Distraction fragments attention. AI tools fragment it further—ChatGPT is a slot machine with a keyboard. Feed it a prompt, watch the spinner, receive your unpredictable reward. The mechanism mirrors Instagram’s infinite scroll, but with higher-bandwidth novelty.
Packy McCormick recently observed that despite unprecedented abundance—AI tools, infinite information, zero distribution costs—original ideas seem to be declining. The paradox resolves once you see the trade: we exchanged depth for access.
The cost: creativity disappears. Creativity requires sitting with an unresolved problem long enough to see new angles. When you pull the lever every three minutes, the angles never form.
This isn’t an argument against AI tools. It’s an argument for understanding curiosity’s compounding effects and the loss of curiosity in the age of abundance.
Creativity doesn’t arrive unprompted. It emerges from a specific engine: curiosity that persists long enough to produce unexpected combinations.
The loop works like this: a question lodges in your mind. You turn it over. You try one approach, then another. The approaches themselves generate new questions. Each answer reveals adjacent problems you hadn’t considered. Eventually, a connection fires that wouldn’t have occurred to someone who stopped earlier.
This is curiosity operating as creative fuel. It’s not just passively flicking through Wikipedia or Youtube - it’s the active pursuit that compounds.
I speak from experience, not expertise. I’ve performed music, recorded albums, written books. Not because I’m particularly skilled at any of them, but because I’m not blocked. Skill-wise, I’m below average in most domains, and the reason why I didn’t grow those skills further is exactly what I cover below. But the output happened. Most people I know with more talent produce less. The difference isn’t ability. It’s that the loop keeps running.
The loop has a vulnerability: it requires sustained attention. Break the chain early and the compounding stops. This is where distraction becomes lethal.
But curiosity doesn’t always ignite. Sometimes you approach a domain and feel nothing: no pull, no drive to continue. The standard explanation is that curiosity is innate: you either have it or you don’t.
This is wrong. The barrier is skill, not disposition.
Curiosity requires traction. A beginner at SQL stares at a query that returns an error. The error means nothing to them. There is no handhold, no incremental win that would pull them back tomorrow. So they close the tab.
Contrast this with a video game: die on level one, respawn in four seconds, try a different jump. The feedback is fast, the cost is trivial, the reward is visible. Curiosity survives because competence accumulates in small, undeniable steps.
The difference isn’t the person. It’s all about feedback.
When someone says they’re “not curious” about a domain, they’re usually describing a skill gap masquerading as a preference. They tried, felt incompetent, received no reward signal, and retreated. The retreat feels like disinterest. It’s actually self-protection.
Here’s the proof that curiosity isn’t absent, just mislocated: everyone was curious as a child. The capacity existed. Factors may have suppressed it: parents who punished questions, schools that rewarded compliance, failures that taught retreat. But the hardware remains.
More importantly, people remain curious in domains they don’t recognize as curiosity.
The person who claims no curiosity spends four hours researching fantasy football trades. Another binge-watches video essays about historical battles. Another knows every detail of their car’s engine. My wife devours psychological thrillers—specifically ones with female antagonists and a twist at the end. She has to know what happens. She has to know how the author constructed the misdirection. They don’t label this curiosity because it feels like recreation. But the mechanism is identical: questions arise, pursuit follows, knowledge compounds.
The problem is you’re attempting forced curiosity in domains where no traction exists, often because someone else declared the domain important (e.g. at work or in school). Meanwhile, your curiosity just bears a different name.
If skill creates traction and traction enables curiosity, then curiosity can be manufactured. But the process requires honesty about what’s worth pursuing.
The Selection QuestionBefore engineering curiosity, ask: should you be curious about this?Life is finite. Attention is zero-sum. Every hour spent forcing interest in knitting is an hour not spent on something with natural pull. Sometimes the correct answer isn’t “how do I get curious about this?” but “why am I trying?”The question has real stakes. People spend years forcing curiosity in careers that never fit, in hobbies recommended by others, in domains where they have no native advantage. The effort rarely converts to genuine engagement. Selection precedes construction.
But assume you’ve decided: this domain matters, traction doesn’t exist, and you want to build it anyway.
The Radio StrategyRadio programmers sandwich the unfamiliar between hits. The new single follows a song you already love; the next track is another you know. Your brain borrows the warmth.
I did this with Napalm Death and other less accessible music, because it felt good to crack the code of an album. At first it’s hard. You dislike what you hear. It’s new and unfamiliar, and all you want to do is go back to the album you’ve been having on repeat for a week. Napalm Death was abrasive: blast beats, vocals like a bandsaw. So I played one track, then something I already liked, then another Napalm Death track. The contrast softened the edge. After a week, I was in love.
The principle generalizes: pair the difficult with the familiar. Interleave the domain you’re building with a domain you already own. Read one chapter of the dense textbook, then something you enjoy. Work on the intimidating project for twenty minutes, then switch to a task where you feel competent. The familiar provides emotional scaffolding for the unfamiliar.
The problem sometimes is that you have a lot of hard things to do, and so you switch from one hard thing to another, and so of course you struggle. When I was starting out in product, every task was difficult, every day, all day long. The only time I felt relaxed was when I could crunch some numbers in Excel. Unfortunately, this task wasn’t common among the rest of the work as there wasn’t always data to crunch, so it was a real grind.
The Feedback Loop AuditManufactured curiosity requires constant reassessment. The questions to repeat:Am I gaining traction? Do I still believe this domain matters? Is the effort converting to genuine engagement? You don’t even need to ask these questions directly, you’ll know.
The loop either tightens - you’re building skill, feedback arrives faster, curiosity strengthens - or it doesn’t. If three months of effort produces no pull, the honest answer might be abandonment. This isn’t failure. It’s efficient reallocation.
Life’s too short to spend it on something that even after a while doesn’t pull you in. But I would vouch that you almost certainly will be pulled in if you persist a little longer than the first 3 tries.
The payoff for sustained curiosity is networked.
Early in any domain, learning feels like collecting isolated facts. You know this term. You can execute that procedure. But the pieces don’t connect—each one sits alone, inert.
Then a threshold passes. The isolated nodes start linking. A concept from chapter three illuminates a problem from chapter seven. A technique you learned for one task suddenly applies to another. The network begins to hum.
This is what it looks like when neural connections fire together. A single synapse is useless. A cluster begins to pattern-match. A network starts solving problems you never directly practiced.
And then the meta-level unlocks: learnings from one domain transfer to another entirely. The structure you internalized in music—tension, release, rhythm, repetition—appears in writing. The debugging logic from programming surfaces when diagnosing a business problem. The pattern recognition you built in chess shows up in strategic planning.
This is the compound interest of curiosity. But it requires surviving the early phase—the isolated-node phase where nothing connects and the feedback is sparse. Distraction kills you here, before the network forms.
AI tools, social networks are slot machines. Each pull delivers a reward. The reward feels like progress.
But notice what’s missing: you didn’t build the network. The AI answered the question before your own neurons could fire together. The connection that would have formed—the one linking this problem to that prior experience—never materialized. You outsourced the curiosity loop.
This is the extraction problem. AI tools simulate the satisfaction of curiosity without building the internal architecture that makes curiosity compound. You feel like you learned something. You retain nothing that transfers.
The person who asks ChatGPT to explain a concept and the person who struggles with the concept for an hour have different experiences. The first feels informed. The second built circuitry.
The difference becomes clear in how you use the tool.
One approach: you ask AI to write your annual goals for you. It produces a tidy list—exercise more, read more, save more. You nod, save the document, forget it by February. The AI did the thinking. Nothing transferred.
Another approach: you use AI to interview you about your goals. It asks questions you wouldn’t ask yourself. It surfaces contradictions between what you say you want and how you actually spend time. The goals that emerge are yours—the AI just held the mirror.
I did both recently. First, I had Claude interview me about my priorities and strategies for the year ahead. Most people do annual planning by staring inward, hoping clarity arrives. Using an external tool to structure the inquiry—to ask the second and third follow-up question—produced different results than introspection alone.
Then I went further: I used AI to gamify the goals. Not to tell me what to do, but to apply game design principles I wouldn’t have connected to personal planning. The AI didn’t generate my motivation. It helped me construct a system that sustains motivation using mechanics from a completely different domain. I will still have to do the work and struggle, but I am augmented with AI in areas where it would have taken me weeks to do by myself.
In both cases, the AI was additive, not substitutive. It didn’t replace the struggle. It created new surfaces.
That’s the distinction. Extraction: the AI answers, you receive, nothing builds. Construction: the AI assists, you struggle, the network forms.
This doesn’t mean AI tools are useless. It means they’re costly in ways the interface hides. Every pull of the lever is a trade: immediate answer for deferred capability.
The argument isn’t about abstinence but awareness.
Distraction in general—and AI distraction specifically—intercepts the curiosity-creativity loop before it can compound. The intervention points are clear:- Protect solitude. Creative insight requires uninterrupted time with an unresolved problem. This is non-negotiable. The length varies by person and problem, but the minimum is longer than you think. Thirty minutes without task-switching is a starting point, not a maximum.
The tool that answers before you struggle is the tool that ensures you’ll need it again. Curiosity compounds. Convenience doesn’t.