Andrei Cornea: How AI Will Kill Us

I do not believe that AI will kill us violently, freeing the Earth from the presence of the human race, as some people—among those most involved in perfecting AI, from OpenAI, Anthropic, or Meta—have begun to predict, if we fail to control it very quickly. We probably will fail to control it, mainly because of the competition between the United States and China. And yet, I do not believe that a violent disappearance awaits us. What seems to be coming is rather a gradual degeneration: once left to ourselves, we have for some time now been becoming, under the influence of all digital and automated systems, not only AI—from PCs to mobile phones and all the way to GPS—ever more indolent, ever more dependent, more helpless, increasingly inept, less intelligent—in short, ever stupider. This process of dumbing down began—it seems—a long time ago, at least decades ago. That is what some studies of IQ testing over time appear to show.

It is no wonder: when not used sufficiently, the mind, with its various abilities, begins to “flatten out,” losing the acuity and ingenuity it developed over many millennia of harsh competition. With electricity, refrigerators, running water, electric ovens, supermarkets, malls, increasingly sophisticated automobiles, airplanes, and all sorts of smart devices, why should we still bother too much to learn how to survive in an uninhabited forest or on a desolate plain? The primitive person, deprived of so many conveniences, was alert and intelligent, had extraordinarily sharp senses, and possessed fundamental knowledge about plants and animals that ensured their survival. Why would we need all these skills when a computer and a bank account take care of all our everyday needs?

For example, why should we preserve a good memory by memorizing thousands of lines of poetry, like the ancients? We always have a Google search at our disposal. I do not even need to learn the multiplication table: any small calculator performs all the calculations in an instant. I do not have to remember dozens of telephone numbers, since I have an electronic address book. Soon, even routine accounting tasks will be carried out by machines. Why should drivers learn to navigate independently in a city? They have GPS and WAZE, so they can completely ignore the city’s topography: they will surely reach their destination. Some time ago, we replaced handwriting with typing; but soon even writing in general will no longer have any use: there are already quite a few people who dictate to a machine, which instantly transcribes their words into graphic signs. Is it even still worthwhile to have books in general? For now, yes, but, as can be seen, more and more people are living comfortable and prosperous lives without them. Reading books, perhaps even reading in general, will soon become a cool form of snobbery. Thus, autonomous vehicles will make the driver’s manual and psychological abilities useless, translation machines will make learning foreign languages unnecessary (they have already begun this operation quite successfully), and audio-video systems for transmitting information will spare us the task of writing and reading. Teachers will be replaced by robots, while AI systems will make most medical diagnoses. Perhaps, someday, we will be given a choice between being operated on by a human (more expensive) or by a robot (cheaper), and we will hesitate. Machines will also compose music or paint—perhaps they will not produce a Chopin or a Klimt, respectively, but it will not be bad either. Then why would we still turn to “geniuses”? Even this politically incorrect word will have to be used as rarely as possible in the AI “brave new world.”

Until then, we are already beginning to rely more and more on AI to make choices and decisions: where to spend our vacation, where and how much to invest, what courses of study to plan for our child, and so on. It is convenient and useful (even if errors sometimes occur); but in this way, are we not unlearning our own searching and effort? If we are no longer able to decide for ourselves, do we not become like minors incapable of moral autonomy?

The domination of screens and clicks—can we still live without it? Less and less. If only we understood what lies behind our clicks! But of course most of us have no idea, contenting ourselves with the fact that the machines perform their task, usually very efficiently. (Lately, it seems that even some of the creators of AI no longer understand everything the “monster” does and how it works.) But when, for a moment, they no longer perform it, what do we do? We are desperate when there is a power outage, because we suddenly perceive our vulnerability, something none of our ancestors had. Our dependence has become terrible, depriving us even of those qualities of which humanity was most proud: ingenuity, creativity, sound intuition, the courage to innovate, and moral autonomy. Not to mention the fact that, dependent as we are on networks, we are becoming increasingly incapable of socializing face to face. Once confronted with a different opinion, we immediately become hysterical; we are easily frightened, dependent on rumors going viral on the Internet, always easy to manipulate and deceive. The herd mentality, which several hundred years of humanistic education had managed to keep somewhat under control, has now infected the networks. And how could we bear ourselves if we did not compulsively check our mobile phone upon waking, even before coffee, in order to hear and comment on the Internet’s perpetual gossip?

In any case, to annihilate us, AI does not need to resort to some apocalyptic scenario from thrillers—nighttime laser attacks, unleashing pandemics, destroying sources of water and food, and so on. The simplest, but also the most intelligent course of action (and in this way it would prove itself to be at the level of its reputation) would be to let things continue as they have until now: soon we will be good for nothing.