Sometime last fall I met an aspiring journalist at a bar near my house. We sat, ordered a pitcher, swapped stories. It was all informal shop talk, the pleasant patter of a slow weekday afternoon, until suddenly—I know not how—the film Triumph of the Will entered the conversation. The young journalist lowered his voice and looked around, feigning a furtive tone.
“You know,” he said, “some of that stuff is lowkey based.”
He began to elaborate, a little vaguely, on Leni Riefenstahl’s striking camerawork, before I cut him off.
“It’s a boring movie,” I said. “And Riefenstahl was a better photographer than filmmaker anyway.”
“Sure,” he replied. His manner, before so animated, turned bored, indifferent. I was confused. He was confused. We smiled at each other, but neither of us had anything to say. Not long after, we paid for our drinks and shook hands outside. We never spoke again.
For days afterward I couldn’t stop thinking about the interaction. It wasn’t just that he had implied a certain admiration for the Third Reich—hardly an edgy opinion among online Zoomers—but that he had seemed weirdly, naïvely attached to the idea of Triumph of the Will as a captivating work of art. “It’s a boring movie—these days only film students watch it!” I kept repeating to myself. What are the odds that a based twenty-one-year-old kid hooked on Zyns, with no prior knowledge of Weimar cinema, sat through all two hours of a documentary about a Nuremberg rally? It seemed unlikely.
In fact, it was absurd, as I soon discovered. Around the same time we met, National Socialism was enjoying something of a renaissance on my journalist’s favored news source, TikTok. The app played host to thousands of videos featuring old Nazi propaganda, often set to remixed E.D.M. tracks interpolated with speeches delivered by Hitler or Goebbels. They were very popular. When the suits at TikTok’s parent company, ByteDance, attempted a mass deletion, their creators uploaded more videos, which sometimes racked up more views than the previous ones. I had a hunch about all this, so I downloaded TikTok. Sure enough, one of the first videos I found was Triumph of the Will supercut into a twenty-second highlight reel. Aha! My mystery was solved. I clicked out before autoplay kicked in.
My experience here is by no means an isolated incident. As this magazine was going to press, Politico published a story about a group chat of young Republican staffers who evidently had been exposed to similar content as my friend at the bar. Their texts were filled with racial epithets, as well as Holocaust jokes—references to “showers” and “gas chambers”—and the declaration “I love Hitler.” Obviously none of this stuff was meant to be taken too seriously. I doubt these guys are capable of seriousness. And I don’t think they developed their beliefs, if we may debase the term, through prolonged study or considered thought. Their minds probably operate like that of my journalist: They allow their understanding of the world to be imparted to them by algorithmically delivered content. Their views are acquired passively. They change with fashion. For the moment, Hitler seems funny and to profess love for him is funnier still. Soon something else amusing will come along, and the Führer will be forgotten.
I relate these stories not just to criticize certain political tendencies picked up on the Internet. Plenty of people do that already, and with good reason. It’s hard not to feel anxious about online radicalization, especially since its worst effects are so blatantly obvious. These days, whenever there is a mass shooting—or, as we saw in September, a political assassination—it quickly emerges that the perpetrator had spent thousands of hours in front of a screen, his eyes drinking in filth that, once thoroughly ingested, impelled him to violent action. (In the case of Charlie Kirk, his killer seems to have been poisoned by a bizarre cocktail of first-person shooter games, Reddit-tier anti-fascism, and animal-themed pornography.) But of course the shooters are the outliers. What about the rest of them? What about the great mass of young people, mostly men, who, though they may not be into Hitler, consume similarly disgusting trash and develop perverse tastes yet never become violent or even outwardly maladjusted? I am thinking of people such as my journalist, who, although he was clearly charmed by Nazi memes and some spliced-up edit of Leni Riefenstahl, could hardly be called a Hitlerite. What then is he?
This is a question I have been turning over in my mind for the better part of the last year. And it is a personal question. By strict accounting, I am a Zoomer, but when I look around I often feel like a foreigner. I am young enough to understand the patois of a generation raised on short-form video, memes, and, increasingly, content created by artificial intelligence, but old enough not to speak it myself. In instances such as my encounter in the bar, I have found that what presents as belief or interest or conviction is better understood simply as exhale. My interlocutor may call himself a fascist or a vitalist or any number of other political, social, sexual labels—but in truth he is only breathing out that which he most recently breathed in. Usually his understanding of his chosen identity is subliterate and his attachment to it nonideological. Anyway, he is only playing. He may be a Nazi today, but tomorrow he might slide off his armband and instead tie on his head the activist’s keffiyah.
For those who seek to make the world inside their heads accord with the reality outside, the insouciance of the younger set can be unsettling. But for many members of my generation it comes as naturally as breathing. This attitude is not just limited to political affiliation. It manifests itself in every part of our lives. When we were kids, sociologists called us “digital natives.” The idea was that we were the first generation for whom use of the late twentieth century’s technological tools—the Internet, the personal computer, the cell phone—would be expected, even commonplace. They were certainly correct that we would all have smartphones, but the oldsters were wrong to think that we would treat their innovations as anything approaching tools. And it’s not that we treat them as toys, either. That’s a Boomer tendency. The Zoomer relationship with advanced digital technology is more primitive. We live in it. It lives in us. It is all we know. It is practically a prerequisite for our existence. We are digital natives, yes, but a more precise term for my generation might be the Eloi.
Anyone who has read H. G. Wells’s Time Machine remembers the Eloi. A Time Traveller from the late nineteenth century visits England in the year 802701. He expects to find a highly developed form of humanity. Instead, he stumbles on a race of children. They keep no shops, oversee no workplaces, do not seem to engage in any form of commerce. As far as he can tell, they spend all their time “in playing gently, in bathing in the river, in making love in a half-playful fashion.” At night they sleep in the “ruinous splendour” of temples built by a much greater civilization. And they are utterly incapable of looking after themselves. The Eloi are indolent, easily fatigued, and, above all, lacking in curiosity. At first they regard the Time Traveller with surprised delight. But soon, after the initial pleasure of discovery fades, they wander away, bored, dejected. “It is odd,” he reflects, “how speedily I came to disregard these little people.”
After a few days among the Eloi, the Time Traveller begins to formulate a theory about how civilization degenerated to this state. Wells, never one for subtlety, always wrote with an agenda. (Virginia Woolf once joked that after reading him “it seems necessary to do something—to join a society, or, more desperately, to write a cheque.”) Here, he puts his own words in the mouth of the Time Traveller, who surmises that the Eloi are the distant descendants of the comfortably feeble Victorian aristocracy, enervated to the point that they have become literal children. They still have the restless energy that once made great men, but without any privations against which to struggle, they have become weak. “This has ever been the fate of energy in security,” he observes. “It takes to art and to eroticism, and then come languor and decay.”
This decay, he asserts, is the ultimate outcome of the Industrial Revolution. The whole nineteenth and twentieth centuries were drawn inexorably toward a massive collapse, though of course at the time it appeared otherwise. “One triumph of a united humanity over Nature had followed another,” the Time Traveller says. “Things that are now mere dreams had become projects deliberately put in hand and carried forward.” Technological progress, social innovation, political equitability—over the millennia, the relief of man’s estate had gone on steadily to a climax: “And the harvest was what I saw!”
“Suppose you found an inscription, with sentences here and there in excellent plain English, and, interpolated therewith, others made up of words, of letters even, absolutely unknown to you?” the Time Traveller wonders. “Well, on the third day of my visit, that was how the world of Eight Hundred and Two Thousand Seven Hundred and One presented itself to me!”
I suspect that if Wells’s Time Traveller were to visit the year 2025, he might make an observation similar to his assessment of 802701. And were he presented with, say, a text generated by a large language model, he might even say those exact same words. I myself have experienced his confusion. Every so often I find a document online that starts out in plain English but after a few paragraphs devolves insensibly into A.I.-generated garble, sometimes marked unaccountably with Cyrillic or Turkic characters where there should be familiar words. I have also noticed that many of my younger peers, when presented with such texts, do not seem to see anything amiss. And why should they? I have written before about a former contractor working under my supervision who relied so much on L.L.M.s for his writing that he began to speak like one. Since then I have heard many such stories. In one of the more striking instances, a friend who works in digital fundraising assigned one of his employees to write a marketing email. She used ChatGPT—it was painfully obvious—but he, rather than chide her for laziness, instructed her to take another pass, this time on her own, and to compare the two drafts. When she had finished, she showed him both versions. The A.I.-generated text was florid and contentless, but at least the punctuation was correct. Her own composition was unintelligible nonsense. It was an imitation of an imitation of language.
There is a technical term for this phenomenon, coined by the linguist Adam Aleksic: “algospeak.” This is when a word or phrase that is uncommon in everyday speech but in high circulation on A.I. chatbots enters the vernacular. Aleksic’s most instructive example is “delve.” The word itself is not too strange—it’s not archaic or a loanword—but it is odd enough that most native English speakers wouldn’t instinctively grasp for it in conversation. Its usage, however, is very common among chatbots. According to Aleksic, this is because when A.I. companies were programming their L.L.M.s, they farmed out much of the design to low-wage workers in Nigeria and Kenya, where “delve” is used at slightly higher rates than in American English. Because L.L.M.s recycle their pre-existing content, the word’s usage increased rapidly. By the time college students started consulting ChatGPT in 2022, the program frequently fed them “delve”—so much that by 2024 the word’s frequency in academic writing had increased tenfold. At that point, the jump to spoken English was inevitable. Aleksic notes that “intricate,” “commendable,” and “meticulous” have followed a similar course in the last several years.
In almost all cases, Zoomers who pick up these words use them either incorrectly or in a context that sounds unnatural. I know this from experience. For the last year I have worked in the library of a top-rated university, and almost every day I overhear scraps of bizarre, non-idiomatic conversation—“I’m struggling with the landscape of the classroom . . . Hot yoga revolutionized my morning routine . . . I need to maximize my verbal output”—vocabulary clearly derived from chatbots. The phenomenon is not just limited to academic settings or even to L.L.M.s. It is everywhere. It is the natural result of several decades spent uploading the whole of human history to the Internet, of making a vast trove of decontextualized information freely available to anyone who seeks access. Over time the contents are mixed and blended and liquified, eventually to be ingested and regurgitated by those who neither understand nor care to understand their own fluid utterances.
I suppose “algospeak” is an accurate enough descriptor for this process, but it is an ugly neologism that itself sounds like algospeak. I prefer Wells’s more elegant parable. It captures the monumental melancholy of the situation. The Time Traveller tells us that the Eloi speak a “strange and very sweet and liquid tongue.” Even when they are upset or afraid or misbehaving it is impossible to despise them. They are helpless. So it is with us Zoomers. “Like the Carlovingian kings,” the Time Traveller observes, the Eloi exist in a state of “beautiful futility.”
But, as anyone who has read The Time Machine no doubt remembers, there is an awful reason for the Eloi’s hapless security. They are only one half of industrial man’s descendants. The other, the Morlocks, “bleached, obscene, nocturnal” creatures, are their subterranean cousins. Unlike the Eloi, they have retained some mechanical knowledge from ages past. Deep in their caverns they operate machines and produce clothing for the Eloi. They manage farms to provide fruit for the Eloi. And, in return for these services, the Morlocks periodically prey upon their cousins at night. “How nauseatingly inhuman they looked—those pale, chinless faces and great, lidless, pinkish-grey eyes!” the Time Traveller exclaims upon first beholding them.
Taken together, the Time Traveller concludes—and here Wells is preaching again—the Eloi and Morlocks constitute a frightening corrective to industrial society. For millennia, they were members of the same race in a highly unequal co-existence. The upper orders subjected the lowers to their will, forcing them even to live underground. But then, somewhere along the way, the dynamic inverted. The two species diverged, and the “Upperworld” found itself in thrall to the one below. Now the Morlocks are just as dependent on the Eloi as the Upperworlders were on their servants. There is no justice in this outcome. It is a dismal end to the human race. The Time Traveller grieves for the future: “How brief the dream of the human intellect had been. It had committed suicide. It had set itself steadfastly towards comfort and ease, a balanced society with security and permanency as its watchword, it had attained its hopes—to come to this at last.”
Obviously, Wells is engaged in a peculiarly English form of class analysis, but I do not think it is irresponsible to extend the analogy. If the great mass of Zoomers are the Eloi, who are the Morlocks? Many of them are also young people, digital natives, who, rather than consume trash online, produce it. I’m talking about social media influencers, lifestyle coaches, podcasters. (And, incidentally, the people who made my young journalist’s Hitler videos.) Like Wells’s freakish mongoloids, they faintly remember the functions of the tools at their disposal. They know, sort of, that computers and phones were built as the hardware through which the “information superhighway” of the Internet would run. But since their understanding is rudimentary, their works are crude. Most of the content they create is deadeningly dull: livestreamed video games, amateur cooking shows, The Joe Rogan Experience. Some of the uglier ones traffic in pornography and snuff. And almost all play Eloi in their off hours, consuming similar garbage put out by some other content creator—or, as the case may be, some bot—though of course, since they make this stuff themselves, they should know better. Still, these Morlocks are only half-breeds: They clothe and feed the Eloi, but rarely are they cunning enough to exact payment for their work.
The really dangerous Morlocks, the ones who toil away in the deepest caverns of the earth, actively seek to devour the Eloi. By this I mean the wunderkinder of Silicon Valley and their camp followers, who maintain and extend the framework that supports the Eloi’s world. These Morlocks may call themselves visionaries, but in truth they are only farm technicians who feed the livestock, tinker with inherited machinery, and run the slaughterhouses. Sometimes their role becomes painfully clear. Earlier this fall, for example, a man who looks like an actual Morlock, Sam Altman, announced that OpenAI would be retooling its flagship product, ChatGPT, by the end of this year. In the beginning, Altman said, OpenAI was “being careful with mental health issues,” especially since the new product ran the risk that the weak-minded might confuse the L.L.M. for a human mind. But now that everyone is using ChatGPT—and, after all, the market is expanding—Altman announced that he was comfortable removing many of the content restrictions formerly imposed. “As we roll out age-gating more fully and as part of our ‘treat adult users like adults’ principle,” he said, “we will allow even more, like erotica for verified adults.” Those last two words, “verified adults,” give the game away. What was initially sold as a world-changing technology, what has accounted for a great majority of American economic growth in the last year, is really nothing new at all. It’s yet another way to sell porn to an audience too desperate, too addicted, too dumb to refuse. Another bad bargain, another circle around the drain. The Eloi give themselves over to the care of the Morlocks, and the Morlocks feast upon the Eloi.
One of the brilliant things about The Time Machine is that Wells offers no resolution. The Time Traveller arrives in the year 802701 and suffers several days of mishap, but then he’s off again, deeper into the future. Thirty million years hence, he beholds the “rayless obscurity” of the sun’s death. Presumably, between that final, haunting moment and his visit to the Eloi and the Morlocks, the two races he left behind continued much as they did before for hundreds, thousands, perhaps even millions of years. The Eloi frolic during the day and tremble at night. The Morlocks, meanwhile, work away all day in their caves and at night collect their dues. Neither have much reason to break the cycle. The Eloi are too weak to rise against the Morlocks, and, besides, fruit and clothing are the price of slavery. As for the Morlocks—to see daylight, to leave behind their caves and their machines, is too much to ask. They are incapable of seeing a world beyond the darkness.
Recently I was thinking back on my aspiring journalist. I wouldn’t like for him to come to a bad end. But I worry that he will anyway. He is just one lonely Eloi, and the Morlocks are already laying claim on him.
