Will AI Make Us Lazy?

The Science (and Soul) of Cognitive Offloading

“Musicians should go to a yard sale and buy an old f***ing drum set and get in their garage and just suck... and then they’ll f***ing start playing and they’ll have the best time they’ve ever had in their lives and then all of a sudden they’ll become Nirvana. You don’t need a f***ing computer or the internet or The Voice or American Idol.” Dave Grohl, Foo Fighters (and Nirvana)

Let’s talk about cognitive offloading … the process of delegating mental tasks to external tools, like writing things down, setting a reminder, or asking your AI assistant what day of the week Christmas is this year. Cognitive offloading is nothing new. Humans have been doing it since we scratched calendars into bone or tied knots in string. But in an age of AI and ambient intelligence, cognitive offloading has gone full steroidal. Instead of calendars, we now have scheduling AIs that anticipate your next meeting. Instead of remembering facts, we outsource to ChatGPT. If we need a picture of a cat, we ask DALL-E. We want to better manage our time, we go to Taskade etc. etc.

This leads to an obvious question … Is AI making us lazy?

If you have seen the Disney/Pixar film WALL-E, you have seen a future where humans, who have evacuated Earth, live such a pampered sedentary life that they have become obese and immobile and rely on hoverchairs to get around, consuming most of their meals through straws. Is there an intellectual version that that, where we become ‘flabby thinkers’ … a condition brought about by the comforting embrace of AI?

Extended Minds, Lazy Times?

In 2016, Evan Risko and Sam Gilbert wrote a field-defining paper titled Cognitive Offloading in Trends in Cognitive Sciences. Their thesis? Humans naturally reduce internal cognitive load by pushing memory and problem-solving tasks into the environment. It’s not cheating; it’s evolutionarily brilliant. That shopping list? Offloaded memory. Your phone’s GPS? Spatial cognition, outsourced. The AI writing your emails? Language production, delegated.

Cognitive science has evolved from seeing the brain as an isolated processor to something more like a networked node. Philosophers like Andy Clark and Rob Rupert talk about “extended cognition,” where your tools (including AI) become part of your thinking system. Philosopher Richard Heersmink goes even further, saying these tools aren’t just add-ons—they shape who we are. We become what we offload into.

The science is broadly clear: offloading works. It saves energy. It allows higher-level cognition. It prevents overload. But here's the flip side … offloading changes us. It changes how we think, and whether we think at all.

Exercising The Grey Matter. The Muscle You Don't Use.

Let’s go back to Dave Grohl. His quote about buying a crappy drum kit and just sucking your way to genius has serious cognitive resonance. What he’s describing is a form of embodied learning. Not delegating the struggle but wading into it and celebrating it. Grohl’s argument? No shortcut replaces the garage. No software simulates what you will learn whilst you’re sucking at playing the drums.

This is where cognitive offloading starts to get existential. Because not all thinking is utilitarian. Not all problem-solving should be frictionless. Sometimes it’s the doing that is the learning. The screw-ups, the failures, the moments when you reach for an answer and there’s no AI there to catch you. Because if AI is there to catch us every time, do we ever build the muscles to think?

The Subtle Erosion of Struggle

This isn’t nostalgia for chalkboards and phonebooks. It’s a serious question about skill decay and cognitive atrophy. And it’s already happening.

Smartphones have demonstrably eroded our spatial memory (GPS) and face-name memory (contacts and social media). The term for this is the Google Effect—the tendency to forget information we think we can easily retrieve later. Now imagine this multiplied by generative AI that drafts your reports, answers your questions, even solves your work tasks before you understand the problem.

One recent study showed that people who used AI to generate written responses on professional tasks became more reliant on the AI with each iteration; and less confident in their own unaided thinking. That’s offloading with a feedback loop.

And as we offload, the “offloaded” parts become less practiced. Less sharp. Less available when we need them. What if your AI goes down? What if the tools vanish? What if you need to improvise?

But let’s not get too hand-wringy. Offloading doesn’t automatically equal laziness. In many cases, it enables better thinking. It frees us from mental drudgery—memorizing bus timetables, proofreading 5,000-word docs … who needs that if we can free up time for interpreting, connecting, imagining. That’s the point Risko and Gilbert make … offloading isn’t avoidance; it’s delegation. Used well, AI allows us to become meta-thinkers, people who think about how they think. You’re not just solving the puzzle you’re redesigning the puzzle box. That’s a whole new level of cognition.

But used passively, AI risks turning us into... well, couch potatoes of the mind. People who look productive because the system is humming—but who haven’t had an original thought in weeks.

The Skill of Not Offloading

This is the weird paradox: in the age of ubiquitous AI, not offloading becomes a skill.

Choosing to wrestle with the raw stuff of thought. Choosing to write the paragraph yourself. Choosing to remember the fact without asking Alexa. Not because you have to—but because you want to. That feeling—the friction of thinking—is where real learning lives. Not just information, but transformation. As cognitive scientist David Kirsh put it, “Thinking is often done with external representations. But some kinds of understanding require you to internalize them first—to wrestle.”

So here’s a challenge: Do something this week the hard way. Write your own draft. Sketch a diagram on paper. Give directions without GPS. Cook without googling the recipe. Build your own mental drum kit, even if it sounds like trash.

You’ll suck. And then you’ll get better.

As AI becomes ubiquitous and ambient, we need to get sharper at asking: What am I offloading? And more importantly: What am I losing when I do? Cognitive offloading is not the enemy. But unconscious offloading? That’s a risk.

Grohl was right. The best stuff often comes not from skipping steps but from being in the room when it sucks. When it’s messy. When it’s real.

Maybe the future belongs to those who can blend both worlds— people who know when best to offload, and when to choose not to.

“In this day and age, when you can use a machine or computer to simulate or emulate what people can do together, it still can't replace the magic of four people in a room playing” Dave Grohl

Next week: What if we could map the kinds of struggle that truly sharpen the mind? In Part II, we introduce a mental model—The Cognitive Friction Cycle—to help defend the thinking that AI can’t replace.