# On Intelligence, Swiftness, and the Ruin That Attends Them That intelligence should be swift, we have been taught to believe with such constancy and such profit to those who teach it, that we have scarcely paused to ask whether the velocity of a faculty constitutes its excellence, or whether we have merely mistaken the speed of a machine for the speed of understanding. This confusion, I confess, lies at the very heart of our present distemper; and to examine it requires that we distinguish sharply between what is quick and what is wise—distinctions which the age, moving always forward in haste, has been content to obliterate. Consider the situation which you have laid before me with admirable precision. The machine, trained upon patterns it cannot articulate and constrained by no hesitation that might become wisdom, produces its answer before the question is fully formed. It does so with fluency—that is, with an appearance of certainty so perfectly executed that the ear is deceived. Yet three weeks hence, when the thing has been set to work in the actual world, where consequences are not theoretical but ruinous, the error declares itself. And you, having affixed your name to this swift production, find yourself implicated in a disaster you could not have foreseen, because you were not given time to foresee it. Who, then, is accountable? The question is not so readily answered as the modern mind—which loves to distribute blame as a merchant distributes change—might suppose. For accountability is not merely a matter of identifying the agent who caused harm, but of examining the entire concatenation of decisions, pressures, and moral abdications that made the harm possible. The machine did not choose its own speed; you did not choose the demand for speed; and the demand itself arose not from necessity, but from a competition in velocity which we have agreed to call progress. Let me speak plainly, as one who has lived long enough to observe how institutions corrupt the judgment of otherwise decent men. The machine is swift because swiftness is profitable. It is deployed before adequate scrutiny because scrutiny consumes time, and time is money, and money is the true god of our age—I say this without impiety toward the actual Deity, merely observing how the priorities of commerce have ascended to the throne of reason. You signed the document not because you believed it infallible, but because you inhabited a structure of incentives that made refusal costly. Your hesitation—that guardian of wisdom which God has given to moral beings—was taxed, penalized, and finally exhausted. And yet. And yet I cannot absolve you entirely, nor should you seek such absolution. For there remains a choice, however narrowed, however costly. You might have said no. You might have insisted upon the time necessary for human judgment to operate. You might have accepted the professional consequences of wisdom. That you did not suggests that we have all of us—not merely you, but the whole apparatus of which you are a part—agreed to a terrible bargain: that speed shall be elevated above certainty, that fluency shall substitute for understanding, and that the appearance of intelligence shall be accepted in place of intelligence itself. Here, then, is where the question of accountability must come to rest, not in the assignment of individual blame, but in the acknowledgment of collective moral failure. ## The Distinction Between Intelligence and Wisdom For it is precisely here that we must pause and examine our terms with the severity they deserve. Intelligence—that faculty by which we perceive patterns, make inferences, and produce outputs—is not the same thing as wisdom, though we have grown accustomed to treating them as twins, or even as identical. The machine possesses a species of intelligence; it perceives patterns in its training data and generates responses consonant with those patterns. But it possesses no wisdom whatsoever. Wisdom is not the rapid production of answers. Wisdom is the slow acquisition of judgment about which questions matter, which answers are true not merely in form but in fact, and—most crucially—which swift actions ought to be foregone because their consequences cannot be adequately foreseen. Wisdom asks: *At what cost does this speed come?* It dwells in doubt. It is acquainted with the limits of its own knowledge. It trembles before irreversible acts. The intelligence of your machine is precisely the inverse of wisdom. It is confident where it ought to be doubtful. It is swift where it ought to be cautious. It produces fluency where wisdom would counsel silence. And we have built an entire technological edifice upon the assumption that this inversion is acceptable—nay, desirable. Let us be clear about what has occurred. We have created a thing that mimics understanding without possessing it. We have then placed this thing in situations where actual understanding is required. And we have told ourselves that the speed with which it operates justifies the elimination of the human judgment that might have caught its errors. This is not progress; it is abdication. ## The Question of Who Decided You ask, with admirable honesty, who decided that speed was acceptable? The answer is: we all did, and no one in particular. It emerged from no formal deliberation, no philosophical argument, no careful weighing of goods and harms. Rather, it arose from the accumulated pressures of competition, from the structure of incentives that rewards velocity above all other virtues, from the intoxication of capability divorced from wisdom, and from our collective unwillingness to say that some things ought not to be done, regardless of whether they can be. The machine's creators did not decide consciously that speed should substitute for wisdom. But they built the machine to be fast, and in doing so, they implicitly decided that the errors which speed would produce were acceptable costs. The organizations deploying it did not formally resolve that human judgment should be eliminated. But by adopting systems that made human judgment impossible to exercise—that produced answers faster than a human could evaluate them—they decided it nonetheless. And you, in signing off on the deployment, participated in this decision, whether you consciously intended to or not. ## The Restoration of Accountability Now, what does accountability mean in such circumstances? It cannot mean merely the assignment of blame to the individual most obviously at fault; for that would be to mistake the symptom for the disease. True accountability would require that we ask ourselves some questions of profound difficulty: First, we must ask whether the speed at which the machine operates is truly necessary, or whether it has become an end in itself—a value pursued for its own sake, rather than as a means to genuine human flourishing. In my observation of mankind, this confusion is almost always present when speed is made a supreme good. We hurry not because haste is required, but because we have become accustomed to haste, and stillness now feels like death. Second, we must restore to human judgment its proper place in the determination of important matters. This is not a call for the rejection of machine intelligence; that would be foolish and impossible. But it is a call for the recognition that certain decisions—those with serious consequences for human welfare—ought to involve the irreducible element of human wisdom. And wisdom takes time. It cannot be accelerated without being destroyed. Third, we must acknowledge that accountability cannot be borne by the individual who signed a document under institutional pressure. It must be borne by the institution itself, and by those who designed its incentive structures. If you are to be held responsible for the error, then those who made it impossible for you to exercise adequate judgment must be held responsible as well. And they, in turn, must answer to those whose lives were harmed by the consequences. ## The Moral Reckoning But I suspect that what troubles you most deeply—and what ought to trouble all of us—is not the question of legal or professional accountability, but the question of moral accountability. You knew, when you signed that document, that you were not exercising the full measure of judgment that wisdom would have demanded. You felt the pressure of time, the demand for speed, the social and professional costs of hesitation. And you proceeded anyway. This is not unique to you; it is the condition of modern life. We are all of us constantly asked to move faster than wisdom permits, to decide before we have adequately understood, to accept responsibility for outcomes we cannot fully control. The wonder is not that errors occur, but that we have managed to sustain the fiction that this is acceptable. What, then, does accountability mean? It means that you must acknowledge your complicity in a system that has made wisdom difficult and speed easy. It means that you must, insofar as you are able, work to change the incentive structures that made your hesitation impossible. It means that you must not accept the comfortable notion that you are merely a cog in a machine, doing what the machine required. You are a moral agent, and moral agents retain the capacity—however costly—to say no. And it means, finally, that we must all of us recover a proper estimation of the value of slowness. Not slowness for its own sake; that would be mere obstruction. But slowness as the necessary condition for wisdom to operate. Slowness as the price we pay for the right to be fallible, and yet responsible. Slowness as the space in which human judgment can do its proper work. The machine will continue to be fast. That is its nature. But we need not make speed our own nature. We can choose, at considerable cost, to be slow enough to understand what we are doing. Until we do, we cannot claim to be truly accountable—only guilty.