From Cognitive Offloading to Cognitive Parkour

 

In Back to the Future Part III, Doc Brown finds himself in a saloon in the American Wild West. Drunk, heartbroken, and far too honest for his own good, he begins telling the locals about the future. He explains that one day, people will travel in automobiles. They will no longer need to run everywhere. And then he says something that sounds completely absurd to them: in the future, people will run “for fun.”

The men in the bar laugh at him.

From their perspective, the idea is ridiculous. Running is not leisure. Running is not self-expression. Running is what you do because you must. You run from danger. You walk because you have to get somewhere. Physical effort belongs to survival, work, and necessity. The thought that people would voluntarily run, stretch, climb, lift weights, practice gymnastics, do yoga, or jump between walls for the sheer joy of it would have seemed almost insane.

And yet, Doc Brown was right.

Modern transportation removed much mandatory physical effort from daily life. We no longer need to walk for hours to buy food, carry heavy loads across long distances, or travel by foot as a basic condition of survival. This has created real problems. Sedentary life can harm the body. Cars, screens, and convenience can make us physically inactive. But the solution was not to design cars that force people to pedal while driving in order to “keep them physically engaged.” That would be a strange form of human-centered design.

The proper role of ergonomic design in a car is not to turn driving back into walking. It is to prevent the car from actively damaging the body. Seats should not destroy the spine. Visibility should not create unnecessary strain. Controls should not be confusing or dangerous. Safety systems should prevent injury. In other words, cars should be physically ergonomic.

But beyond that, the liberation from mandatory physical effort created something extraordinary. It created time and space for voluntary physical excellence. Running became sport. Climbing became recreation. Dance became a field of bodily expression pushing choreography to ever higher limits in terms of complexity and sophistication. Gymnastics pushed the body into possibilities that ordinary survival work would never have produced. Parkour turns urban space into a playground of movement. Yoga transforms posture, breath, attention, and flexibility into a lifelong practice. These new bodily expressions, in many ways, are useless. They serve no practical purpose. Yet a gymnast can still walk to the store to get food, if necessary. But gymnastics is not “useful” in the same narrow way walking to the store is useful. Its purpose is not survival. Its purpose is to expand the range of what a body can do.

This is the analogy we urgently need for AI.

For decades, digital tools have helped us offload cognitive tasks. Search engines coupled to digital repositories remember facts for us. GPS calculates routes for us. Calculators perform arithmetic for us. Calendars remember appointments for us. Now large language models can draft, summarize, translate, plan, brainstorm, code, and argue for us. Each of these tools removes a cognitive burden that used to require effort.

There is nothing inherently wrong with this. Cognitive offloading is part of human intelligence. Writing itself is cognitive offloading. Books, maps, diagrams, notebooks, clocks, abacuses, libraries, and computers all externalize thought. Humanity became intelligent not by keeping cognition trapped inside the skull, but by distributing it across tools, symbols, institutions, and shared environments.

The problem is not that we offload cognition. The problem is what we do with the freedom that offloading creates.

Research on cognitive offloading shows that the use of external tools can change what we remember and how we think. The “Google effect” suggests that when people expect information to remain available online, they are less likely to remember the information itself and more likely to remember where to find it. Studies on GPS use suggest that habitual reliance on navigation systems can be associated with weaker spatial memory when people later navigate without assistance. Research on calculator use has long raised the question of how automation can improve performance while reducing engagement with underlying concepts. More recent work on AI tools points in the same direction: heavy reliance on AI may reduce critical thinking, cognitive engagement, memory ownership, and independent effort, depending on how the tools are used.

In a previous article I discussed the concept of “noetic ergonomics”, which is broader than “cognitive ergonomics”, dealing more specifically with the impact of using digital tools on human cognition.

Human-computer interaction has often asked how interfaces can be more usable, more accessible, less frustrating, more aligned with human needs. That remains essential. If a tool manipulates attention, increases anxiety, induces dependency, hides errors, or encourages passivity, it is badly designed. In the same way that a chair should not damage the back, an AI system should not damage the mind. A cognitive tool should not make users less capable, less curious, less confident, or less able to think without it.

But noetic ergonomics should go further than harm reduction. It should ask how cognitive tools can help humans develop richer forms of thought.

At the moment, much of AI is being deployed in the opposite direction. Instead of freeing people from burdensome cognitive work so that they can engage in higher forms of thinking, AI is often used to increase the volume of low-value cognitive work. We automate email so we can send more emails. We automate reports so organizations can demand more reports. We automate summaries so people can process more documents, attend more meetings, and respond to more administrative expectations. The burden does not disappear. It multiplies.

This is the tragedy of badly used cognitive automation. It does not liberate the mind. It accelerates mandatory uninteresting and burdensome paperwork. How ironic to deplore the diminishing skills people have in manually filling out their tax forms because of AI automation. This should be celebrated. Using human cognition for form filling is an insult to human intelligence and cognition to begin with.

Imagine if the invention of the car had not created leisure time for sport, play, dance, and exploration, but only required people to commute farther, sit longer, and work more. In many ways, that did happen. Physical automation created freedom, but capitalism often captured that freedom and converted it into productivity. Now the same pattern is repeating at the cognitive level.

AI can write faster, therefore humans are expected to produce more text. AI can summarize faster, therefore humans are expected to read more. AI can schedule, optimize, analyze, and draft, therefore the complexity of ordinary work increases until the saved time disappears. Cognitive automation becomes not liberation, but intensification.

A human-centered AI interface should not simply minimize friction. Sometimes friction is good. A gym is full of deliberate friction. A piano lesson is full of deliberate friction. Chess, poetry, mathematics, martial arts, debate, and philosophy all involve difficulty that is not a bug but the source of growth. The goal is not to remove all effort from human life. The goal is to remove degrading, repetitive, meaningless effort so that people have more energy for meaningful, playful, developmental effort.

This is the difference between cognitive offloading and cognitive parkour.

Cognitive parkour would be the design of activities, tools, and cultural spaces that push the mind beyond its ordinary routines. Not because survival demands it. Not because a boss demands it. Not because productivity software demands it. But because thinking can be playful, beautiful, difficult, social, embodied, and fun.

A society that uses AI well would not tell children: “You no longer need to write, calculate, navigate, remember, or argue.” It would say: “Now that some burdens can be automated, let us create new games of thought.” Mental mathematics as play. Memory palaces as art. Navigation without GPS as urban adventure. Debate as cooperative exploration. Writing not as administrative output but as style, voice, and self-discovery. Philosophy as intellectual dance. Coding as creative construction. Scientific reasoning as puzzle-solving. Storytelling as world-building.

The future of AI should not be a world where humans become passive consumers of machine-generated cognition. It should be a world where machines take over cognitive drudgery so that humans can practice more demanding, joyful, and expressive forms of cognition.

This also means rethinking leisure. If leisure time is captured by passive consumption, then cognitive automation will weaken us twice: first by doing our thinking for us at work, and second by feeding us endless streams of content when we are tired. A healthy cognitive culture needs active leisure, not only entertainment. It needs spaces where people play with ideas, not only consume them. It needs tools that invite users to stretch their minds, not merely outsource them.

The risk of AI is not only that it will become too intelligent. The risk is that we will become comfortable becoming less alive in our own thinking. To become “cognitively sedentary”.

The best AI systems should be like good physical infrastructure. They should automate uncreative burdensome cognitive tasks, protect us from unnecessary cognitive harm, while leaving room for challenge, play, and development. A good city has roads, but also parks. It has public transport, but also walking paths. It has elevators, but also staircases. It does not force everyone to exercise all the time, but it makes movement possible, attractive, and socially meaningful. Walks are not just functional, but especially enjoyable. 

Likewise, a good cognitive environment should include automation, but also spaces for thinking without automation. It should include assistance, but also invitations to struggle productively. It should include answers, but also better questions.

The saloon men laughed at Doc Brown because they could not imagine running for fun. Today, we risk making the opposite mistake. We can imagine machines doing our thinking for us, but we struggle to imagine what higher forms of human thinking might become once freed from drudgery.

That is the real opportunity of AI.

Not a future where we never need to think, but a future where thinking becomes an art of living, and a mode of human expression and experimentation;  future where thinking becomes more playful, more creative, more demanding, and more human than ever, liberating human thought much like bodily expression was enhanced in the last century.

Hopefully, one day, human CVs will be populated with select video game performance benchmarks or familiarity with recognized cognitively enhancing board games.

by Martin Schmalzried , AAIH Insights – Editorial Writer