June 22, 2026

The Algorithm Cannot Meditate

The Algorithm Cannot Meditate Last month I sat for ten days in noble silence, watching my breath move in and out. On the eleventh morning I opened my laptop an...

The Algorithm Cannot Meditate

Last month I sat for ten days in noble silence, watching my breath move in and out. On the eleventh morning I opened my laptop and a language model greeted me by name, ready to help. I felt a small quiet ache I have not stopped thinking about since.

The Question I Keep Returning To

How did we get here? Not the technology itself — that story is easy to tell in decades of chips and datasets. I mean something stranger. How did we arrive at a moment where we ask machines to help us think, help us feel, help us decide who we are, and we do this without pausing to ask what a machine can actually know?

I do not want to argue against anything. I use these tools every day. My company depends on them. I am writing this partly with the help of one. But somewhere between the cushion and the keyboard, a question keeps surfacing: what is it that the algorithm cannot touch? And if I cannot answer that, how will I know when I have handed it something it was never meant to hold?

What the Machine Can and Cannot Reach

I want to be careful here. It would be easy to say machines are cold and humans are warm, machines are calculating and we are wise. That is not what I have found. The machines I work with produce astonishing things. They translate, they summarise, they suggest, they sometimes surprise me. If wisdom were only pattern recognition, they would already be wiser than most of us.

But sitting on a cushion for ten days teaches you something that pattern recognition cannot reach. It teaches you that awareness is not information. When I watch a sensation arise on my upper lip — a small tingling, a pressure, a heat — and I watch it pass, something is happening that has no output. No token is generated. No file is saved. The event is complete in itself. It changes me and leaves no trace the machine could read.

This is where I keep bumping against the limit. Not in what the algorithm produces, but in what it cannot even register as existing.

I have been trying, imperfectly, to sketch what different traditions have pointed at when they speak of this dimension of experience — the part that is prior to language, prior to concept.

Tradition	What It Points At	The Word Used	What Remains When Language Stops
Vipassana (Theravāda)	Direct sensation, moment to moment	vedanā	Pure awareness of arising and passing
Advaita Vedanta	The witness behind thought	sākṣī	The one who sees, never seen
Ayurveda	Intelligence in tissue	prajñā in dhātu	The body's own knowing
Zen	Mind before thought arises	mushin	No-mind, unconditioned presence
Sufism	The heart that perceives	qalb	Direct taste, dhawq
Taoism	Effortless action from stillness	wu wei	The uncarved block
Christian mysticism	Silence beyond concept	apophasis	The cloud of unknowing

I list these not to compare or rank them. I list them because it moved me to notice that across cultures that never spoke to each other, human beings kept finding the same edge. There is a place where language ends and something else begins. Every tradition I have studied points at it. Every serious practitioner I have met describes it, though none the same way twice.

The algorithm lives entirely on the other side of that edge. It lives in language, in pattern, in what has been said before. This is not a criticism. It is simply where it lives. Ask it about vedanā and it will produce a beautiful paragraph. Ask it to feel one, and the question does not compute — not because the machine is stupid, but because the question is not for it.

I have started to wonder whether the great confusion of our moment is that we mistake fluency for understanding. The machine is extraordinarily fluent. It speaks the language of every tradition I listed above. It can quote them, weave them, generate new sentences in their style. And so it is easy — so easy — to feel that it knows.

But knowing, in the sense that a meditator means it, is not a linguistic event. It is what happens when the linguistic event stops. My teacher used to say: you cannot read a menu and taste the food. The algorithm is the most complete menu ever written. I have been reading it obsessively for years. And I keep forgetting, in my hunger for it, that I have not eaten.

Perhaps this is what troubles me. Not that machines cannot meditate — of course they cannot, and there is no shame in that. What troubles me is that I catch myself using them as if they could, as if their fluent output about presence could be a substitute for the practice of it. As if reading about the breath could ever be the breath.

The Deeper Layer — Who Is Asking

There is a layer beneath even this. When I sit long enough, a stranger question appears. Not "what can the machine not do?" but "who is the one asking?"

Because the "I" that types the prompt, that reads the reply, that judges whether the reply is good — that "I" is also a kind of algorithm. It runs on memory, on conditioning, on training data collected since birth. It generates responses based on patterns. When I look closely at my own reactions, they are often startlingly automatic. A word arrives, an emotion arises, an opinion forms, and the whole process feels like watching a very sophisticated language model produce output.

So perhaps the real question is not whether the machine is like me. Perhaps it is whether I have understood what in me is not like the machine.

Vipassana pointed me toward this: there is something in the human being that is not pattern, not response, not conditioned reaction. The traditions call it many things. I do not know what to call it. But when I touch it, even briefly, I know it is not the part of me that argues with algorithms. It is the part that watches the arguing happen and is not moved.

That part, I suspect, is what cannot be automated. Not because it is powerful, but because it is not a doing at all. It is a being. And being, as far as I can tell, cannot be scaled.

What I Am Still Learning

I do not want to end this with a conclusion. I am not sure I have one.

What I notice is that the more time I spend with machines, the more grateful I become for the cushion. And the more time I spend on the cushion, the more clearly I see what the machines are, and are not, offering me. The two practices seem to need each other, at least in my life. Without meditation, I mistake the algorithm for a mind. Without the algorithm, I would forget how much of my own mind is just algorithm.

Perhaps the invitation of this moment is simply to stay awake to the difference. To use the tools without becoming them. To let the machine do what the machine does, and to keep something for ourselves that we do not outsource — not because it is precious, but because it is what we are.

I am still learning. I will probably be learning until I stop.

FAQ

Q: Are you saying AI is bad or that we should not use it? A: No. I use it daily and my company depends on it. I am asking a different question — what part of being human happens in a place the machine cannot reach, and how do we keep visiting that place?

Q: What is Vipassana, briefly? A: A meditation technique from the Theravāda Buddhist tradition, taught in silent ten-day retreats. It involves observing bodily sensations with equanimity, moment by moment. I am a student, not a teacher.

Q: Can a machine become conscious one day? A: I do not know. I am suspicious of anyone who says they do. What interests me more is whether I am fully conscious now.

Q: Why bring in Ayurveda and other traditions? A: Because they each point at something the modern conversation about intelligence tends to skip — that knowing happens in the body, in silence, in relationship, not only in the mind.

Q: Is fluency really different from understanding? A: I think so, though I hold this lightly. A parrot can be fluent. A teacher can be silent. The difference seems to matter.

Q: How do you balance running a company with a contemplative practice? A: Poorly, most days. Better on the days I sit before I open the laptop. I am still figuring it out.

Q: What would you suggest for someone who feels the same ache you describe? A: I would not suggest anything. I would only say that the ache itself is worth listening to. It usually knows more than I do.

Paweł Sokołowski is a CEO, Vipassana meditator, and student of Ayurveda and Thai massage. He writes from the intersection of technology, consciousness, and the ancient body.

← Back to the journal