompting ChatGPT before even reading the brief. Outputs look impressive — but clients report a ‘loss of spark’ in the recommendations.
It’s pointless to assign any blame. In the tech space, products and services are launched with little consideration for the cognitive impact they may have. This is unlikely to change. The frictionless interfaces being created by tech firms are desirable to facilitate the outcome of human critical thinking, communication, and creativity.
But they are not a substitute for actually carrying out the mental task. This is a slippery slope. In many organisations, productivity dashboards reward deliverables, not discernment. This makes offloading thinking to AI not just easy, but rational. According to Adobe Digital Trends 2025:
70% of marketing teams in the U.S. report using AI for campaign ideation
Yet 42% of CMOs admit their creative differentiation has declined
At the risk of sounding very European, I would suggest creating a cognitive impact score, similarly to the environmental metrics we see on home appliances. Think of it as a ‘cognitive nutrition label’ for every AI tool — showing how much reasoning, reflection, and decision-making it preserves for the user.
Reasoning: Does the tool leave you solving the problem, or does it do the logic for you?
Reflection: Does it prompt you to pause, question, and weigh alternatives, or simply provide the answer?
Creativity: Does it challenge you to combine ideas and explore novel solutions, or simply generate outputs on autopilot?
The score could be visual, colour-coded like a traffic light:
Green = high cognitive engagement. The tool acts as a workout partner for your mind.
Yellow = moderate engagement. You’re partially offloading effort.
Red = low engagement. The tool does almost all the thinking, leaving your neurons idle.
Just as we check calories or nutrients before eating, we could one day check the ‘brain calories’ or cognitive value of our tools before using them. The idea is not to avoid AI. It’s to choose tools that exercise your mind rather than atrophy it.
As GenAI continues to be more widely adopted, the frontier between efficiency and plain laziness becomes tenuous.
We’re entering a world where knowledge is free, but wisdom must be earned. We may witness massive technological progress powered by GPUs, but the real workplace revolution will happen with those who still choose to think.
In this new era, the most valuable skill will not be knowing more, but knowing how to engage your mind deliberately — to question, integrate, and create. The cognitive gym is no longer optional; it’s essential. Those who treat thinking as a daily practice, stretching their neurons and challenging assumptions, will emerge as the cognitive athletes of the machine age.