What is intelligence?
The question seems simple, but perhaps it is only superficially so. Today we use the word with enormous ease, as if its meaning were obvious. We talk about intelligent people, intelligent devices, intelligent systems, intelligent cities, intelligent weapons. The adjective has spread so much that it hardly requires explanation anymore. And that's the problem: the more easily a word circulates, the greater the chance that it has begun to lose its definition.
When discussing artificial intelligence, for example, the immediate response is almost always to whether or not the machine is intelligent. But this question presupposes another, far more fundamental one: what is intelligence? Before deciding whether a machine possesses it, it would be prudent to define what we are talking about.
Alan Turing (1912-1954), whose authority is often invoked in these debates, was more cautious than many of his intellectual heirs. In his famous 1950 paper, he didn't dare define "thinking." He did something else. He replaced the original question with a more operational version. Instead of asking whether machines think, he asked whether a digital computer would be able to pass for a human in a written question-and-answer game. The maneuver was ingenious. But it was, nevertheless, a maneuver. The problem wasn't solved; it was circumvented.
This deviation is revealing. It shows how our era tends to proceed when it loses certainty about the essence of things. Instead of asking what something is, it asks what it does. Instead of seeking the nature of a faculty, its apparent effects are examined. What matters is no longer Being, but performance. And so Intelligence ceases to be what the mind is and becomes what it is able to display.
But it wasn't always thought that way!
There was a time when intelligence did not primarily mean quick thinking, efficiency in problem-solving, or skill in manipulating symbols. In the classical tradition, and particularly in the Thomistic one, intelligence concerned above all Truth. The human intellect was understood as the power by which man apprehends the true. Reason, in turn, was the discursive movement by which the mind passes from one known thing to another. The distinction is important. Reason walks; intellect sees. Reason discourses; intellect apprehends. Reason seeks; intellect achieves.
If this is forgotten, everything else becomes confused. For, in this context, Intelligence is not simply the capacity to reason well. Reasoning undoubtedly has its dignity. But it is not the ultimate term of Intelligence. It is its discursive instrument. The decisive point is another: Intelligence, in its highest sense, is the capacity to perceive the Truth.
This formulation sounds almost strange to modern ears. And it's not hard to understand why. The very word "truth" has lost its centrality among us. It hasn't disappeared, of course, but it has become uncomfortable. Today it seems more elegant to say that "everyone has their own truth." The phrase has become commonplace. And yet, it is intellectually devastating. For if everyone has their own, then Truth ceases to be that to which Intelligence must conform, and becomes merely a subjective experience, a private perspective, a psychological construct. Truth ceases to measure the intellect. The intellect begins to produce "truths."
Once this inversion is made, the very idea of Intelligence changes. It is no longer seen as an ordering towards truth, but as operational capacity. Intelligent becomes someone who responds quickly, who calculates well, who argues with agility, who finds effective solutions, who processes a lot of information, who exhibits high performance. The word is dragged from the plane of Truth to the plane of efficiency. And this semantic shift ends up producing very concrete philosophical effects.
Consider, for example, the extreme case of so-called idiot savants , a condition now appropriately designated by the more neutral expression savant syndrome . The phenomenon is well-known: people, often within the autism spectrum, reveal extraordinary abilities in very specific domains: prodigious memory, date calculation, musical reproduction after a single hearing, numerical speed, meticulous drawing. Such abilities are real, sometimes astonishing, and deserve admiration. But do they, without further ado, deserve the name of Intelligence? The answer depends, again, on what is meant by the word. If Intelligence is simply exceptional performance in some mental operation, then the answer is affirmative. But if Intelligence is the ability to perceive the Truth, the question becomes much less trivial.
For memory is not the same as judgment. Calculation is not the same as comprehension. Faithful reproduction is not the same as grasping meaning. A particular capacity can reach an extraordinary degree without resulting in full intelligence. There can be localized excellence without a vision of the whole. There can be brilliance without sufficient light. There can be, so to speak, a splendor of the faculty and at the same time a profound limitation regarding the apprehension of Reality.
It is precisely here that the modern use of the word "intelligence" begins to show its shortcomings. If we call everything that impresses cognitively intelligent, sooner or later we will lose the ability to distinguish between skill and wisdom, between reasoning and judgment, between mental agility and intelligence proper. And, once this distinction is lost, the result becomes visible in everyday life. People universally considered intelligent end up doing things that are clearly stupid. This is not an unsolvable paradox. It is simply a clash between two different meanings of the word.
In common usage, someone is called intelligent because they reason quickly, master abstractions, have a good memory, pass difficult tests, speak well, and argue better than average. All of this may be true. But none of these things, taken in isolation, is enough to define Intelligence in its highest sense. A person can be very capable in the discursive operations of reason and yet be very deficient in judging Reality. They can be quick to infer and slow to see. They can be brilliant in formal structures and blind to the closest Truth.
Popular wisdom understands this better than certain scholarly discourses. "How can someone so intelligent do something so stupid?" The question is excellent. And the answer is perhaps this: because what we call Intelligence today is often only a part of it. We call Intelligence what, at most, would be rational agility. We call Intelligence what would perhaps be better called discursive skill. We call Intelligence what, in the Thomistic tradition, would only be an aspect of rational operation, and not its highest end.
The same applies to machines. The expression "artificial intelligence" seems natural only because the concept of intelligence has already been previously reduced. If intelligence means the ability to manipulate symbols efficiently, recognize patterns, generate plausible answers, and solve delimited tasks, then there is really no great difficulty in applying the term to machines. But it is worth realizing the price of this naturalness. It was not achieved because we elevated machines to the classical concept of intelligence. It was achieved because we lowered intelligence to a functional level compatible with the performance of machines.
In this sense, the machine is less a metaphysical novelty than a semantic mirror. It shows us what we have done to the word "intelligence." If we no longer demand from it any intrinsic relationship with Truth, if we no longer understand it as the capacity to apprehend Reality, if operational competence is enough for us, then it really becomes easy to talk about artificial intelligence. But this says less about the machine than about our conceptual decay. Because a machine can calculate without understanding, answer without judging, correlate without knowing, imitate without seeing. And yet, it will be called intelligent in an era when many men are also intelligent only for equally impoverished reasons.
Perhaps, therefore, the solution to the problem lies in recovering an old and very sensible distinction. Reason has its place. Calculation has its place. Memory has its place. Technical skill has its place. But Intelligence, in its highest sense, is not exhausted by any of these things. Its most proper criterion remains its relationship with Truth.
This does not mean demanding infallibility from men. That would be absurd. Human intelligence is finite, fallible, subject to error, illusion, passion, ignorance, and weakness. But it is one thing to recognize its limitations; it is quite another to redefine it in such a poor way that any functional simulation comes to deserve the same name.
Perhaps it would be better, then, to preserve the concepts of Intelligence and Truth as the Thomists understood them. Not out of an archaeological spirit, nor out of scholastic attachment to ancient formulas, but out of simple precision. If Intelligence is the capacity to perceive Truth, much that is called intelligent today will have to be renamed. Certain people will remain admirable, but for more specific reasons. Certain machines will remain useful, but without usurping human attributes in a strong sense. And certain paradoxes will disappear on their own.
The problem has never been that supposedly intelligent people do stupid things. The problem is that we no longer know, with sufficient clarity, what intelligence means. Perhaps, in the end, we have traded the ability to see the Truth for the ability to impress. And then we call that intelligence.
Cover photo:
