In 1864, Dostoevsky published a novella in two parts that would go on to unsettle every generation of readers that encountered it. The first part is a monologue delivered by a forty-year-old former civil servant ranting from his apartment in St. Petersburg, addressing an audience he simultaneously craves and despises. He is spiteful, contradictory, and fully aware of both qualities. He tells us he is sick and refuses to see a doctor, not because he can't afford one but because treating the illness would be the rational thing to do. Rationality is precisely what he has set himself against.

The underground man is responding to a specific intellectual current of his time. The Russian nihilists and utopian socialists like Chernyshevsky believed that human beings, once educated about their true self-interest, would naturally build a perfect society. Chernyshevsky called this vision the crystal palace, borrowing the name from the actual Crystal Palace built in London for the Great Exhibition of 1851, a monument to industrial progress and rational design. The underground man takes this image and turns it into an indictment. His objection, the thing that makes his argument so difficult to dismiss, is that the palace would actually work, that it would deliver on every promise, and that this would be the worst possible outcome for the people living inside it.

"What sort of free will is left," he asks, "when we come to tabulation and arithmetic, when it will all be a case of twice two makes four?" His answer is that a person confronted with this perfect calculus would choose to act against it on purpose, would insist that two plus two equals five. Not out of ignorance, but as an assertion of the only thing the system cannot account for: the capacity to refuse. The argument runs deeper than a simple defense of personal freedom. What he is really getting at is that the capacity to choose badly, to act against one's own interest, to suffer when suffering is entirely unnecessary, constitutes the foundation of consciousness itself. A system that forecloses on that capacity has not liberated anyone. It has constructed a cage whose inhabitants have been persuaded to mistake comfort for flourishing.

The second part of the novella dramatizes this philosophy through a series of painful encounters. The underground man attends a dinner party for a former schoolmate he despises and humiliates himself spectacularly, oscillating between grandiosity and self-loathing in a way that is almost physically uncomfortable to read. Later, he meets Liza, a young sex worker, and delivers a long, seemingly compassionate speech about the trajectory of her life, moving her to tears. When she comes to him seeking the genuine human connection he appeared to offer, he responds with cruelty. Real intimacy would require vulnerability, and vulnerability would mean surrendering the control that his underground existence is designed to protect. He chooses isolation and spite over connection and warmth. Dostoevsky's point is that this choice, however ugly, is authentically his in a way that the rational alternative would never be.

I think about this constantly in the context of what artificial intelligence promises to become. Not the tools as they exist today, but the vision that animates their development. A world without mundane work, where personal agents handle your decisions before you've fully articulated them, where commerce and communication and even thought itself have been stripped of all resistance. The pitch is always the same: we will remove the things that slow you down, and you will finally be free. But the question Dostoevsky would press is whether the friction was doing something important that we failed to notice. Whether the things that slow us down were also the things holding certain kinds of meaning in place.

The space where meaning forms

I read physical books and often write things down by hand on paper. The digital equivalents are cheaper, more searchable, and more efficient by almost every measurable standard. The feeling of reading something printed and writing something with a pen still outweighs whatever I would gain from the optimized alternative. For a long time I dismissed this as a personal quirk, maybe even a kind of nostalgic stubbornness. Dostoevsky helped me see it as something else entirely.

The friction itself generates meaning in ways that are difficult to articulate but impossible to ignore once you've noticed them. The slowness of handwriting forces you into a different relationship with your own thinking, one where the pace of the hand constrains and shapes the movement of the mind. The weight and texture of a physical page creates a kind of presence that a screen does not replicate. These qualities are not inefficiencies waiting to be solved. They are part of the cognitive experience itself, and stripping them away in the name of speed transforms the process into something fundamentally different. Something that may be more productive in a narrow sense but that has quietly lost the thing that made it worth doing.

Dostoevsky believed that suffering was generative. The struggle to understand, to create, to reach another person across the full distance of human difference was the place where meaning actually took shape. Not in the outcome, but in the labor of arriving there. The underground man's encounter with Liza is the purest illustration of this. He has the opportunity for real human connection, the kind that requires effort and risk and the willingness to be changed by another person. He destroys it because the frictionless alternative, retreating back underground, is easier. The tragedy is that he knows exactly what he is doing. When AI collapses the distance between intention and output, when a rough thought becomes polished prose in seconds, the time saved is real. But so is the elimination of the space where understanding used to develop. We may not recognize what has disappeared until the disappearance is well behind us.

The optimization treadmill

Here is where I find myself genuinely uncertain, and I think the honest thing is to sit with that uncertainty rather than resolve it prematurely. Optimization is deeply human, possibly the thing we are best at. When we run out of meaningful problems we begin optimizing the solutions to old ones. There is something beautiful in that impulse, something that reflects our constitutional restlessness and our refusal to accept the world as we find it.

The standard argument holds that optimization gives us more time for the things that matter, that automating tedious work frees people to think, create, and connect with one another. I want to believe this, but I'm not convinced it describes what actually happens. The time we save seems to fill immediately with further optimization, each automated task revealing three more that could be automated in turn. The promise is always liberation. The result looks more like a treadmill whose acceleration we have learned to call progress.

The underground man would understand this treadmill perfectly because he lives on one of his own. His monologue circles and doubles back on itself endlessly, each thought generating a counter-thought, each justification spawning a new grievance. He is optimizing his misery with the same relentless energy that the rationalists apply to their utopia, and the result in both cases is a kind of motion that never arrives anywhere. Dostoevsky never resolved this tension, and I find some comfort in that. He simply kept pointing at it, kept insisting that his readers look carefully at the machinery they were constructing and ask themselves whether they truly understood what it would cost them. Nobody knows the full consequences that AI will have for human society. Not the engineers building these systems, and not the critics warning against them. That uncertainty deserves to be taken seriously on its own terms rather than smoothed over with either optimism or dread.

Originality and the discipline of refusal

The effort required to create something genuinely original feels like it has never been higher, while the barrier to producing something that looks original has never been lower. Everyone now has access to tools that generate text, images, music, and code with remarkable fluency. But the output draws from patterns in existing human work, which means it tends to converge toward a kind of sophisticated average of everything that came before. Originality, almost by definition, lives outside that average, in the places where pattern recognition breaks down and something ungovernable emerges.

There are people who refuse to use these tools entirely, and I have deep respect for that choice. I think it reflects something Dostoevsky would have recognized immediately. Refusal, in this context, functions as a form of creative discipline, a deliberate decision to preserve the conditions under which original thought remains possible. The underground man's refusal leads nowhere obviously productive and contributes nothing measurable to the world around him. Yet it stands as the most honest gesture in the entire novel, because it insists that certain things matter more than efficiency, more than output, more even than outcomes. Dostoevsky lets him be ugly, petty, and self-destructive, and still somehow right about the one thing that matters: that a person who cannot say no to the system has already ceased to be a person in any meaningful sense.

The question I keep returning to is not whether AI is good or bad, or whether automation represents a gift or a curse. I think that framing already concedes too much to the rationalist worldview, to the belief that everything can be weighed on a scale and sorted into tidy categories. The deeper question concerns what happens to us in the space between intention and action. In the struggle and the friction and the uncertainty that Dostoevsky spent his life insisting we protect. Whether we are willing to defend that space even when someone offers, convincingly and with the best of intentions, to make it disappear. Whether we understand, as the underground man understood in his bitter and luminous way, that the crystal palace, for all its beauty, was never designed with a door.