What is Intelligence? Beyond the Loop
From intellect to intelligence through the algorithms of awareness in the age of AI
The Contained Intelligence
August in Kanpur was hot, humid, and restless. The rains had left the roads pitted with potholes, brimming with muddy water. Tempos rattled past, sending up sudden splashes that tested the reflexes of anyone on foot, while a mango vendor called out the season’s last bargains over the low rumble of traffic.
The year was 1993, and that month I began my 11th grade—and with it, the long march toward the IIT Joint Entrance Exam (JEE). The exam was still two years away, but it hovered over everything, an invisible deadline tightening the frame around my days. I’d been told it was the most competitive exam in the world—harder than anything I could imagine—and I didn’t know a single person who had made it through.
School, by contrast, felt easy and almost incidental. I was ahead of the syllabus, often finishing problem sets before the teacher introduced the material. Most days were lighthearted, spent playing football in black leather shoes and sharing patties from the canteen. In the background, there was always that exam.
Evenings were sacred. The narrow street in front of our house became our cricket pitch, the ball wrapped in tape to soften its bounce. Rules shifted depending on how many turned up: a one-handed catch after a single bounce was an out; a ball into a neighbor’s house, the same. I had already given up tennis—a sport I adored, but one I could no longer play after we moved houses and the court was too far to reach. Cricket still gave me glimpses of that feeling: no friction between mind and body, no inner voice judging. Just flow.
Eventually, the street would empty. The sound of play would fade, replaced by the hum of ceiling fans and the clink of dishes being washed. The shift was almost automatic: I would move from the open air to my study room on the second floor, as if the day itself had narrowed. My companions were a rickety chair with wheels and a stack of books that looked older than me, by authors I had never heard mentioned, alongside great scientists or mathematicians: Resnick & Halliday, Irodov, Hall & Knight, and Loney. These weren’t guides; they were opponents.
I’d begin around 10 p.m., sometimes earlier, often after everyone had fallen asleep. And once I started, I couldn’t stop. I wasn’t solving problems out of love, but for that exhilarating feeling of conquest.
Power cuts were routine, sometimes lasting ten hours or more. Most nights, the oil lamp cast a flickering circle on the page. I worked through conservation of momentum, projectile motion, trigonometry—sometimes sitting with a single problem for half an hour until—suddenly—clarity, as if I had taken something wild and forced it into a structure. Then it was on to the next one.
By 2 or 3 a.m., my left hand would be smeared with ink, my back aching, my mind still running, estimating my chances of getting through the exam. Stopping felt like betrayal. I told myself I was focused, disciplined, and preparing for greatness. Yet inside, it felt more like chasing some version of myself I hadn’t yet earned the right to be.
And while I couldn’t name it then, that was a form of suffering. Not the loud kind. I wasn’t in pain, and nothing outward was wrong. But something in me—something tender and alive—was being slowly replaced. I was trading curiosity for certainty in the future.
I thought I was building intelligence, but what I was really doing was containing it.
The Intelligence We Forgot
I got into IIT. The exam that had shaped my days, defined my nights, and contoured my very sense of self—it was over. For the first time in years, there was no next problem set waiting, no clock ticking toward the future I’d been chasing. The release was strange: part relief, part emptiness, as if I had stepped out of a tunnel into a space too wide to know where to look.
I moved to Mumbai, where the rains made Kanpur’s monsoons feel almost gentle. The downpours were relentless, drenching the city in minutes, yet the roads held firm—no potholes, no muddy craters to dodge. As I rushed half-awake to classes, cows wandered the streets, and with them came the smell of wet earth mixed with cow dung.
New friendships formed quickly. We stayed up late in dorm lounges, arguing about life and love, free will and physics. Conversations drifted between posturing and genuine inquiry—flowing from quantum mechanics to cricket scores to what it meant to live well. On some nights, we’d wander out to the campus lawns after midnight, lying on the damp grass and talking until the air grew cold. Even then, I often caught myself steering debates toward “right” answers, as if an ungraded conversation could still be won. One night, someone reached for a worn copy of the Bhagavad Gita.
We began reading passages aloud—not with the reverence of ritual, but with the curiosity of people meeting an old text for the first time. I remember stopping over a line, its clarity hitting harder than any proof I had learned: The intelligence of the resolute is one-pointed, but the minds of the irresolute branch endlessly.
I couldn’t say I understood it. But I felt it. Something in me recognized a truth I had been living without seeing: that intellect alone wasn’t enough.
I had trained my mind to solve, but now it was beginning to ask: What is worth solving? I had mastered the game, but some part of me kept asking: What is the game for? That question stayed with me for years, until I found language for it in a distinction many Eastern traditions hold sacred but modern discourse—especially in AI—rarely makes: the difference between intellect and intelligence. And with it, another often-misunderstood term: wisdom. Put simply: Intellect is the power to model the world. Intelligence is the capacity to step outside the model. Wisdom is knowing when to do which. In other words, by “wisdom” I mean long-horizon, context-sensitive regulation: knowing when to apply or suspend a model.
I later found echoes of this distinction in others’ work. Yuval Noah Harari warns that in an age of algorithms, self-observation may be our last refuge from being modeled. Pankaj Mishra shows how the Buddha’s diagnosis of suffering still speaks to modern restlessness. Both point to the same truth I had been circling: that discernment is not just a personal discipline but a cultural necessity—our ability to step outside the momentum of our loops before they carry us, unexamined, to their next conclusion.
The intellect is the part of the mind that reasons. It organizes, categorizes, and calculates. During my IIT preparation, it was the engine I relied on entirely—solving problems late into the night, ranking options, eliminating possibilities until only the right answer remained. It excels at abstraction and manipulation. Language is its primary instrument—shaping thought into symbols, slicing the fluid continuity of reality into neat, nameable parts. This is the faculty logic gates emulate—the one today’s AI tries to scale with vast compute and data.
An old line many teachers use calls the intellect “a beautiful servant but a dangerous master”—a sharp knife with no hand to guide it. For me, the intellect wasn’t just a servant—it was my whole identity. This knife had let me cut through the most complex problems and carried me from Kanpur’s sweltering monsoons to one of the most competitive institutions in the world.
Intelligence, however, is something deeper: the capacity to discern. Known in Sanskrit as buddhi (from budh, “to awaken”), it is the subtle, regulating faculty that sees what matters, and why.
During my IIT preparation, I don’t think buddhi stirred much at all. My nights were an unbroken current of problems and solutions, the intellect sprinting from one conquest to the next. But on the tennis court back in Kanpur, or in the narrow street where we played cricket, it flickered without my knowing the name for it.
In tennis, it was the moment I could feel—before the ball even crossed the net—that my opponent was off balance, and that going for the flashy winner would be the wrong choice. In cricket, it was letting a slower ball pass instead of swinging, because something in me sensed the trap. These weren’t calculations in the way I solved physics problems; they were pauses. Adjustments. The body knowing when not to act. And yet, even here, I sometimes caught myself treating the game as a puzzle to be solved, as if joy were just another problem with an optimal solution.
That is how buddhi works. Unlike the intellect, it does not rush to solve. It listens. It orients. It can hold back when the obvious move is to push forward. It does not merely produce outcomes; it governs them, aligning action with a deeper sense of rightness.
Eastern traditions call it the “awakening” faculty—not because it reveals mystical truths, but because it wakes the mind from the automatic momentum of intellect. It is what turns cleverness into clarity and knowledge into right action.
And then there is wisdom—not a separate faculty, but a state that arises when intelligence is practiced over time, with care. Wisdom is not just knowing what matters. It is knowing how to live in accordance with it. If intellect is the tool, and intelligence is the hand, then wisdom is the whole body—moving with coherence through a complex world.
The intellect finds the loophole. Intelligence refuses to exploit it. Wisdom sees the system that produced it.
The intellect wins the debate. Intelligence knows when the debate is hollow. Wisdom walks away.
The intellect sees the pattern. Intelligence feels when something is off. Wisdom knows when not to speak.
The intellect builds the system. Intelligence questions its purpose. Wisdom asks who it serves, and at what cost.
So, when in the Bhagavad Gita, Krishna tells Arjuna, “The intelligence of the resolute is one-pointed, but the minds of the irresolute branch endlessly,” that one-pointedness is not rigidity. It is clarity—the ability to perceive what truly matters amid the noise. The irresolute mind branches infinitely—mimicking what large language models do today: fluent, endless, ungrounded generation.
This is what makes AI so disorienting. It scales intellect without intelligence. It simulates fluency without discernment. It cannot pause, refuse, reflect, or reorient. It knows how, but not why.
Even the warnings amplify the myth. A prominent deep-learning pioneer has warned that advanced systems could outstrip us, pursue misaligned goals, and defy control—yet that framing still equates intelligence with scale and pattern recognition, crowning the machine on the very terms we should question.
By forgetting these distinctions, we commit a category error of civilizational scale. We conflate mimicry with the mind. We mistake fluency for wisdom. And we begin to redefine intelligence downward.
For me, these were nice theoretical ideas. Intellect, intelligence, wisdom—I could now name what I had begun to sense during those long nights in Mumbai. But I still lived almost entirely through my intellect. It had gotten me this far. It had earned me scholarships, accolades, and a place in the Algorithms, Combinatorics, and Optimization Ph.D. program at Georgia Tech.
So I doubled down.
At first, research felt different than preparing for IIT. It felt like freedom—open-ended, creative, almost playful. Yet even in that openness, I still measured progress by the proofs I could complete, the papers I could produce. Slowly, something began to fray. The joy of inquiry was replaced by the pressure to produce. Metrics, grants, rankings, reputations—everything became a race. And in that race, I saw things I couldn’t unsee. Things that chipped away at the quiet idealism I had carried with me. The pursuit of knowledge started to feel transactional. Strategy replaced sincerity. Attention splintered. And in ways I didn’t yet understand, so did I.
Eventually, I returned to India in late 2004. I joined IBM Research in Delhi, where the rules were surprisingly humane: spend half your time solving industry problems, and the rest on whatever research you cared about. After years of deadlines, the open space felt intoxicating. I wandered through bookstores and cafés in Khan Market, reading aimlessly, letting the hours stretch without consequence. But even here, I sometimes caught myself looking for “productive” takeaways from books that didn’t need to yield anything.
Incidentally, within a year, I reached a pinnacle in my mathematical life. Subhash Khot—my batchmate from IIT—and I co-authored a paper that disproved a long-standing conjecture. It was a breakthrough: celebrated and professionally validating. The kind of result that, years earlier, I might have imagined as a destination.
But no destination arrived. The rush lasted days, maybe a week. Then the old restlessness returned, sharper than before. I was already scanning for the next summit, the next conquest—as if still chasing an updated version of myself from Kanpur.
I had begun to suspect that reflection could turn in on itself, looping until it became a kind of self-made cage. I didn’t yet know how to step out. I only knew that the intellect—my most trusted guide—was beginning to feel like a compass that spun but never pointed anywhere.
Which is why, when something wholly different appeared before me, I noticed. By then, I could name the distinction between intellect, intelligence, and wisdom—but naming is not the same as knowing. I had yet to see what it looked like, embodied.
First Contact with Meditation
It was February 2006, a cool Delhi morning, when I wandered into the India International Centre. A close friend from Georgia Tech, Manu Sharma—visiting his family in Delhi—had mentioned a conference on mind and cognition, vague enough to pique my curiosity. He also mentioned the Dalai Lama would be speaking, but at the time, I knew him mostly as a political figure. I didn’t expect much.
I found myself in the auditorium as the inaugural session began. There was an invocation, a few words of welcome—and then he walked in: smaller than I had imagined, maroon robes draped lightly over his frame, an easy smile that seemed to include everyone.
It wasn’t his words that struck me most. It was his way of being. He listened with his whole body—eyes sometimes closed, head tilting in thought. When he spoke, he didn’t try to impress. He joked. He paused. He admitted what he didn’t know. In a hall filled with scientists and scholars, his laughter felt like the most intelligent thing in the room.
I don’t know why, but I wept—quietly, without warning. Not from sadness or joy, but from a deeper recognition: that something in me had gone quiet long ago, and here was someone who had carried such quiet through exile, loss, and dislocation—and yet seemed to live entirely from it.
Only weeks earlier, I had been riding the high of a major mathematical result—the kind of achievement I once imagined as a summit. Yet here was something that made it feel strangely small. For the first time, I felt that intelligence beyond intellect—wisdom as something lived—was not just philosophy. It could be experienced. And if it could be experienced, perhaps it could be cultivated.
On my way home, I stopped at a bookstore in Khan Market. I didn’t know what I was looking for, only that I wanted to try meditation—not just read about it in passing. I had heard of yogis who could focus for hours, control their breath, even “see God.” But even if I saw God, then what? What would that change in the restless undercurrent of thought that never seemed to quiet?
I wandered the shelves until a slim paperback caught my eye: Stages of Meditation: Training the Mind for Wisdom by Kamalashila, translated by the Dalai Lama. It mapped the mind’s progression in practice: from scattered to focused, from focus to compassion, to equanimity, and finally to a direct understanding of suffering and wisdom. The instructions were plain: begin with the breath; return to it when distracted; hold attention gently; let the mind settle.
I tried. I sat cross-legged on the thin carpet of my Delhi apartment, the book open beside me, eyes half-closed, attention at my nose. Within seconds, my mind darted away—replaying conversations, working through proofs, thinking about dinner. I’d catch it, pull it back, lose it again. Over and over.
The book said this was normal. Meditation wasn’t about silencing thoughts but about noticing their wanderings and returning without judgment. I understood this in words, but my habits were still intellectual: measuring myself, keeping score. It felt like a contest I was losing to my mind.
Still, something had shifted. I had brushed, however briefly, against another way of knowing. But the seed did not sprout. I went back to living as I always had—seeing the world as a stage for the self: its security, its validation, its expansion. In that separation, compassion thinned. People became means to an end. The familiar poisons—anger, pride, desire, resentment—found space to grow. And so suffering continued, not as a crisis, but as the quiet background hum of a life lived through the intellect alone.
Life moved on. Work called me elsewhere. My career, for the moment, reclaimed the center of the stage—though the memory of that morning in Delhi stayed, quiet but persistent, like an unanswered question.
From Formula to Algorithm
In 2009, I moved to Microsoft Research (MSR) in Bangalore. If IBM had given me half my time for open-ended exploration, MSR gave me even more. IBM, by then, was starting to sag under the buzzword of “Services,” and the change in atmosphere made me restless. At MSR, I felt lighter, freer—surrounded by people who lived in the deep end of their fields.
Instead of cafés in Khan Market, I began exploring the rain-soaked campus of the Indian Institute of Science. One of my old dorm-mates from IIT Bombay, Narendra Dixit, was now a professor there—one of the most brilliant and intellectually versatile people I knew. We started meeting for thali lunches at Nesara, a bustling canteen shaded by rain trees.
One afternoon, after we’d exhausted the usual catching-up, I asked what he was working on. Narendra described his research on the evolution of viruses—how they adapt under immune pressure, the competing dynamics of infection and treatment. I was hooked before the food was cold.
The math behind his models felt familiar; the biology did not. We decided to collaborate on a problem that had stumped his field for years: determining the error threshold for HIV—the mutation rate beyond which the virus can no longer sustain itself.
Here we hit a wall. In his world—drawing from physics—the instinct was to hunt for a closed-form formula: an equation you could plug parameters into for an exact answer. This problem resisted that entirely. My instinct was different: forget the formula, find an algorithm. Build a step-by-step procedure, prove it would converge, and let computation do what algebra could not.
We did just that—developing an algorithm that could provably estimate the threshold where traditional theory had come up empty.
That project left me with a conviction: much of the physical world looks intractable only because we insist on seeing it through the lens of formulas. Shift the lens to computation, and whole classes of problems become tractable—not always elegant in the mathematician’s sense, but solvable, understandable, provable.
I didn’t know it then, but I would need that lesson years later—in a completely different domain.
During the COVID years, I saw the irony. My whole professional life had been about algorithms—designing them, proving them, teaching them. I had convinced others to see the value of an algorithmic approach when formulas failed. Yet in meditation, I abandoned that lens entirely. I wanted a neat, closed-form recipe for calm or insight—a single expression that could be understood without being lived.
It took me far too long to see that meditation was also an algorithm. Not the shallow kind you find in a list of steps, but the deep, state-transforming kind I had worked with all my life. I had first glimpsed it, without naming it, on that quiet February morning in Delhi.
Meditation as Algorithm
I had spent most of my life in the mathematics of algorithms—always in the abstract, never in breath. From the sweltering monsoons of Kanpur to lecture halls at IIT, then to Georgia Tech and Microsoft Research, I was trained to break problems into parts, to define states and transitions, to know exactly what would happen when an input passed through a function. That training felt rarefied at the time, but the loops it reinforced—anticipation, problem-solving, prediction—were not. They run in every mind, whether in a lab, on a factory floor, or waiting at a bus stop.
Years later, I realized I’d been hunting for a formula in a domain where only an algorithm would do—one I’d been holding all along.
The boy in Kanpur, hunched over Irodov by oil lamp, tracing step after step through conservation of momentum—he was already running it. Not the solution in a single stroke, but the slow, iterative process that changes the state of the system until something new emerges.
In mathematics, a formula is a direct expression. You plug in the inputs, do a finite amount of calculation, and out comes the answer. If you want the acceleration of an object under a given force, Newton’s law F = ma does it in a single line. If you want the sum of all integers from 1 to n, we have: 1 + 2 + … + n = n(n+1)/2.
Now imagine a pile of cards to order from Ace to King. With a few, you can jump straight to the answer. As the pile grows, the question shifts from What’s the order? to How do we get there? That’s where algorithms come in: step-by-step procedures that transform the system until the goal is reached. In selection sort, you repeatedly move the smallest remaining card to the front of a growing sorted pile; two things never change—the front stays ordered, and the pile keeps growing until the deck is complete. The answer doesn’t appear at once; it emerges through execution.
Meditation, I came to see, is exactly this kind of process—not an algorithm in the shallow sense of “a set of steps,” but in the deeper computational sense: an iterative procedure that transforms the system’s own rules. It operates on the substrate of the mind itself, altering how thoughts and feelings arise, persist, and fade.
Here is the crucial difference: most algorithms are designed to converge to a specific goal. Meditation is a deconditioning algorithm—its purpose is not to lock the mind into a new attractor, but to loosen the predictable pathways it habitually follows. Where intellect optimizes for prediction, control, and self-expansion, meditation optimizes for direct awareness. Its cost function is not Am I winning? or Am I safe? but Am I aware?
This is where AI, for all its computational brilliance, reaches a hard limit. A machine can execute an algorithm, but it cannot alter the update rules of its awareness—because it has none. It can simulate the outputs of intelligence, but not the inward shift of perception that this form of intelligence requires.
For a long time, I expected more from it—some decisive insight, a clean solution I could write down. The turning point came when I stopped looking for a definitive description and committed to the practice itself. That’s when I began to see a way out of the loop I had traced earlier: awareness folding into identity, identity into craving, craving into control—until suffering becomes self-perpetuating.
In computational terms, that loop is like a feedback system whose state space has collapsed into a tight orbit: a local minimum. Meditation changes the update rule and retrains for direct awareness.
The techniques vary—breath, body scans, visualizations, open awareness—but structurally they share the same move: detect when the mind has wandered, return it gently to an anchor, repeat. This is not about suppressing thoughts or chasing bliss; it is about re-training the system’s dynamics so that awareness can remain stable regardless of the content it encounters.
Over time, this changes the topology of the mind’s state space. The familiar attractors—anger, pride, desire, fear—still arise, but they lose some of their gravitational pull. Compassion becomes less something you will into being, and more something that arises naturally when the boundaries between “self” and “other” soften.
From the outside, this may look passive. Inside, it is one of the most active forms of learning a mind can undertake. You are not merely acquiring information; you are altering the machinery by which information is processed.
Some traditions describe this process differently. J. Krishnamurti, for example, spoke of a shift that is not cultivated over time but happens in a single act of perception—attention without a center, free of method, free of progression. In his view, meditation is not algorithmic at all, but a movement outside time that begins when conditioning ends. My framing is closer to the computational: practice as an iterative retraining of the mind’s dynamics. Yet both point to the same horizon—the possibility of meeting each moment without the residue of the last.
The promise—echoed across traditions from Kamalashila to the Upanishads—is that all beings, once they strip away the layers of conditioning, are capable of the same direct experience. Not similar. Not approximate. The same. Because what is realized is not personal—it is universal.
If intellect is about building better models of the world, meditation is about loosening the modeler’s grip. In that loosening, the loop of suffering can unwind.
None of this is new. The Buddha described the process more than two millennia ago. What he called “training the mind” we might now describe, without changing its essence, as an algorithm: an iterative procedure for shifting the dynamics of thought and perception—a direct, non-discursive knowing many have called “perennial.”
The novelty lies not in the process itself, but in the translation, placing it in the same conceptual frame we use for machine learning, optimization, and control. For those fluent in these domains, it reveals meditation not as mysticism or self-help, but as a rigorous method for altering the operating conditions of the mind.
Much of this transformation unfolds beneath words. The deepest shifts in perception and response don’t announce themselves in sentences and can’t be cleanly encoded in equations. Yet the literacy-centered schooling of the last two centuries in the West—and now worldwide—has elevated linguistic and mathematical reasoning at the expense of the sensory, emotional, and embodied. That narrowing makes it easy to misidentify large language model (LLM) output as “thought”: first, we thin our notion of thinking to what can be written or formalized, then we marvel when machines meet that stripped-down criterion.
And just as in computer science, an algorithm is only half the story. We also ask: How do we know it works? In mathematics, we prove correctness. In practice, we test. In meditation, the proof cannot be written down, but it can be run. This isn’t an appeal to authority; it’s an invitation to experiment.
If you’re tempted to ask whether better instruments could arbitrate this from the outside, consider a simpler case. Imagine I hand you a laptop and every instrument you could want: oscilloscopes, logic analyzers, nanosecond-resolution traces. From those signals alone, could you tell whether it’s running quicksort or mergesort—drafting a love story or an obituary? Probably not. The same low-level traces can implement many high-level programs; without code or the right interpretive language, meaning stays underdetermined.
Our neuroscience tools—fMRI, MEG, multi-unit spikes—are those meters for the brain. They give exquisite correlates—voxels, couplings, spectra—not semantics. Presence, discernment, and compassion are organizational properties visible only from the first-person stance. Neuroscience can and should constrain and corroborate that observation; it can’t replace it.
This points to the only lever we directly control: practice from within.
Meditation and Predictability
I had spent much of my life designing algorithms that narrowed possibility—procedures to make the next step as foreseeable as possible. Meditation flips that logic. Instead of tightening the loop, it loosens it. Instead of making the next move inevitable, it makes it open.
If meditation is a counter-algorithm—an update rule that loosens the grip of the loop—then one of its side effects is unpredictability.
For me, this is not abstract. Most days, I sit for 5, 10, maybe 15 minutes. Sometimes it’s a single deep breath held in stillness. Sometimes it’s broadening my visual field with eyes closed. Sometimes it’s anulom vilom—alternate-nostril breathing, my grandparents practiced.
However small, these techniques make me less predictable. Not in a mystical way, but in the very real sense that my next reaction isn’t dictated by habit. A tense email arrives, a driver cuts me off, a memory surfaces—and instead of the usual cascade of triggers, there’s a pause. The update rule has been interrupted.
Algorithms, whether human or machine, thrive on patterns. The fewer degrees of freedom you exercise, the easier you are to model. Meditation reintroduces those degrees of freedom. It creates gaps in the loop—moments in which the “next move” is no longer inevitable.
Recently, an old schoolmate from Kanpur—now a seasoned audit director in New York—confided that he was afraid of AI. “It’s eerie,” he said. “I typed `what am I thinking?’—and it gave a pretty good guess.”
For him, the fear wasn’t the inanimate AI, but the handful of people who control it—people who, with access to its inner workings and data, could read the patterns beneath a life. At root, it’s a governance problem—concentrated modeling power with few checks. The danger isn’t the blade; it’s the absence of a governor—the discernment to decide how, and whether, it should be used.
What’s new isn’t surveillance—someone could already read your emails or messages—but the leap from seeing what you’ve said to modeling why you say it: tracing subconscious imprints that drive your choices, and predicting “spontaneous” decisions before you make them. It’s the difference between overhearing a conversation and reverse-engineering the architecture of a mind.
A tool like this is not inherently evil, but in the wrong hands, the power to model and manipulate human behavior at that depth becomes dangerous.
If the threat is in being fully predictable, the countermeasure is not opacity but fluidity—reducing the grip of those patterns on our behavior. That is precisely what meditation trains: loosening the loops so awareness can meet each moment without being locked into the past. A mind less governed by old imprints is not only freer; it’s harder to model.
Which is why I keep coming back to my seat, to my breath, to that widening field. Meditation isn’t just a private refuge. It’s a way of stepping off the grid—not digitally, but cognitively. When you can notice the loop before it runs, when you can see your own thoughts without automatically acting on them, you become harder to model. Harder to predict. Harder to hack.
That unpredictability comes from something deeper. Awareness is also a way out of prediction. I remember sitting in a small room in my house near New Haven, early winter light spilling in, realizing that my mind was doing exactly what my models did: scanning past data, projecting forward, narrowing possibility to what had already been seen. Prediction lives in the past—it watches what you did, what others did, and draws its lines forward.
Awareness breaks that loop. To be aware is not to predict the next thought, but to see it arise and dissolve. To not act automatically. To not be nudged by history or modeled by data. In awareness, you are not a function of your inputs. You are not optimized. You are free.
This is what no model can replicate—not because it’s too complex, but because it is not a computation. It is a presence, a witnessing, a stillness that cannot be inferred from patterns—only returned to.
And here the body matters. A scientific study of intelligence will always be incomplete if it excludes the human body as an instrument. For too long, we’ve treated intelligence as something that happens in the brain, or worse, in abstract symbolic systems. But cognition is not suspended in the air. It is shaped by breath, posture, gut, gait, skin, hormones, and heartbeat. Your capacity to attend, remember, intuit, or reflect is not just a function of neural activity—it’s embodied regulation.
Traditions like yoga, tantra, and vipassana understood this long before neuroscience caught up. They don’t ask you to conceptualize intelligence; they ask you to sit, breathe, and observe. To run the body as the instrument—not as a metaphor, but as a substrate. And once you run that algorithm, you discover a form of intelligence no model can simulate: intelligence as presence.
Until we include the body—not just as a vessel, but as a participant in knowing—our models of intelligence will remain disembodied, and therefore, partial.
Some traditions go further. They don’t just acknowledge the body—they give it specific, reproducible algorithms for shifting the mind. One of the simplest begins with the breath.
“Attend to the turning point of the breath.” Not breath control. Not relaxation. Just notice the exact moment the breath reverses direction—a brief, natural pause, when neither inhaling nor exhaling.
The Loop Breaks — Most of the day, we live in loops: thought predicts feeling, feeling triggers memory, memory shapes the next thought. Even breath becomes automatic—linked to emotion, tension, fear. But at the turning point, there’s no direction, no goal. Just pause. The loop doesn’t feed itself, and suffering—which rides on anticipation—momentarily stops.
Craving Can’t Find Purchase — Desire and aversion rely on movement: toward or away, inhale or exhale, grasp or resist. In the turning point, there is no “toward,” no “away.” Just being. The mind has nothing to chase. No thought to hold. No self to protect. Craving doesn’t apply.
The Self Isn’t There — Try to locate yourself in that turning point. You won’t find a storyline, a goal, or a possession—only breath, only awareness. This is not annihilation. It’s unhooking: from identity, from narrative, from the constant simulation of control. And in that unhooking, something incredibly quiet begins to show—a version of awareness untouched by suffering. Not a better story, but the absence of needing one.
In practice, this doesn’t just change what you notice; it changes how predictable you are.
The Full Circle
What happens in that breath is not unique to breath. It’s the same shift I’ve been circling in this essay—the moment intelligence steps outside its loops. The old insight still holds: the mind alone can analyze the mind. But for years, I misunderstood what that meant. I thought meditation was about reaching some special state, a goal I could measure. And because I could not reach it, I kept thinking I was failing.
It was algorithmic thinking that finally loosened that grip. In computation, an algorithm may have a goal—sorting a list, finding a path—but you don’t get there by willing the final state into existence. You get there by running the steps faithfully, letting each comparison and swap do its quiet work. Skip the steps, and the goal never materializes. Meditation, I began to see, worked the same way.
We began with the myth of superintelligence, a story that cast intelligence as prediction, control, and speed. It was a seductive frame, but I came to see it for what it was: a narrow definition of intellect, not of intelligence. From there, the scope widened—first to photons and slime molds, then to the architecture of the mind, and finally to the wisdom traditions that ask what lies beyond the mind’s loops altogether.
On a trip to Kanpur, I sat in a small wooden boat on the Ganga. The city’s intersections were still as I remembered—chaos resolving without plan—but here, the river moved without obstruction. I tipped a clay pot over the side; the ashes drifted, broke apart, and vanished into the current.
It was the opposite of the loops I had spent my life studying—no return, no recurrence, no optimization. Only release. The air stilled, like the turning point of the breath: no direction, no goal, just the pause before the next movement. Meditation is one path into it; there are others. But whatever the path, I’ve come to see the mark of intelligence not as prediction or control, but as the capacity to meet the moment unbound by the patterns that brought you here—a place no machine will ever reach, no matter how “superintellectual” it becomes. In that unbinding, intelligence moves unconstrained.
References
Radhakrishnan, S. (1948). The Bhagavadgita. Harper & Brothers.
The Dalai Lama and Kamalashila (2003). Stages of Meditation. Snow Lion Publications.
Knuth, D. E. (1997). The Art of Computer Programming. Addison-Wesley.
Tripathi, B., Balagam, R., Vishnoi, N. K., & Dixit, N. M. (2012). Stochastic Simulations Suggest that HIV-1 Survives Close to Its Error Threshold. PLoS Computational Biology, 8(9): e1002684.
Mishra, P. (2004). An End to Suffering: The Buddha in the World. Farrar, Straus and Giroux.
Harari, Y. N. (2018). When Tech Knows You Better Than You Know Yourself. Wired (Interview).
Krishnamurti, J. (1996). Total freedom: The essential Krishnamurti. Harper.
This was a beautiful read.