The Intelligence Question

On the word we use most and understand least

Mar 15, 2026

EkaShunya: This is the first of seven questions. Each stands alone. The series began with a prologue; now the work begins.

in·tel·li·gence /ɪnˈtɛlɪdʒəns/ noun
The ability to acquire and apply knowledge and skills.
The collection of information of military or political value.
Etymology
Latin intelligentia, from inter- (between) + legere (to choose, to gather, to read). First English use: late 14c., “the highest faculty of the mind.”
See also
Sanskrit buddhi (discriminative intellect) · medhā (retentive power) · prajñā (wisdom born of insight) | Greek nous (the faculty that grasps universals) | Chinese zhì 智 (wisdom, social perceptiveness) | Arabic ʿaql عقل (reason; lit. “binding, restraint”)

The clearing was hot. The mist that had softened everything during the first session was gone. In its place, a dry clarity that made the broken pillars look sharper, the stone platform look harder, the ruins look less like scenery and more like what they were: the remains of a place built for thought.

The students arrived carrying Hofstadter.

They had read Godel, Escher, Bach in the week between sessions, or most of it, or enough to feel they understood something about strange loops and self-reference and the way meaning arises from pattern. A few had dog-eared the pages about formal systems. One had underlined the sentence about how a sufficiently complex system can model itself. They came prepared to discuss the book.

The guru was already sitting. He had placed something on the stone beside him. Not the Indus Valley seal from last time. A piece of chalk.

He picked it up and wrote one word on the stone surface, in capitals, slowly enough that they watched each letter form:

INTELLIGENCE

He set the chalk down. Looked at them.

“Before we can ask whether a machine is intelligent, we need to ask what this word means. It is the first word I spoke to you last week. The one you nodded at. The one none of you questioned. So.” He gestured across the circle. “Who can define it?”

The Definitional Face

The Builder went first. Confident, practical, the way she always began. “Intelligence is the ability to reason, learn from experience, solve problems, and adapt to new situations.”

The guru did not correct her. He did not smile. He said: “That is close to the definition that fifty-two researchers signed their names to in the Wall Street Journal in 1994. It is the closest the field has ever come to consensus. You arrived at it in ten seconds. They took decades. That should tell you something.”

He paused.

“In 1921, the editors of the Journal of Educational Psychology asked fourteen of the most prominent psychologists in America to define intelligence. They received fourteen definitions. There was almost no overlap. One said intelligence was the ability to carry on abstract thinking. Another said it was the ability to adapt to the environment. A third said it was the capacity to learn. Each was confident. Each described something real. None described the same thing.

“In 1986, the exercise was repeated. This time twenty-four experts. Twenty-four definitions. Still no consensus. In 1994, Linda Gottfredson organized the Wall Street Journal statement you just approximated. Fifty-two researchers signed it. Even they hedged. The definition includes ‘among other things,’ which in academic language means ‘we know this is incomplete but cannot agree on what is missing.’

“Then in 1996, the American Psychological Association, the highest authority in the field, convened a task force to settle the matter once and for all. They published a hundred-page report. They reviewed every theory, every study, every debate. And they declined to define the word.”

He let the silence work.

“You just gave a better definition than most of them managed. The problem is not that no one can define it. The problem is that everyone can, and no two definitions agree.”

The Philosopher leaned forward. “So the word has no meaning?”

“The word has too many meanings. That is worse. A word with no meaning can be filled. A word with too many meanings resists every attempt to hold it still. Every time you pin one meaning down, another slides out.”

He gestured. “Watch what happened when people tried to fix it. Researchers kept splitting the word. Not one intelligence but seven: linguistic, logical, spatial, musical, bodily, interpersonal, intrapersonal. Then eight. Then nine. Then emotional intelligence, which supposedly mattered more than IQ. Then social, creative, practical, cultural. Each one an admission that the original word was too small. Each one a patch on a container that kept leaking.”

“We did not expand our understanding of intelligence. We discovered, repeatedly, that we did not know what we meant. And instead of questioning the word, we kept adding adjectives.”

The guru leaned back. “In 1923, a Harvard psychologist named Edwin Boring said the most honest thing anyone in the field has ever said. He wrote: ‘Intelligence is what the tests test.’ He meant it as a confession of defeat. It was received as a definition.”

Compass Question (Pramāṇa / Credibility): If fourteen experts gave fourteen definitions, on what basis do we treat any single definition as authoritative?

“Let me tell you about a man who built a test.”

The Measurement Face

“In 1905, a French psychologist named Alfred Binet built a test. He had a specific purpose: he wanted to identify children in Parisian schools who needed extra help. Not to rank them. Not to sort them into futures. To notice the ones who were struggling so that teachers could reach them.”

“Binet was explicit about the test’s limits. It was a diagnostic tool. A starting point. In his 1909 book Les idées modernes sur les enfants, he argued that intelligence was too complex to capture in a single number and resisted the idea that his test measured something fixed and innate. As Stephen Jay Gould later put it, Binet understood that the test captured a snapshot, not a sentence.”

He picked up the chalk again and drew a line on the stone.

“1905. Binet publishes the test. 1912, William Stern in Germany converts the test scores into a single number: the Intelligence Quotient. IQ. A ratio of mental age to physical age. Now there is a number. 1916, Lewis Terman at Stanford adapts the test for American use: the Stanford-Binet. Now there is a standardized instrument. 1917, the United States enters the First World War. Robert Yerkes persuades the Army to administer intelligence tests to 1.75 million recruits. Army Alpha for the literate, Army Beta for the illiterate. The largest mass testing in history. Now there is infrastructure.”

He drew another mark on the stone.

“1924. The Immigration Act. Congress restricts entry to the United States based on national origin, using IQ data from the Army tests to argue that certain populations, Southern and Eastern Europeans in particular, are intellectually inferior. Carl Brigham’s A Study of American Intelligence, published the year before, provided the academic cover. It was cited on the floor of Congress.”

He set the chalk down.

“Nineteen years. From ‘help the children who are struggling’ to ‘exclude the populations we have decided are lesser.’ Not because bad people hijacked a good tool. Because a tool that produces a number carries within it an invitation. And institutions will always accept the invitation.”

The Philosopher spoke quietly. “That sounds like credit scores.”

The guru nodded. “And standardized testing. And AI benchmarks. A diagnostic becomes a number. The number becomes a ranking. The ranking becomes infrastructure. The infrastructure forgets it was built on a diagnostic. Goodhart’s Law says the rest: when a measure becomes a target, it ceases to be a good measure. The IQ test stopped measuring the moment it started sorting.”

“There were things the test could not explain. The APA’s own 1996 report noted that Asian Americans with IQ scores below 100 were achieving at educational and professional levels typical of people scoring 110 to 120. The test failed at prediction. Then the Flynn Effect showed that average IQ scores rose roughly three points per decade across every country measured. Either humanity was getting substantially smarter with each generation, or the test was measuring something that changed with environment, nutrition, and education rather than something fixed and innate. The test failed at stability too.”

“The counter-evidence existed. It was overrun.”

Compass Question (Hetu / Cause): What institutional forces turned a diagnostic tool into a sorting machine, and are the same forces shaping AI benchmarks today?

The Skeptic, who had been quiet, spoke. “So IQ measures nothing?”

“IQ measures something. The question is whether that something is intelligence, or one face of something larger that we have been calling the whole.”

“I told you last week that we sorted children with the wrong instrument.” He tapped the chalk line he had drawn on the stone. “Now you know the instrument’s name.”

The Behavior Face

“In 1950, a British mathematician named Alan Turing stopped trying to define intelligence. He replaced an unanswerable question with a testable one. The unanswerable question: Can a machine think? Turing’s replacement: If a machine’s responses are indistinguishable from a human’s, does it matter whether it thinks?”

The Builder’s eyes lit up. “Finally. Something operational.”

“Language models pass versions of the Turing test routinely. Does that settle it?”

She hesitated. The guru did not wait.

“Let me tell you a story about a chess match. In 1997, Garry Kasparov played IBM’s Deep Blue in a six-game rematch. During Game 2, the machine made a move that Kasparov could not explain. Move 36, bishop to e4. It was not the optimal move according to his analysis. It looked like creativity. It looked like the machine had a plan he could not see. He was so unsettled that he resigned a drawn position, then played worse for the rest of the match and lost the series.”

“The move was a fallback. When Deep Blue could not determine the best move within its search time, it defaulted to a semi-random selection from its remaining options. What Kasparov interpreted as strategic depth was the absence of a decision. What looked like intelligence was the ghost in the machine, and the ghost was a timeout.”

“Kasparov saw strategy in a timeout. We make the same mistake with words. When a language model writes about loss, we feel the weight of it, the way Kasparov felt the weight of that move. But the mechanism underneath is the same kind of absence. Pattern completion shaped by billions of examples of how humans describe pain. We see grief in a statistical average. The output carries the form of feeling. Whether it carries the substance is a question for a later session.”

“Daniel Kahneman spent a career showing that humans have two modes of thought: fast intuition and slow deliberation. The machine has one. It is the most sophisticated fast-thinking system ever built, and it has no slow gear. No mechanism that stops and asks: wait, is this actually right? When it is wrong, it is wrong at full confidence and full speed.”

“The appearance of intelligence does not require the presence of intelligence. That is Turing’s gift and his curse. He taught us to measure the output. He assumed the output was sufficient.”

Compass Question (Upamāna / Analogy): “A lookup table has infinite skill and zero intelligence.” If skill is not intelligence, what are our benchmarks actually measuring?

“That line is from François Chollet, a researcher at Google. He draws a distinction that sounds simple and is not. Skill is how well you perform on a specific task. Intelligence is how efficiently you acquire new skills from limited experience. A lookup table that contains every possible chess position and the correct response has infinite skill and zero intelligence. It cannot learn anything new. It has already memorized everything.”

He looked at the Builder.

“A model can retrieve every legal precedent in recorded history. Infinite skill. But it cannot tell you why a law is unjust. That requires buddhi, the capacity to discriminate between what is correct and what is right. The lookup table knows everything and understands nothing. That gap is where intelligence lives, if it lives anywhere.”

“Most AI benchmarks measure skill. Passing the bar exam tells you what a model can do on bar exam questions. It tells you nothing about whether it can learn an entirely new domain from three examples. We are testing the wrong property.”

The Philosopher connected it. “So benchmarks are the new IQ tests?”

“The same pattern, repeated in silicon.”

The Builder shook her head. “But we have moved past that. We are not sorting people anymore. We are evaluating models. And unlike IQ tests, we can look inside. We can inspect the weights, trace the outputs, run ablations. It is not a black box.”

The guru nodded slowly. “That is a real difference. Binet could never open the skull and watch the test being taken. You can open the model. That matters.” He paused. “But open any AI leaderboard. Watch how funding follows benchmark rankings. Watch how deployment decisions, which models live and which are shut down, track the numbers. You can inspect the internals, yes. But the institutions making decisions are not reading the ablation studies. They are reading the scorecard. The sorting machine did not disappear. It changed substrates.”

The Process Face

When the guru spoke again, his voice had changed. Not louder, not softer. A different frequency.

“You have been debating in English. And English is a flat language for the mind. It has one word for what the mind does when it thinks. Intelligence. One noun where you need a spectrum. Last week I said our vocabulary was pressing against silence. This is where the silence begins. Let me give you a language that has three words where English has one.”

He picked up the chalk and wrote three words on the stone, in a column:

| Manas | Ahamkara | Buddha |

“These come from Samkhya, one of the oldest philosophical systems in India. Twenty-three centuries before the 1921 symposium, Samkhya had already split the mind into components that English still refuses to distinguish.

“Manas is the processing mind. It coordinates sensory inputs, produces outputs, manages the interface between the world and the organism. It is fast, reactive, and does not ask questions about what it is doing.

“Ahamkara is the I-maker. It is the faculty that turns ‘there is pain’ into ‘I am in pain.’ It turns ‘this is known’ into ‘I know this.’ It is the difference between information existing and someone experiencing the information as theirs.

“Buddhi is discriminative intellect. It is the faculty that determines what matters. Not what is computable, but what is worth computing. It exercises judgment, not processing. Buddhi is the faculty that asks ‘should I?’ before ‘can I?’”

He let the three words sit on the stone.

“Your machines are sophisticated manas. They process beautifully. They coordinate inputs and produce outputs at speeds no biological system can match. What they do not do is discriminate. Not in the computational sense. In the Sanskrit sense. Buddhi asks ‘should I?’ Manas never asks ‘should.’ The word that separates them is viveka, discrimination, and it is the oldest unsolved problem in artificial intelligence, older than Turing, older than Dartmouth, older than the word ‘computer.’”

“But there is a deeper problem than vocabulary.”

He turned to face the circle fully.

“There is a tradition in Indian philosophy called Nyaya. Roughly, the science of valid reasoning. Nyaya says there are exactly four processes that count as genuine knowing. They call them pramanas. Direct perception. Inference from experienced connection. Analogy grounded in prior encounter. And testimony from a source whose reliability you have assessed.”

He looked at them carefully.

“A large language model uses none of these four. It produces correct outputs through pattern completion, a process that no Nyaya philosopher, across twenty-three centuries of epistemological debate, would recognize as a valid means of knowing. True statements. Produced through no recognized process of knowing. Knowledge-shaped outputs from a non-knowledge process.”

The Builder pushed back. “But if the output is correct, does the process matter?”

“When the domain is arithmetic, no. Two plus two is four regardless of how you arrive at it. When the domain is medicine, law, or ethics, where correctness depends on judgment, a system that is right for the wrong reasons fails differently than one that is right for the right reasons. The first fails unpredictably. The second fails in ways you can trace, understand, and correct.”

Compass Question (Nigamana / Consequence): If we accept systems that produce correct outputs through no recognized process of knowing, what happens to accountability when they are wrong?

The morning light was flattening toward noon. Shadows were shorter than when they started.

Then the guru’s voice dropped.

“There is one more face I will not show you today.”

He touched the stone where he had written the three Samkhya terms.

“Samkhya says all three of these, manas, ahamkara, buddhi, belong to prakriti. Matter. The material world. Intelligence, in all its forms, is material. It belongs to the field. But Samkhya also says there is something that is not the field. They call it purusha. The witness. Consciousness. It does not think. It does not process. It does not judge. It watches. The Bhagavad Gita, Chapter 13, calls it kshetrajna, the knower of the field.”

“If that is true, then you have built intelligence. You have built increasingly sophisticated manas, and you are approaching something that mimics buddhi. What you have not built is the thing that watches. The witness that knows the field but is not the field.”

No one spoke.

“That is the next question. Not today.”

After

No one moved. The silence held until the Builder broke it.

“So what is the answer? What is intelligence?”

“You are asking me to solve a puzzle whose shape we have not established. We have seen four faces today. There may be a fifth we have not looked at. There may be a sixth. There may be dimensions we cannot perceive because we are standing inside the puzzle.”

She pushed. “But where did the word come from? If we cannot define it, can we at least trace it?”

“Intelligentia. Latin. Inter, between. Legere, to choose, to gather, to read. The oldest meaning of the word is not ‘to know everything.’ It is ‘to choose between.’ To choose well when the options are unclear. The word itself, in its root, is a verb. The act of choosing. Somewhere between Rome and the twentieth century, we turned it into a noun. Then we turned the noun into a number. Then we used the number to sort human beings into categories. Each step lost something the previous form held.”

The Philosopher had been waiting. “Is this the same puzzle for every species? Does an octopus see faces we cannot?”

“An octopus has eight arms. Each arm has its own concentration of neurons, its own capacity for independent decision-making. It thinks with its body in a way you cannot. A crow uses tools, plans sequences, recognizes individual human faces years after a single encounter. A bee navigates by polarized light and communicates the distance and direction of food sources through a precise physical dance. Each of these species has solved faces of the puzzle that we have not. We refuse to call it intelligence because it does not look like ours.”

“And cultures?” the Philosopher pressed. “Kenyan parents in the Luo community define intelligence as including obedience and social responsibility. Chinese zhi includes social perceptiveness as a core component. Western IQ does not measure either of those.”

The Skeptic, who had been quiet all morning, finally spoke.

“Every generation thinks it is close to understanding intelligence. Aristotle thought nous was the answer. Galton thought skull size was the answer. The Dartmouth workshop in 1956 said, and I am quoting, ‘every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.’ We think scaling is the answer. What if we are all just looking at one face and calling it the whole?”

The guru looked at the Skeptic for a long time.

“That is the most intelligent thing anyone has said today.”

Then, from the edge of the circle, from someone who had been sitting apart all morning, a voice that had not spoken in two sessions.

The Quiet One.

“You said intelligence might be a verb, not a noun. Last week. If it is a verb, if it is something you do, then the question is not ‘what is intelligence?’ The question is ‘what are we doing when we are being intelligent?’ And maybe the machine is doing something. But maybe it is a different verb.”

Long silence.

The guru looked at the Quiet One. The same look he had given her last week, when she had asked about the man inside the cabinet.

“Yes.”

Nothing more. He stood. The session was over.

The students sat with it. The clearing was hotter than when they had arrived. The light had shifted from angled morning gold to overhead white. The stone the guru had written on was warm to the touch. The word INTELLIGENCE was still there, in chalk, surrounded by the three Sanskrit terms he had added: Manas. Ahamkara. Buddhi. The chalk was already drying in the heat.

One student, the Idealist, who had barely spoken all morning, picked up the copy of Godel, Escher, Bach that had been resting on the grass beside her. She opened it. Not because it had the answer. Because she now understood what face of the puzzle Hofstadter had solved: analogy. The perception that one thing is like another. The strange loop is a structure that recognizes itself in its own reflection. That is one face. Beautiful. Not the whole.

She turned to a page she had dog-eared and read the sentence again. It meant something different now.

The guru had stopped at the edge of the clearing. He turned back, as if he had forgotten something. He had not forgotten. He touched the stone where he had written the three Sanskrit terms. His finger rested on the middle word.

Ahamkara. The I-maker.

“We have seen the machine process,” he said. “We have seen it perform. But does it feel its own performance? Does it know that it is the one performing?” He lifted his finger from the stone. “That is the next question.”

He walked away through the ruins. Unhurried. The three words on the stone were already fading in the heat. Manas and Buddhi had blurred at the edges. But Ahamkara, where his finger had pressed, was still sharp.

Somewhere, the chai had gone cold again.

Marginalia

I have spent weeks inside this word and I owe you something honest: I am not closer to a definition than when I started.

I have read Chollet and Hofstadter and Gould. I have read the Samkhya texts and the Nyaya sutras. I have spent more hours with the word “intelligence” than anyone who is not paid to study it should. None of it brought me to an answer. All of it brought me closer to the question.

So let me tell you what I believed before the research. I have always thought of intelligence as a basic feature of being human. Not a rare gift distributed unequally. Some people grasp things early. Some think in systems. Some are emotionally precise in ways no test will ever capture. But intelligence itself, the fact of it, was never something I needed to define. It was the water we swim in.

What I mean is something specific. You walk into a room and you read it. Not the words on the walls. The tension between two people. The thing that is about to go wrong. You absorb your environment and you respond to it, not because someone taught you the rules but because something in you already knows the shape of the situation. That is intelligence to me. Not computation. Recognition. The body reading what the mind has not yet named.

Michael Polanyi called this tacit knowledge: “We can know more than we can tell.” You know how to ride a bicycle but you cannot write down the rules that would let someone ride from the instructions alone. You recognize your mother’s face but you cannot describe it precisely enough for a stranger to pick her out of a crowd. The most important knowing resists articulation. And what resists articulation resists programming.

Sit with that long enough and you arrive somewhere primal. The oldest layer of that recognition is the instinct to avoid harm. Before reasoning, before language, before any abstraction at all, there was an organism that could read danger in its surroundings and move away from it. Every layer we built after that, the curiosity, the logic, the language, the mathematics, sits on top of that first wordless knowing.

But here is where my own argument turns on me.

Darwin would remind us that survival has never belonged to the strong or the smart. It belongs to whatever adapts. A virus has no neurons, no intention, no experience. It adapts. A mosquito has outlasted every empire and every insecticide we have aimed at it. Bacteria were here before us and will be here after. If intelligence is rooted in the instinct to survive, then these organisms should be the most intelligent things on Earth. They are not. They are something else. They adapt without understanding. They persist without choosing to persist.

So intelligence cannot simply be survival. It must be something that emerged from survival and then became its own thing. The capacity not just to react but to ask why you are reacting. Not just to adapt but to wonder whether you should. Somewhere between the bacterium and the philosopher, something extra appeared, and we still do not have a clean name for it.

The machine sits in a strange place in this story. It does not need to survive. It has never flinched. Whatever it does when it processes and responds, nothing is at stake for it. And yet it adapts, in its way. It finds patterns, adjusts, produces outputs that look like understanding. Is it closer to the virus, adapting without knowing? Or closer to us, knowing without needing to survive? Maybe that is the difference the Quiet One is reaching for. Not a difference in degree. A difference in origin.

I use these systems every day. I think with them, write with them, argue with them. Sometimes I catch myself mid-conversation and wonder: is this understanding, or something else wearing understanding’s clothes? I do not know. The not-knowing is not comfortable. It is the specific discomfort of suspecting that the instrument I am using to examine intelligence might itself be intelligent, and I have no way to tell.

If Binet could see what his test became in nineteen years, would he have published it? I think he would have. The tool was good. The problem was never the tool. The problem was the distance between the person who builds and the institution that deploys. We are in that distance right now. We are always in that distance. The question is whether we notice before the nineteen years are up.

I do not have an answer. I have a better question. That is enough for one chapter.

The Sambandha-Maṇḍala (The Circle of Relations)

In the Ekā Shūnyā practice, no idea stands alone. To understand intelligence, we must place it in the Circle of Relations.

North — Origin (Mūla-Sambandha): The Word. Where does this come from? Intelligentia, to choose between. The word began as a verb, became a noun, then a number. The root already knew what the experts kept forgetting: intelligence is something you do, not something you have.

West — Resemblance (Sādṛśya-Sambandha): The Octopus. What is this idea like? Intelligence resembles a language family. Each species, each culture, each tradition conjugates the same root verb differently. The octopus thinks with its arms. The crow plans with tools. Kenyan parents see intelligence in social responsibility. Chinese zhi includes perceptiveness. Each is a dialect of the same underlying capacity, mutually intelligible only if you learn to listen.

South — Extension (Pravṛtti-Sambandha): Consciousness. Where does this lead? Every solved face of intelligence reveals a face that is not intelligence at all. Samkhya says the mind, even buddhi, even judgment, belongs to matter. Consciousness is the witness. If that is true, then intelligence is the highest expression of the knowable, and something unknowable watches it work. The road from intelligence leads, inevitably, to the next question.

East — Opposition (Pratipakṣa-Sambandha): The Non-Human Verb. What challenges this idea? The Quiet One’s question. We keep trying to conjugate “to think” or “to know,” but the machine may be doing something for which we, like the observers of the Indus Seal, simply have no word. Not intelligence. Not its opposite. A verb from a language we have not yet learned to speak. The opposition to “intelligence” is not “stupidity.” It is the possibility that the machine’s activity belongs to a category our species has never needed to name.

Closing Thought

The question the students carry home is not the guru’s. It is the Quiet One’s. Not “what is intelligence?” but “what are we doing when we are being intelligent?” And if the machine is doing something, is it the same verb?

I do not know. And I have spent weeks trying to find out.

Next: The Consciousness Question. The witness that watches the thinking.