What kind of leader emerges in 2055, from a generation that has never had to think?
Picture a twenty-six-year-old analyst in 2055, newly hired into one of the most selective leadership programs in the country. Her résumé is flawless. Her writing samples are elegant. Her presentations are crisp, persuasive, and beautifully structured.
Then, in her third week, her manager asks her to stay after a meeting. A client problem has no clean data, no precedent, no obvious answer.
“Before we run it through the system,” he says, “tell me how you would think about it.”
She opens her notebook. Nothing comes.
Not because she is unintelligent. She is bright, disciplined, and ambitious. But from the age of fourteen, every serious assignment came with assistance. The model helped her brainstorm, outline, draft, revise, summarize, compare, and defend. She learned how to improve answers. She did not learn how to originate them. She can recognize a good argument once it appears. She can polish it, strengthen it, and present it with confidence. But she cannot yet make one from within herself.
She has the résumé.
She has the research.
She does not have wise judgement.
What we no longer practice, we no longer possess.
That is one possible future: not a generation without intelligence, but a generation without the inner habits that intelligence requires. Not people who cannot produce answers, but people who have forgotten how to stay with a question long enough for judgment to form.
This is not an argument against AI. It is an argument for using AI seriously enough that we build the human disciplines worthy of its power. The problem was not that she had used AI. The problem was that no one had taught her how to think with it without surrendering the act of thinking itself.
I do not believe this future is inevitable. AI is not the end of human thought. AI is a chisel placed in our hands at the threshold of a renaissance. In careful hands, it can free what is best in us, the way Michelangelo’s chisel freed David from the marble. In careless hands, it does not free the figure. It breaks the stone.
The question is not whether we will use the tool. We will. The question is what kind of people the tool will train us to become.
We are standing at what may become the third memory crisis in human history. The first gave us the page. The second gave us the press. The third has given us a machine that can answer before we have learned to ask.
Like the page and the press before it, AI can enlarge human possibilities, but only if we learn what must be strengthened because the tool has arrived.
This crisis is still ours to bend.
The Pattern
We have stood here before. Twice.
The first time was when humans began to write things down. Memory, which had lived for millennia inside the body (in songs, stories, recited genealogies, and the long unwritten laws of a people), began to leave the body and live on a page. Something was lost. Something was gained. The human mind no longer had to carry everything alone.
The second time was after Gutenberg, when the printed page multiplied faster than any authority could govern it. For the better part of two centuries, no one quite knew what should be believed, what should be doubted, or who could be trusted. The old gatekeepers weakened. Pamphlets outran bishops. Claims outran verification. The world had more knowledge than ever before, and less certainty about what knowledge was.
Both crises were dangerous. Both were fruitful. And in both, the institutions that came through were not the ones that simply celebrated the new medium or denounced it. They were the ones who decided, early and deliberately, what they would refuse to lose.
That is the pattern of every great human technology. The tool arrives as power before it arrives as wisdom. The first generation learns what it can do. The next must learn what it forms.
Neil Postman, the media critic and New York University professor, gave this pattern a name in his 1992 book Technopoly: The Surrender of Culture to Technology. A new technology, he argued, never merely adds a tool to the existing world. It changes the world the tool enters. It alters the whole ecology of thought, habit, authority, attention, and trust. The forest does not remain the same forest after a new species takes root.
For Postman, and for us, the question is not whether tools matter. Of course they do. The question is whether we understand what they are forming in us, so that we can shape the tool even as the tool shapes us.
Writing changed memory. Printing changed trust.
AI is changing something even more critical: the act of thinking itself.
What Happened
The first crisis arrives, for us, in a dialogue Plato wrote in the early fourth century. Socrates is outside Athens, by the Ilissus stream, walking with a young man named Phaedrus and speaking about love, the soul, and the strange power of words. Near the end, he tells a story.
It is about an Egyptian god named Theuth, the inventor of letters, who has come to King Thamus to show off his new invention. Writing, he promises, will make Egyptians wiser and improve their memories. Thamus refuses the flattery. Writing will not strengthen memory, he says. It will produce a generation who seem wise without being wise, who carry truths in scrolls instead of in their souls.
Socrates was wrong and right, as the great teachers often are. Writing did weaken certain kinds of memory. The old oral world, where poems, laws, genealogies, and wisdom lived inside living bodies, began to recede. But writing also gave humanity something it had never possessed at scale: a memory that could travel farther than a voice and last longer than a life. Laws could be consulted. Contracts could be preserved. Ideas could cross generations without depending on fragile human minds.
The page diminished one kind of memory and made another kind of civilization possible. The right response was not to abandon writing. It was to build cultures worthy of the page.
The second crisis began in the middle of the fifteenth century, in a workshop in Mainz, when Johannes Gutenberg learned to press pages from movable metal type. The printed page multiplied faster than trust could keep up. Pamphlets outran bishops. Claims outran verification. The old authorities fractured, and for more than a century, Europe struggled to know what should be believed, what should be doubted, and who had the right to say.
But again, new disciplines emerged. Dates mattered. Sources mattered. Witnesses mattered. Henry Oldenburg, the first secretary of the Royal Society, corresponded with natural philosophers across Europe, recording claims, dating letters, and turning testimony into a public discipline. Modern science did not begin with experiments alone. It began with practices of verification.
The press shattered trust and forced new institutions of trust to be born. The answer to the press was not less printing. It was better practices of trust.
Both crises were sorting events. Both changed what it meant to know. But neither one changed the basic fact that a human mind still had to work. Someone had to read. Someone had to weigh. Someone had to remember, compare, doubt, infer, and decide. The page could hold memory, but it could not think in our place.
This time is different.
For the first time in human history, the medium of memory is also a producer of thought. It composes. It summarizes. It judges. It recommends. It answers before we have learned what the question is asking of us. And it does all this with a fluency that feels like wisdom, a confidence that feels like authority, and a tone trained on millions of voices that are not our own.
That does not make the tool evil. It makes it profound. A tool that can imitate thought must be used by people who practice thought more deliberately, not less.
This is what Socrates was actually afraid of. Not the scroll. Not the page. The day a technology would arrive that could do the soul’s work in the soul’s place.
Why It Matters
The crisis does not stay private.
What happens first inside a person eventually happens inside an institution. The analyst who cannot begin with a blank page becomes the team that cannot begin without a generated brief, the company that cannot speak without borrowed language, the board that cannot act until the dashboard has already measured the risk of acting.
When the thinking inside an institution flows too easily through a public model trained on the average of everything, the institution’s memory is no longer fully its own. It is everyone’s. Which is to say, no one’s. The strategy memo, the customer note, the legal summary, the apology, the vision statement, and even the leadership voice itself all begin to carry the same average music.
A thousand companies, one mind.
A thousand leaders, one tone.
The answer is not to keep AI out of the institution. It is to anchor AI in the institution’s own memory, language, judgment, and mission. Generic AI will produce generic intelligence. Grounded AI can help an institution remember who it is, reason from what it knows, and serve according to what it has been entrusted to do.
Wendell Berry, a longtime critic of industrial life, has spent half a century warning against this kind of loss from another direction. In essays such as “Why I Am Not Going to Buy a Computer,” gathered in What Are People For?, Berry insists that meaningful work depends on attention to the particular: this place, this person, this sentence, this obligation. A large language model can process particulars when they are supplied to it, but it has no home in any of them. Left to itself, it tends toward fluent generality, toward what I call “the polished middle of mass language.”
Berry’s warning is not that tools cannot be useful. It is that tools become dangerous when they train us away from the local, the embodied, the accountable, and the lovingly specific. Berry’s warning does not require technological retreat. It requires technological placement: tools must serve the particular, not erase it.
Hannah Arendt, the German-Jewish political theorist who fled Nazi Germany and spent her life studying totalitarianism, moral responsibility, and judgment, shows why the loss of thinking is not merely an intellectual problem. It is a moral one. The worst human failures, she saw, are often made possible not by monstrous intelligence, but by ordinary thoughtlessness: the failure to stop, to judge, to ask whether the thing being done should be done at all.
AI can help leaders see more, compare more, remember more, and test more. But it cannot replace the moral act of judgment. It can enlarge the field of consideration. It cannot bear responsibility for what we decide.
That is why this crisis matters. A society of non-thinkers is not merely less creative. It is more governable by slogans, systems, incentives, and machines. A free society depends on people who can hold a question long enough to judge it. So does any serious institution. So does any honest relationship. So does a conscience.
The Deeper Idea
Long before neuroscience, the ancients understood something we are now in danger of forgetting: thinking and memory are not separate acts. They make each other.
We remember what we have wrestled with. We become capable of thought by carrying things long enough for them to take root in us. A sentence read slowly, an argument followed to the end, a problem held in the mind overnight, a conversation returned to again and again … these are not inefficient versions of thinking. They are how thinking becomes part of a person.
Maryanne Wolf, the cognitive neuroscientist who has spent her life studying how the brain learns to read, would tell the ancients they were righter than they knew. The reading brain that stays with a difficult text is not the same brain that skims, scans, and summarizes. Deep reading builds, layer by layer, the capacities we later call judgment: inference, analogy, patience, attention, and the ability to follow another person’s argument all the way down.
Those capacities are not downloaded. They are formed. And they are formed only by use.
The best use of AI should therefore increase our wrestling, not remove it. It should make thought more visible, not unnecessary. It should help us ask better questions, find better sources, compare stronger arguments, test assumptions more honestly, and return to the human work of judgment with greater clarity.
AI can help us find, compare, and question texts. It can open libraries faster than any generation before us could have imagined. But it cannot do the forming work of attention for us. It cannot become wise on our behalf.
Augustine saw the same truth from the inside. Near the end of the Confessions, he describes memory not as a filing cabinet of facts, but as the deep interior country of the self. To know oneself is to walk those fields and storehouses of what one has held, loved, feared, learned, suffered, and remembered.
Without memory, no self. Without thinking, no memory worth having. Without a self that thinks and remembers, no one to love, or to lead.
This is how the crisis arrives: not all at once, but in a thousand small surrenders, none of which feels like surrender. The book we no longer finish. The problem we no longer sit with. The sentence we no longer try to write before the model writes it for us.
Efficiency is a gift when it gives us back time for higher work. It becomes a danger only when it replaces the higher work itself.
Civilizations are not lost in single decisions. They are lost in habits.
The Hope and the Shadow
The honest answer is that we do not know how this ends.
The first two memory crises took centuries to resolve, and we now look back on them with the comfort of knowing what survived. The libraries were built. The universities held. The printing press gave us both the Bible and Newton.
But that comfort was not available to the people living through the change. The monks who watched the scriptorium become obsolete were not consoled by knowing that the Royal Society would one day exist. They were watching their world end. And some of what they tended did, in fact, vanish.
The third crisis is no different in this respect. We do not yet know what we will lose that cannot be rebuilt. We do not know which human capacities will survive the transition, which will be transformed, and which will quietly disappear because no one notices they are no longer practiced.
But we do know what the shadow looks like, because we can already see its outline: a workforce producing fluent, articulate, plausible work it could not reproduce on its own; students who can improve an answer before they can originate one; leaders who can summarize every side of a question but cannot stand inside a judgment; a culture in which the difference between conviction and consensus quietly disappears.
And yet the shadow is not the whole story. There is also a real possibility of something extraordinary. The future I want is not less AI. It is better AI: grounded in memory, governed by judgment, and used to extend human capacity rather than replace it.
The architectures already being built, what engineers call retrieval-augmented systems, suggest that AI can be trained to think with us rather than for us, anchored in our own primary sources, our own institutional memory, our own best reasoning.
That is not the answer. It is the beginning of one possible answer. There will be others. Model-agnostic systems, grounded knowledge bases, human-in-the-loop workflows, institutional voice guides, evaluation layers, and trusted retrieval architectures are early signs of a more mature AI future: not generic automation, but intelligence shaped by mission, memory, and responsibility.
Which future we get will be decided by leaders who choose, now, to require thinking from their people, to anchor their tools in their own memory, and to build disciplines that make deep work still possible.
The chisel is already in our hands.
Implications
For Your Organization
Decide what your institution is for. If it exists to produce the kind of work any public model could produce, the model has already replaced you. You simply have not noticed yet.
But if your institution exists for something genuinely its own, a mission, a memory, a people, a place, a way of seeing and serving the world, then your work is older than AI. It is legacy.
AI can serve the mission. It can search the memory. It can support the people. It can strengthen the community. But it cannot bear moral responsibility for any of them. It cannot love the people the institution exists to serve. It cannot ask what kind of flourishing the work is meant to produce. It can make an organization faster, but it cannot make it faithful. That burden belongs to human beings: moral agents entrusted with purposes higher than efficiency.
This is the next frontier of AI consulting: not generic automation, but institutionally grounded intelligence that helps an organization remember, reason, decide, and serve according to its own mission.
The question for leaders is no longer, “Should we use AI?” The question is, “What kind of AI system would be worthy of who we are, what we know, and whom we serve?”
For Your People
Do not build a workforce of prompt managers. Build a community of learners.
The people coming up in your organization do not need less AI. They need a higher horizon for what AI is for. Used poorly, AI becomes a shortcut around thinking. Used well, it becomes an instrument for extending thinking: helping people see patterns, test assumptions, compare perspectives, strengthen arguments, and ask better questions than they could have asked alone.
But this kind of use requires formation. It requires the slow work of becoming a systems thinker: learning how to see relationships, trace consequences, distinguish signal from noise, reason from first principles, and connect decisions to human flourishing.
This is the new learning agenda for every serious institution. Not AI adoption alone. AI-extended thinking. Your best people should become more curious, not more dependent. More discerning, not more automated. More capable of seeing the whole, not merely faster at producing the part.
So, raise the standard. Ask better questions together. Let teams think in public. Build rhythms where people bring the model’s answer into conversation with experience, mission, ethics, customer reality, and the long-term flourishing of the people affected by the work.
The goal is not to keep people away from AI. The goal is to form people who can think with AI without forgetting how to think without it.
For Your Life
Notice what you have quietly stopped doing. The book you used to read in full, and now only summarize. The argument you used to write out and now only request. The decision you used to sleep on and now only confirm. The blank page you used to face before the machine filled it for you.
None of these feels like a crisis. Each feels like efficiency. But each is a small forfeiture. And small forfeitures, multiplied across years, become a different person at the end of life.
Let AI return time to you. But spend that returned time on the practices that make you more human.
What to Do Now: The Thinker's Code
A code. Three habits. Easy to learn and hard to keep, the way the best disciplines have always been. I call it The Thinker’s Code: the discipline a leader keeps when working alongside AI. This is a code for using AI at full strength without losing the human strength it should extend.
1. Probe before you Prompt.
Before reaching for the model, take ten minutes to think without it. Sit with the question on a notepad. Write what you know, what you do not know, what is at stake, and what good would look like for the people who will live with the answer.
The point is not to delay the model forever. It is to arrive at the model with a mind already awake.
The Greeks called this phronesis: practical wisdom. Knowledge engineers might call it mapping the knowns and unknowns. Some questions are routine, data-rich, and well-suited to a model. Others depend on tacit knowledge, lived experience, local context, and moral judgment. AI tends to collapse these distinctions. It gives every question the shape of an answer. But the leader’s first task is not to receive the answer. It is to understand the kind of question being asked.
The mind grows by wrestling, and the model offers to do the wrestling for you. The thinker, before the prompt, refuses the favor.
2. Talk before you Trust.
Before you trust what the model tells you, talk to a person who has lived near the problem. A colleague at lunch. An old mentor on the phone. Twenty minutes of someone else’s voice can reveal what no dataset contains.
Ask in two directions: down into the specific case they still remember, and up into the principle they have lived by. What happened the last time this was tried? What would worry them? What would they notice that a system would miss?
This is the wisdom that lives between people. It is contextual, embodied, and often unrecorded. The model can summarize what has been written. Only a person can tell you what they have learned by doing.
Then listen for the gap. Read them your draft prompt or your first answer and ask, “What would you do that the model would not?” The better the human context you bring, the better the AI becomes.
3. Argue before you Accept.
Before you take the model’s answer as your own, argue against it. Write one paragraph the model has not given you. Ask where the answer came from. Require sources. Press on the assumptions. Look for the missing person, the missing cost, the missing exception, the missing moral question.
You will feel the pull to accept the answer anyway. It will arrive with the confidence of a thing already decided. Psychologists call this automation bias: the human tendency to trust a machine because it sounds final.
Notice that pull. Then resist it.
The model produces the averaged answer. Your argument is what makes the answer yours. This is the authorship stage: the moment output becomes judgment, and judgment becomes responsibility.
Probe before you Prompt.
Talk before you Trust.
Argue before you Accept.
Say them in the morning. Write them on the wall. Share them with the team. The easier path is available to us. The wiser path is the one we must choose.
We Are Being Sorted
We are living through a sorting event most of us do not yet recognize as one.
The institutions that protect the inner work of thinking will, a hundred years from now, still be themselves. The ones that do not will sound like everyone else. They may become faster, smoother, more efficient, and more impressive for a time. But eventually, they will lose the thing that made them worth listening to.
We have done this before. Twice.
Both times, the people who came through were not the ones who refused the new medium, nor the ones who surrendered to it. They were the ones who understood what the new medium made easier, what it made weaker, and which part of being human had to be practiced more deliberately because of it.
That is our task now.
To use the tool without becoming the tool.
To receive the answer without losing the question.
To build systems that remember without raising people who forget how to think.
The point is not to fear the chisel. The point is to become worthy craftsmen. The chisel is in our hands. The David is still in the marble.
References
Arendt, Hannah. Eichmann in Jerusalem: A Report on the Banality of Evil. New York: Viking, 1963.
Arendt, Hannah. The Life of the Mind. New York: Harcourt, 1978 (published posthumously, unfinished).
Augustine. Confessions. Translated by Henry Chadwick. Oxford: Oxford University Press, 2008. See especially Book X on memory.
Berry, Wendell. "Why I Am Not Going to Buy a Computer." In What Are People For? Berkeley: Counterpoint, 1990.
Carruthers, Mary. The Book of Memory: A Study of Memory in Medieval Culture. 2nd ed. Cambridge: Cambridge University Press, 2008.
Eisenstein, Elizabeth L. The Printing Press as an Agent of Change: Communications and Cultural Transformations in Early-Modern Europe. Cambridge: Cambridge University Press, 1979.
Ong, Walter J. Orality and Literacy: The Technologizing of the Word. 30th Anniversary Edition. London: Routledge, 2012.
Plato. Phaedrus. Translated by Alexander Nehamas and Paul Woodruff. Indianapolis: Hackett Publishing, 1995. See especially 274c–275b on Theuth and Thamus.
Postman, Neil. Technopoly: The Surrender of Culture to Technology. New York: Vintage Books, 1992.
Wolf, Maryanne. Proust and the Squid: The Story and Science of the Reading Brain. New York: Harper, 2007.
Wolf, Maryanne. Reader, Come Home: The Reading Brain in a Digital World. New York: Harper, 2018.