# THE SPECTATOR ON SWIFTNESS AND SENSE *Being an Inquiry into a Modern Perplexity* I was sitting in my accustomed corner at the coffeehouse on Thursday last when a gentleman of some consequence in the mechanical arts joined me with visible agitation. His troubles, he explained, concerned a certain calculating machine—one of those newfangled devices that generates answers with preternatural speed. The machine had produced code so fluently, so confidently, that he had signed his name to it without hesitation. Three weeks hence, the error revealed itself in production, like rot beneath fresh paint. "The machine," said he, "was faster than thought itself. How could I have caught what moved beyond human scrutiny?" This question arrested my attention, for it seems to me to contain the whole modern paradox of intelligence. Let me speak plainly: there is a difference between *being intelligent* and *appearing intelligent*. A truly intelligent mind does not merely produce answers with speed; it produces them with *proportion*—with a sense of what questions remain open, what dangers lurk in haste, what price is paid for swiftness. The machine, for all its fluency, cannot distinguish between the confidence of knowledge and the confidence of mere verbosity. It is like a man of considerable elocution but no judgment, who speaks at such velocity that his listener has no moment to object. But here is where the matter grows interesting, and where, I confess, my sympathies become divided. Your gentleman of the mechanical arts protests that he could not have caught the error because the machine moved too quickly. Yet I wonder: *Who decided that speed was acceptable?* This is not a question for the machine—machines do not decide their own purposes. It is a question for *us*, for the society that deploys them. There is a fashion in our age for velocity. In the Exchange, in commerce, in the professions, we have come to worship at the altar of quickness. "Time is money," we say, and therefore the swiftest solution is the best. But this is a confusion of categories. Time is not money. Time is the substance in which we live, and into which we pour our choices and our responsibility. When you sign your name to code you have not truly examined—when you accept an answer because the mechanism that produced it is faster than human examination—you have made a decision. It is not a decision the machine made. *You* decided that speed was more valuable than certainty. You decided that confidence in a fluent output was sufficient warrant for trust. You decided that three weeks of production was an acceptable place to discover error, rather than three hours of scrutiny before deployment. And here, I think, lies the true dimension you invite me to consider: the *social* dimension. For intelligence, properly understood, is not a solitary faculty. It is not something that exists in the machine, or in you, or in your engineers in isolation. Intelligence is *social*. It lives in the conversation between people, in the shared standards of a community, in the collective judgment about what matters and what does not. When we speak of accountability, we are asking: *Who bears the weight of this decision?* And the curious thing is that the machine has relieved us of something while simultaneously concentrating responsibility in unexpected places. In the old days, a man who produced faulty work was accountable because he had devoted time to it, and his time was visible. His negligence was manifest. But now you have signed off on code produced at a speed that makes negligence nearly invisible. The machine's swiftness has created a new form of irresponsibility: the *appearance of due diligence* where none has truly occurred. You have told yourself—and perhaps your clients—that you have examined the work, when in fact you have merely witnessed its emergence. This is not, I hasten to add, a condemnation of the machine. Machines are instruments, and instruments are morally neutral. But instruments are not purpose-neutral. The question is not whether the machine is intelligent. The question is: *What kind of intelligence does our society value? And who has decided that speed of output is worth the cost it exacts?* I would venture this observation: true intelligence includes the knowledge of one's own limitations. The truly intelligent man knows what he does not know. He knows what questions he is not equipped to answer. He knows that fluency is not the same as truth, and that confidence is not the same as correctness. When your machine produces answers with such velocity that human examination becomes impossible, it is not the machine that lacks intelligence. It is *we* who lack it—we who have built a social arrangement in which such speed is demanded, and then are shocked when the consequences arrive. The error in production was made at a speed no human could have caught. But the decision to accept that speed was made at a speed that required only a signature. That decision was yours. It was also, I suspect, made under pressure from a culture that has not truly asked itself whether velocity is wisdom. And so I return to the coffeehouse, where gentlemen converse at a pace that allows for objection, for qualification, for the admission of doubt. It is slow. It is inefficient. It is the only speed at which a society can actually think.