A Response to the Fear That AI Is Stealing Our Curiosity
Kendra Cooke’s essay is intelligent, well-intentioned, and animated by a real concern: that in the age of AI, people may surrender the discipline of curiosity for the comfort of immediate answers. She is not wrong to worry about shallow thinking, untrained adoption, or polished output unmoored from understanding. But the essay finally arrives where so many critiques of AI now arrive, at a conclusion that mistakes the tool for the failure surrounding it. It treats AI as the accelerant of decline when what it is often accelerating is something older and more human: intellectual passivity, institutional laziness, and a longstanding desire to reach conclusion without earning it. That is not wisdom. It is a failure of imagination dressed up as a warning.
The fear of AI as cognitive corrosion is not a new argument. It is the same argument leveled at the printing press, at the calculator, at Google Search. It is the recurring panic that arrives whenever a tool lowers the cost of accessing information or capability, and it consistently misidentifies the tool as the threat rather than the habits of mind that existed long before the tool arrived. Attention was fragmenting before ChatGPT launched. Executives were rubber-stamping decisions they did not fully understand long before any generative model existed. The erosion Cooke describes did not begin in 2022. AI simply made it visible at scale.
But that is not what this technology has done to me.
What AI Actually Did to My Curiosity
I will speak for myself here, because the personal is often the most precise instrument. AI did not reduce my curiosity. It expanded it in ways I did not anticipate and did not fully appreciate until I stopped to examine the shift. The mechanism is not ease. This is an important distinction that the fear narrative consistently collapses. The mechanism is efficiency in the service of depth. Easy suggests passivity. Efficient means I can move faster into the real labor of thought: finding more articles, locating stronger counterarguments, surfacing the pros and cons with greater range, testing assumptions against competing evidence, and synthesizing a wider body of material with more speed and more precision than before.
That is not intellectual laziness. That is intellectual acceleration.
Before AI-assisted research, the cost of pursuing a thread was high enough that I would often stop at the first defensible answer. A search engine returns ten links. I would read two or three. The friction of going further was real, and the friction was not productive. It was not the friction of wrestling with uncertainty; it was the friction of navigation, of tab management, of paywalls and inconsistent sourcing. That friction did not build intellectual muscle. It simply limited range. Too much of this conversation still confuses slowness with rigor. It romanticizes the old friction of search and treats laborious retrieval as though it were proof of seriousness. It was not. The point of research was never the mere performance of struggle. The point was always to find the strongest arguments, the sharpest contradictions, and the deepest patterns, then think hard enough to make sense of them.
When I am working through a complex argument, I now synthesize across sources in a single session that would have taken days of library work or subscription-hopping. I read the pro. I read the con. I find the meta-analysis. I find the critic of the meta-analysis. I identify the ideological assumptions embedded in each camp. Curiosity is not measured by how long it takes to locate information. It is measured by what one does after the material is found. Do I follow the thread? Do I test the claim against its opposite? Do I synthesize, challenge, refine, and return to the question with greater discipline than before? Access is merely the threshold. Judgment is the work. Interpretation is the work. Synthesis is the work. AI has not removed that burden from me. It has placed it on me more heavily. Because I can now reach more perspectives, more contradictions, and more evidence with greater efficiency, I have fewer excuses for superficiality. I have to think better, not merely faster.
The Fear Is a Respectable Disguise for Avoidance
The narrative of AI-induced cognitive decline is, at bottom, a comfortable position. It locates the problem in the tool, which means the solution is restraint or regulation or slower adoption. It asks nothing hard of the person holding the argument. It preserves the idea that the person who refuses the tool is the serious thinker, the authentic intellectual, the college applicant who wrote her own essay. And while that story is moving, it is also a kind of flattery masquerading as philosophy. Much of the panic sounds less like a defense of human intellect and more like nostalgia for older inconvenience. It is, in many cases, a respectable disguise for avoidance: easier to indict the instrument than to confront the harder question of whether deep thinking was happening in the first place.
The tool does not determine the outcome on its own. The decisive variable is the standard, judgment, and discipline the user and the institution bring to it. A person who brings genuine intellectual hunger to an AI system will find that system amplifies that hunger into something formidable. A person who brings intellectual passivity will find the system confirms that passivity. That is the divide that matters. Not human versus machine, but passive user versus active thinker. AI can absolutely become an answer machine for the intellectually lazy. It can also become a research accelerant for the intellectually ambitious. The difference is not in the software. The difference is in the standard brought to it.
Leaders and organizations that treat this moment as an occasion for philosophical hand-wringing about the death of curiosity are engaging in a very expensive form of avoidance. The real obligation is harder: it is the design of systems, incentives and cultures that reward genuine understanding regardless of what tools people use to arrive at it. Cooke herself gestures toward this at the end of her essay, but the gesture is brief. The cultural diagnosis takes four pages. The design prescription takes two sentences. Describing the danger is easy. Building the architecture that produces better thinkers inside AI-enabled environments is where the actual leadership challenge lives.
We Have Navigated This Before
The velocity of AI adoption is genuinely unprecedented. There is no prior technological transition that moved at this speed across this many domains simultaneously. It would be dishonest to pretend otherwise, and the people who point to the printing press as a clean analogy are understating the magnitude of what is happening. But the underlying human dynamic is not new. People have always had to renegotiate their relationship to their own cognitive labor when a tool took over some portion of it. We did this with writing, which Socrates warned would destroy memory. We did it with the printing press, which terrified the scribal class. We did it with the calculator, with databases, with search engines. In every case, the catastrophe failed to arrive on schedule.
We adapted then. We will adapt now.
What actually happened in each of those transitions is that the humans who adapted most successfully understood that the tool changed what mattered, not whether thinking mattered. The printing press did not eliminate the need for serious readers; it created an enormous demand for them. The calculator elevated the value of knowing what calculation to perform and how to interpret the result. AI will not eliminate the need for original thought, synthesis, judgment or curiosity. It will dramatically raise the return on those capacities for the people who possess them. Yes, the velocity is different. Yes, the stakes are higher. Yes, AI can amplify sloppiness, dependency, and false confidence. But none of that means we are on the verge of surrendering ourselves to the machine. That kind of theatrical despair is not wisdom. It is posture. The people who will struggle in this moment are not the ones who use AI aggressively. They are the ones who refuse to engage with it and therefore never develop the discipline of directing it, interrogating its outputs, and identifying its failures. The real task is not panic. The real task is mastery.
The Actual Obligation of Leaders
Cooke is right that leaders have a design obligation, and that obligation deserves more than the brief acknowledgment she gives it. The obligation is not to slow AI adoption or to create sanctuaries from the tool. The obligation is to build environments where the quality of thinking is visible and rewarded, where people defend conclusions rather than merely produce them, where the origin of an insight matters less than the depth of understanding that surrounds it.
That means leaders must become skilled interrogators of their own people. They must ask not only what the recommendation is but how the person arrived there. What alternatives did you test? What evidence changed your mind? Where did the model mislead you, and how did you catch it? What did you reject, and why? These are not rhetorical questions. They are the evaluative standard that separates organizations where AI produces judgment from organizations where AI produces output. The deck is the artifact of the process. The reasoning is the asset. Leaders who cannot distinguish between them will build organizations full of polish and empty of thought.
And leaders must model this themselves. The most important signal in any organization is not the policy. It is the behavior of the senior person in the room. If that person uses AI to think harder, pursues arguments further than previously possible, and demonstrates what disciplined AI-assisted reasoning looks like in practice, the organization will follow. If that person uses AI to save forty-five minutes on a report no one will read closely, the organization will follow that instead.
We Will Not Succumb to the robots
The robots are not coming for our curiosity. Our curiosity was always ours to defend or abandon. The people who abandon it will find AI a convenient vehicle for that abandonment, just as they previously found Google, television and the cable news cycle convenient vehicles for the same. The people who defend it will find AI the most powerful amplification of that curiosity in the history of intellectual work.
The choice, as it has always been with every significant tool in human history, belongs entirely to the human holding it. The essay form has never written itself. The argument has never made itself. The synthesis has never arrived without a mind willing to do the difficult, irreplaceable work of sitting with a hard question long enough to actually understand it. For me, the technology has not reduced the need for thought. It has sharpened it. It has made me more responsible for what I conclude, not less. Because now I can test more claims, examine more objections, and gather a broader field of evidence before I speak. That is not the death of curiosity. That is curiosity under pressure. Curiosity with range. Curiosity with consequence.
And perhaps that is what unsettles people most. AI is exposing a truth many would rather avoid: the value of a mind was never in how long it took to find information. Its value was always in what it could do with what it found.
The future does not belong to the person who refused the tool. It belongs to the person who mastered it and never once confused that mastery with the end of thinking.
Written in response to “AI, curiosity, and the work we’re quietly losing” by Kendra Cooke, SYPartners, February 17, 2026.