The “lumpenproletariat,” according to “The Communist Manifesto,” is “the social scum, that passively rotting mass thrown off by the lowest layers of the old society.” Lower than proletariat workers, the lumpenproletariat includes the indigent and the unemployable, those cast out of the workforce with no recourse, or those who can’t enter it in the first place, such as young workers in times of economic depression. According to some in Silicon Valley, this sorry category will soon encompass much of the human population, as a new lumpenproletariat—or, in modern online parlance, a “permanent underclass”—is created by the accelerating progress of artificial intelligence.
The idea of a permanent underclass has recently been embraced in part as an online joke and in part out of a sincere fear about how A.I. automation will upend the labor market and create a new norm of inequality. In an A.I.-dominated future, those with capital will buy “compute” (the tech term for A.I. horsepower) and use it to accomplish work once done by humans: anything from coding software to designing marketing campaigns to managing factories. Those without the same resources will be stuck with few alternatives. A sense of dread about this impending A.I. caste system has created a new urgency to get ahead while you still can. “You have 2 years to create a podcast in order to escape the permanent underclass,” one Silicon Valley meme account, @creatine_cycle, posted recently on X, suggesting that perhaps fame can still save you. “Honestly if you don’t want to be a part of the permanent underclass you should probably ship slop asap,” another person posted, using the slang term for any A.I.-generated or augmented content; in other words, start leaning in to A.I. products or stay poor forever. The creator of @creatine_cycle is Jayden Clark, a former musician turned entrepreneur working in San Francisco. His niche posts satirize the id of the tech industry, which he has seen change radically since the advent of the A.I. gold rush. In the future that A.I. hustlers envision, Clark told me, “nobody’s working anymore.” He continued, “whoever hasn’t gotten in, you have no other chance to climb the ladder.”
The worries of a looming deadline for employability stem in part from an influential essay, published last year by the researcher and former OpenAI employee Leopold Aschenbrenner, predicting that A.I. will reach or exceed human capacity in 2027. Aschenbrenner writes that it is “strikingly plausible” that, by then, “models will be able to do the work of an AI researcher/engineer.” At that point, technological progress would become self-reinforcing, operating on a runaway feedback loop: A.I. would build more powerful A.I. on its own, rendering humans superfluous. Nate Soares, a prominent A.I. pessimist, the president of the Machine Intelligence Research Institute, and co-author of a recent best-selling book on A.I. titled “If Anyone Builds It, Everyone Dies,” told me, “In Silicon Valley, it’s like everyone’s seen a ghost.” He continued, “We don’t know how long there is on the clock” before the dawn of full A.I. automation. Regardless of his own existential fears about the technology, Soares said, “people should not be banking on work in the long term.” The tech jobs may be the first victims, not unlike how Frankenstein’s monster killed its creator. Then comes the wider world of digitized labor: writing e-mails, filling out spreadsheets, making presentations. Finally, self-innovating A.I. will develop intelligent machines to better perform physical tasks. Whatever A.I. can do better, it will, according to Soares: “Humans are just not the most efficient arrangement of matter to do almost any job.”
The fear of a permanent underclass stems in part from the progress that A.I. has already made. Whether we want it or not, the technology is creeping into our daily lives. Both OpenAI and Meta have recently launched feeds of purely A.I.-generated videos, auguring an era of social media in which even the most elaborate content we consume is no longer created by humans. Workaday corporate software such as Salesforce is being amped up with A.I.-powered “agents” that can independently perform tasks for users. Waymo cars drive themselves through the streets of major cities. Some economic statistics are already hinting at a hiring slowdown, particularly among new workers; this year, the unemployment rate for recent American college graduates surpassed the national average, an anomaly that an Oxford Economics report primarily blamed on A.I. automation. Entry-level software engineers are facing particular difficulty. Jasmine Sun, a former employee of Substack who writes a newsletter covering the culture of Silicon Valley, told me, of tech workers, “Many are really struggling and can’t find even a normal salary, and some of the people are raking it in with these never-seen-before tech salaries. That creates this sense of bifurcation.”
The new desirable employee archetype is a “cracked twenty-two-year-old,” Sun said, using the slang for a hyperproductive, extremely online programmer who might work “nine-nine-six,” a term adopted from workers in China that refers to a schedule of 9 A.M. to 9 P.M., six days a week. The only way to escape the permanent A.I. underclass, ironically, is to lean in and hustle in a bot-like way. “Rather than being politically radicalized, everyone grinds harder,” Sun said. The reward for the grind might be a role as an overlord of the A.I. future: the closer to collaborating with the machine you are, the more power you will have. Fears of a permanent underclass reflect the fact that there is not yet a coherent vision for how a future A.I.-dominated society will be structured. Sun said, of the Silicon Valley élites pushing accelerationism, “They’re not thinking through the economic implications; no one has a plan for redistribution or Universal Basic Income.” What will be left for the underlings, it seems, is a bleak world of A.I.-generated content and the semblance of companionship from chatbots. As Sun put it, “Are you going to be the piggy or be the one making the slop?”
