For a hot minute in 1995, coders looked like they might follow in the steps of the Neo-expressionists who taxied north from industrial lofts south of Houston to hang in the halls of power. Hackers, featuring Johnny Lee Miller as Zero Cool and Angelina Jolie as Acid Burn, promised a shadow aristocracy of Galliano-loving Linux fetishists demanding deference if not outright submission. But that’s not what we got.
We got CSV skins, dogshit search engines, backfiring misinformation engines, and Burning Man camps full of palefaces who didn’t know how to handle their drugs. Coders did not produce a culture that commanded mass respect. They did, however, produce a big bad more intimidating than Fisher Stevens on a skateboard – an autocomplete effective enough to take their jobs.
A little over 180 days ago when Anthropic CEO Dario Amodei predicted that AI would “write 90 percent of the code” in six months, Discord servers lit up with furious comments by users with names like “rootkit420” and “TorvaldFucks,” but the usual pearl clutchers didn’t reach for their mollusk secretions. Even as job listings started to disappear and Google, Amazon, and Meta announced mass layoffs, the end of coding as we know it didn’t elicit much of a response from cultural elites or non-tech professionals.
It’s no mystery why. There has always been the sense that coders didn’t belong in the professional class. That they were just a bunch of Palo Alto Hillbillies who struck it rich by living on the right part of the spectrum at the right time2. Few said it aloud, but the current semi-silence speaks semi-volumes.

For a long time, coders seemed like they would become, if not a shadow aristocracy, a permanent part of the professional elite. They were in demand – ascendant enough that “maybe I’ll learn to code” became a creative class euphemism for professional capitulation. But most creatives didn’t learn to code. The job didn’t map to their priorities – coders accrued capital without accumulating power or prestige. There were a few reasons for this. The first was that it was fundamentally a skills-based trade. Like carpentry or plumbing, coding focused on less prestigious “object” work rather than more prestigious “people” work. The second was that, unlike professions built around formal certification processes, coding had innumerable on-ramps – everything from elite computer-science programs to state schools to bootcamps to watching YouTube in the basement – and no clear code of ethics. This may explain the lack of concern about the destruction of coding jobs within the professional–managerial class, many of whom see those particular layoffs as unfortunate but not unjust.
Self-described disruptors can’t much complain about being disrupted.
It is uncomfortably, plausible that AI could replace a great deal of coding labor. That's not because coding is not a creative endeavor, but because most coding is not a creative endeavor. Predictive models can generate, refactor, debug, and document – work that is either repetitive or iterative. Over the past two years, jobs listings for “computer programmers” (typically junior, task-oriented roles) have fallen by nearly 30 percent. Listings for “software developers (typically more senior roles involving systems thinking) have declined more modestly. In other words, the ladder is breaking from the bottom and no one is coming to fix it.
It’s not as though this was impossible to see coming. In 1966, social theorist Gerhard Lenski wrote that “distributive systems which allocate rewards in ways that cannot be adequately justified are inherently unstable.” In other words, high pay without widely recognized civic utility does not generate the moral authority necessary to command empathy. Put differently, non-coders have no reasons to care more about coders than code. “Prestige is accorded on the basis of social evaluation, not simply on the basis of economic success.”
Coders did not pass their social evaluation3.

This failure is obvious not only because of the schadenfreude expressed by members a creative class, which has been destabilized by the diversion of ad dollars toward platforms. Many liberal professionals outside the tech industry view techies as politically suspect, seeing them as ambivalent about labor, regulation, and even democracy. Because of the behavior of handful of tax-avoidant edgelords constantly hurling libertarian-adjacent political kinderzeug, coders are often assumed to lean conservative. In reality, voting patterns in Silicon Valley are no different from those of marketing executives on the Upper West Side. Still, the latter group is suspicious of the former because they never darken the door at Bergdorf’s – and because it’s not entirely clear they’re a group at all.
The British sociologist Frank Parkin had a blunt term for the reason coding professionals never quite managed to become a respectable cohort. “Social closure,” he wrote, “is the attempt by one group to secure advantages by closing off opportunities to another group.” Law and medicine don’t just require skill; they require degrees, years of training, licenses, a narrative of public service, and the polish of people worn smooth by a steady stream of prior vettings. Coding hasn’t built anything comparable. It has shibboleths, sure, but no chokepoints on the way in. No way to keep out talented weirdos.
That openness helped the tech industry hyperscale, creating enormous wealth, but not authority and certainly not moral authority. Sure, tech flirted with effective altruism as a moral justification for monopolistic practices and individual greed, but that ethos was never formalized. Without a clear claim to social value backed by a license, an oath, or a shared ethic of restraint, status remained contestable even with tech firms.
Contestable status is at the core of the decades-long cold war persists between rank-and-file coders and better educated, more socially functional product managers. This conflict was always inevitable. Work organized around people reliably commands more prestige than work organized around things, even when the thing-work is harder, more technical, and better paid. Eliot Freidson, in his classic work on professions, argued that authority and status flow not from technical competence alone but from jurisdiction –from deciding what should be done. Andrew Abbott built on this in The System of Professions, showing that occupations organized around interpreting human needs enjoy greater standing than those defined by mastery of processes.
Coders may have mastery, but, as product-management guru Lenny Rachitsky puts it, a PM’s job is to “identify and solve the most impactful customer problems.” The authority lies upstream, in judgment and prioritization, not execution.

In part because of this, coders tend to seek status primarily within the coder community itself. Languages, stacks, frameworks, LeetCode scores, arcane performance metrics—these create dense internal hierarchies that feel meaningful to those on the inside and opaque to the point of irrelevance to everyone else. The prestige doesn’t travel.
Disrespect does.
The number of CS degrees awarded by trend-chasing American colleges has doubled in the past decade as universities and liberal arts colleges built departments that often felt like trade schools gussied up with copies of Gödel, Escher, Bach4. It's certainly not the first instance of administrators chasing the job market – the law degree boom of the 1990s springs to mind – but it may be the most extreme and embarrassing. Colleges went out of their way to train students to work for employers who actively derided undergraduate education as a bad investment.
Recent data from the Federal Reserve Bank of New York puts unemployment among new computer-science graduates above six percent – the highest of any major. Jobs aren’t just disappearing because of AI. They are being offshored to Pakistan and the Philippines. They are being modularized, routinized, and quietly converted into contract and gig work. There are some folks with very strong resumes reduced to selling ass their services on UpWork.
It’s not Zero Cool. It’s zero cool. Maybe they’ll learn to plumb.

