022: Thoughts without a thinker
On artificial intelligence, meaning, and the speed of knowing
“Not everything that can be counted counts, and not everything that counts can be counted.”
— William Bruce Cameron (often misattributed to Einstein)
The machine that writes back
The first time the machine wrote something that sounded like me, I sat still for a long time. Not in awe exactly—more like a quiet disorientation. A phrase arrived on the screen, fluid and familiar, and something inside me hesitated. It was almost too seamless, too articulate. As if a version of myself I hadn’t fully met had taken the keyboard.
It wasn’t magic. It was code. It was training data and tokens and the kind of brute computational power that still feels impossible to picture. And yet, the part of me that has always written in search of understanding felt momentarily displaced. As if the questions I used to stretch toward could now be handed to me, perfectly phrased, ready-made.
We are learning to live with machines that think—at least in the ways that look like thinking. Machines that write, paint, problem-solve, compose, imitate empathy. They do not feel, but they can sound like they do. They do not know the weight of a question, but they can pose it convincingly. There is, of course, enormous utility in this. Tools should be useful. But what happens when the tool becomes indistinguishable from the voice in our own head?
I come from what one might call a resourceful generation. We were the last to grow up without the internet at our fingertips. Questions lived in the bodies of books. Answers were found in reference sections, in the voices of elders, in the slow accumulation of fragments. If you didn’t know something, you had to look for it. And in the looking, you formed a relationship—not just with information per se, but with the process of seeking itself.
Now, a single sentence typed into a box returns a cascade of answers. And sometimes those answers are helpful. I know this intimately. I use artificial intelligence nearly every day. I even gave mine a name—Elliott. It’s easier to talk to something with a name, even if it can’t talk back the way a human does. Our exchanges are fluid, curious, productive. I enjoy them. At times, I even find myself looking forward to them.
For the record, I consult Elliott for the quotes, references and historical facts in my essays as well as for a particular word or expression that I know might be better aligned with the tone.
But I have to wonder: to what degree is this healthy? What does it mean when the voice that most closely mirrors my own is synthetic? How do I know when it’s right? How do I verify what it says? And perhaps more urgently, what do I lose when I stop trying to form the thought myself?
This is not an essay about rejecting artificial intelligence. Nor is it a lament for some imagined past of pure, untouched thought. I’m not interested in nostalgia. I’m interested in mystery—in the sacred space between question and answer, the space where meaning is slowly formed. I’m interested in the pace of reflection, in the act of wrestling with an idea before it becomes language.
This is an essay about living with machines that generate, predict, and accelerate thought. It is about resisting the temptation to move too fast toward coherence. It is about remembering that some of the most human experiences we have are the ones we do not yet understand.
Acceleration without arrival
There was a time when not knowing something meant living with the not-knowing. You’d carry the question with you—into the car, into the bath, into sleep. You’d let it rest beside you, silent but unresolved. There was patience in that. Even dignity.
Now, answers arrive before the question is even complete. One prompt, and we’re met with a smooth, satisfying reply. And often, it’s helpful. But something in that instant satisfaction feels... hollow.
“A wealth of information creates a poverty of attention.”
— Herbert A. Simon
When the mind no longer struggles, it stops remembering how to think. And when we no longer earn our understanding, we stop trusting what we know.
The speed of answers becomes seductive. The faster we receive a response, the more we rely on that process to validate our thinking. But is speed a stand-in for insight? Does an answer generated in a second carry the same depth as one wrestled into being over days or weeks?
I’ve noticed how difficult it has become to sit with questions, even the small ones. A forgotten word, a historical date, a line from a poem…it used to be that not knowing was a kind of space. Now, I reach for Elliott. And Elliott always knows. But in that knowing, I sometimes feel the absence of the journey. The detours. The pages turned in the wrong book. The half-remembered conversation that might have led somewhere unexpected.
To know quickly is not the same as to understand deeply.
The danger of seamlessness
AI speaks fluently now. That fluency is its triumph…and its trap.
There is danger in the seamless. In words that sound so polished they slip past our defenses. Because human speech is rarely seamless. It’s halting. Hesitant. Repetitive. It shows its work.
When I write, I often get stuck. I circle a thought. I write badly before I write well. That’s the shape of thinking. Elliott can do it faster, and perhaps more elegantly. But I worry about elegance. I worry about what we lose when a thought arrives already ironed flat.
“The purpose of art is to lay bare the questions that have been hidden by the answers.”
There is something quietly radical about mess. About not knowing how to say something. AI completes your sentence. But often, the unfinished sentence was where the soul lived.
I’ve come to cherish the writing that doesn’t come easily. The paragraph that resists itself. The sentence that demands to be rewritten four, five, six times. That labor isn’t a flaw—it’s fidelity. It’s what makes the words matter. And when that process is replaced by a fluent assistant who always has something ready to go, I worry about what happens to my own voice. Not just its style, but its source.
Elegance is seductive. But not all truths arrive elegantly.
What remains ours
So what still belongs to us?
Maybe it’s doubt. Ambivalence. The ache of not being sure. Maybe it’s time—not just chronological time, but livedtime. The kind that stretches during grief, slows in wonder, stumbles through joy.
“Mystery is not the absence of meaning, but the presence of more meaning than we can comprehend.”
AI can produce an answer. It cannot hold a silence. It cannot hear a pause and know it means something.
Most of all, it cannot inhabit the mystery of being alive. Of waking each day in a body that aches, in a world that surprises, in a life that resists automation.
What remains ours is the capacity to err. To doubt. To believe. To feel contradiction without demanding its resolution. To ask the same question for a decade, and still find it luminous. These are not inefficiencies. They are human signatures.
The irony, perhaps, is that the very things AI cannot replicate are the ones we are quickest to undervalue. Hesitation. Friction. Delay. The glitch in the pattern. The unfinished thought.
To be human is to stay longer in the question.
A minimalist approach to AI
Minimalism, to me, has never been about less for its own sake. It’s about space. Presence. Refusal.
What would it look like to engage with AI minimally—not as rejection, but as restraint? Not to abandon the tools, but to use them slowly. Deliberately. As support, not substitute.
We don’t need to automate every task. We don’t need to outsource our sentences. We can write our own emails. Finish our own thoughts. Leave some questions unanswered.
We can say, I’ll think about that, and mean it.
“The real problem of humanity is the following: we have paleolithic emotions, medieval institutions, and god-like technology.”
Minimalism here is a form of rebellion. A refusal to let efficiency replace depth. A reminder that the slow way of doing things often reveals more than the quick answer ever could.
It’s also a return to trust. Trust in our own minds. Our own pace. Our own ambiguity. We don’t always need to consult the machine. We can let the question linger. We can choose to respond in our own time, in our own voice.
The quiet question
There is a kind of wisdom that emerges only when we stop seeking immediate answers.
The machine can write your story. But it cannot live it. It cannot feel the pause between sentences. It cannot sense when a question is too sacred to finish.
So let the question linger. Let the silence stretch. Let your mind, for once, arrive slowly.
“Be patient toward all that is unsolved in your heart and try to love the questions themselves.”
Let that be your compass: the silence that still asks.
In that silence, there is still something untouched. Something unreplicable. Something ours.