Let’s say you’re trying to decide where to order lunch. Once upon a time, you might have had a Zagat guide on the shelf—or depending on your lifestyle, a Michelin. Today, you’re more likely to go online. On a website like Yelp, you can find the same collection of opinions, the same star rating system, and the same index of logistical facts like addresses and phone numbers. But while that information was once collected by a staff of experts, it’s now provided by someone else: you. Whatever else Yelp is, it is an enormous repository of labor.

Over the years, Silicon Valley has scrambled the way we think about work and how it generates value. Companies of all kinds have strained to make their offices resemble those of tech giants—everyone is incubating and disrupting. Companies like WeWork, the latest cautionary tale, hope for a windfall of venture capital by standing close to, and speaking the language of, the tech industry. The standards established in that industry have come to influence jobs that seem to have little to do with the now-deserted campuses along California’s US-101. Instead, they have made us, the users, party to those standards—often just by using their products. Among the most influential standards, though not always in obvious ways, are its conceptions of gender.

The collective labor that goes into something like Yelp is of two kinds: A platform has to have both structure and content, and both have to be produced by labor. But who’s doing the work here? And who’s getting paid? What makes the platform attractive to a general user is above all an unparalleled mass of reviews—useful, funny, cool, or not. Reviews that represent probably billions of labor hours—none of which were remunerated. In fact, the whole thing becomes profitable only if we can explain to ourselves why a certain kind of labor doesn’t deserve or require pay. You might argue that reviewers aren’t employees—and platforms often do—but it’s quite another thing to say that what they do isn’t even work.

A platform like Yelp is unprecedented in many ways, but the way labor is rendered invisible in order for it to function is decidedly not. There are a lot of ways to get something from consumers while avoiding giving out a paycheck. The tech industry has come to call many such tactics “gamification”—collecting data or content by tricking the user into thinking they’re having fun. The way this happens is best described with language drawn from the field of care work, a field in which labor—particularly women’s labor—has historically been made invisible. It’s no accident that Silicon Valley relies on the same gendered rhetoric.

This was something that was tacitly acknowledged within the company, as one early employee told me, speaking under condition of anonymity. He got the sense that the initial marketing strategy was deliberately meant to recruit women as users, by means of gifts and gamified rewards that seemed to cater to stereotypical gender roles. Most of the “elite” events, meant to reward high-volume reviewers at Yelp, were spa events, involved skincare products, or took place at hair salons. The reason Yelp targeted women, the engineer speculated, was “a suspicion that they would have a larger number of online social connections to invite to the site and a better writing pool.”

Small decisions like this along the path to start-up success add up, and they reactivate older tropes and project them into the future. Even the word this early Yelp employee chose—“writing pool”—brings to mind the secretarial pool of the mid-20th century workplace. Women were there for the social stuff, for doing the important but ultimately secondary work of providing content for the structures set up by an—as this same employee makes clear—“overwhelmingly male” central organization. Yelp may have epitomized an unprecedented kind of virtual environment, with hitherto unglimpsed forms of labor, but when early employees began imagining and describing this space, you could hear the faint echoes of the clatter of typewriters and the workplace mores of Sterling Cooper. Although more than a half-century old by now, gendered categories like this still influence who gets hired in Silicon Valley, and how their work is understood once they are.

It has hardly escaped notice that the tech industry isn’t exactly exemplary when it comes to gender representation. What has received less attention is the implicit gendering of labor and work product that perpetuates that problem. The Women’s Leadership Lab at Stanford is a research institute trying to untangle what goes wrong around gender and race in Silicon Valley, whether it’s about recruitment, promotion, or retention. What the researchers are finding is that, for all their meritocratic myths and their famously flat organizational structures, Silicon Valley firms are rife with inequalities that ultimately sort employees into those whose work counts more and those whose work counts less.

Tech has developed all kinds of interlocking hierarchies to differentiate labor and hierarchize how it is rewarded: management vs. technical ladder, fuzziness vs. logic, front end vs. back end, engineering vs. design. And then there is the byzantine nomenclature of badges at most large tech companies—badges that distinguish the full-timer from the contractual worker, the worker who deals with the product and the one who runs some kind of support network. The system makes public how an individual worker is salaried by what company, and also what perks one has access to—no shuttle for the blue badge, no all-hands meeting unless you have a white one.

There is an entire value system being telegraphed to workers at these companies, and the platforms they set up often enough pass on that value system to the users. There is work that is the core of the mission; it is arduous, difficult, all-consuming, and isolating. And it is almost invariably figured as masculine. As Miriam Posner, a professor at UCLA, has pointed out, this ideology is “why the problem of women in technology is thornier than shoehorning women into all-male panels.” It’s not just the gender of the employee; it’s the way gender is embedded in the job.

As Sharla Alegria from the University of Toronto has found, gender in tech has often been governed by a set of interlocking, and surprisingly fine-grained, sorting systems. She found that women were less likely to be promoted than men, and when white women were promoted, it was frequently into jobs that seemed to separate them further from the core business of the company, and toward “soft” and “people skills.” This was even more pronounced for nonwhite women, who ended up in fields even more removed from their training. She interviewed women of color who underwent training after training in programming, but ended up in “support” or “help desk” positions.

Within this system, people can make choices, but they are, as Shannon Gilmartin of the VMware Women’s Leadership Innovation Lab told me, “loaded choices.” If the corporate culture recognizes certain jobs as “male” (having to do with the platform, the product, the genius vision) and “female” ones (care, content, and communication), then it won’t take long for those who try to buck the pattern to be treated as interlopers.

Of course, some of these distinctions between “masculinized” and “feminized” jobs reproduce gendered valuations that one finds across American society, from the elementary school classrooms to the New York Stock Exchange. As Posner points out, one way of telling the story of the rise of the tech industry is of a field that had been heavily feminized, dominated by clerical busywork, and accordingly had seen low wages, establishing itself as a domain of male geniuses that command exorbitant salaries. In that way, tech is just part of the jagged intersection of gender and the American workplace more broadly.

But other distinctions between “male” and “female” work in tech are granular enough that they can surprise outsiders when they come across them. Then again, more and more they might not. Because while the obsession with hard logic, purity, and difficulty that underlies them may be somewhat unique to the tech industry, this obsession has long shaped what outsiders picture when we hear words like “tech worker” or “programmer.”

In accounts like Posner’s, or Mar Hicks’s study of the British tech industry, it becomes clear that this development was far from inevitable. As the tech sector took shape, people had to choose to interpret labor and value. Where did this interpretation come from, and how has it proved so resilient? One place to look is the intellectual climate in which the companies—not necessarily the technologies—of Silicon Valley first took shape. The ideas of R. Buckminster Fuller (1895–1983), Marshall McLuhan (1911–1980), and Ayn Rand (1905–1982) came to Silicon Valley via the midcentury counterculture, and have always retained its trappings.

Think of the way the concept of “genius” has shaped the image of the tech industry. “Genius” is an old word, but one that means fairly specific things in the tech context. The tech industry likes to diffuse the concept—think of the “genius bar” at the Apple Store, of the idea of the collective genius of crowd intelligence. But when it comes to creating the companies and platforms, the word seems to refer more to—as the subtitle of Adam Fisher’s recent Uncensored History of Silicon Valley puts it—The Hackers, Founders, and Freaks Who Made It Boom. The genius is a hacker, a founder, a freak. Pure and aloof, the genius is disruptive to the old order and ultimately a little cold. The genius isn’t a team player, often an outsider, does not play well with others. To ask him to play with others means diminishing his brilliance. And the genius is usually a “he.”

The “genius aesthetic,” which holds that the meaning of an object is best understood by understanding the mind of its exceptional creator, comes from the turn of the 19th century. But it is everywhere in contemporary Silicon Valley, which often treats multibillion-dollar companies with thousands of employees as though they were one brilliant individual’s work of art. Elon Musk unites a strange bundle of companies just by the magnetic force of his outsize personalities. Steve Jobs’s personal quirks are embedded in the gadgets that have made him an icon. This way of reducing collective labor to the ideas of one person is likely owed to the influence of Ayn Rand, a favorite of many tech moguls, as is the idea that the job of the broader collective is to accommodate itself to the brilliance of the few.

Marshall McLuhan has had a more counterintuitive impact on the lay theories of Silicon Valley. Sometime in the 1940s, McLuhan, then teaching English at Saint Louis University, began filling grocery boxes with various advertisements from newspapers. His first book, The Mechanical Bride, began its life as essentially one of these boxes sent to Vanguard Press in New York, along with McLuhan’s commentary. The medium (the box of pop cultural flotsam) was itself the message.

What made McLuhan’s work pathbreaking was this ability to stand over a dizzying cacophony of content and establish formal relations between the disparate elements. Focusing on the content of media messages was ultimately a fool’s errand. “The ‘content’ of a medium is like the juicy piece of meat carried by the burglar to distract the watchdog of the mind.” McLuhan framed content as seductive: Obvious, diverting, and uncomplicated, it thrust itself upon the analyst and kept him from the serious work of understanding what was really going on.

McLuhan’s legacies to the counterculture and later to the tech industry are various. But one of the legacies was simply about the kind of persona he modeled. The curious, aloof, slightly ironic man who takes seriously the detritus of our expanding communication networks and figures out how to bring it into some sort of order. There is one type of labor, in other words, that fills magazines with ads, comics, advice columns; there’s another type of labor that clips those out, collates them, and orders them in ways their creators only dimly intuit. And the second type of labor was more real, more important, more valuable.

Whatever else they contributed to Silicon Valley’s emerging ecosystem, figures like McLuhan, along with R. Buckminster Fuller or Stewart Brand, shaped a particular style of Silicon Valley masculinity: techno-libertarian, opposed to traditional structures, and defined by a freewheeling and highly individuated creativity. And that valued free, abstract tinkering, removed as much as possible from the messy political and interpersonal concerns of modern society. That sort of gee-whiz purity always has an implicit opposite—the thing to keep tinkering pure from. That which lies outside the monastery walls.

When it comes to who gets promoted how in Silicon Valley today, Gilmartin told me, there’s still an assumption that “anybody can be a manager, but not everybody can be an engineer.” Dealing with the interpersonal stuff, with care work, with the content, according to this ideology, requires little specialization. As a general rule, the further you are removed from the “front end,” your colleagues, the customer and the content, the more prestige accrues to your position. This applies to female tech workers who get frequently shunted onto management tracks, into UX design or even into HR, even though they have the same degrees and training as their male engineer colleagues. It gets worse once you get beyond them.

In the far reaches of the ecosystem of a tech company—beyond the limits of those the companies even recognize as “their” employees.—the imbalance applies even more forcefully. This is tech’s “shadow workforce,” as the journalist Ellen Sheng puts it: contractors who supplement the efforts of direct employees of the tech giants. As Wendy Liu has written, the contract worker “becomes an atomized unit, forced to display enough perseverance and ‘merit’ to convince their supervisor of their economic value,” in hopes of being made permanent. Conversations with contract workers suggest that the same implicit biases around labor that are rampant in these companies govern who gets hired for these positions, how much they make when they get there, what benefits and privileges their non-employer feels necessary to grant them, and whether they are ever made permanent employees.

Whether you are recognized as an employee or not, based on the mere technicality that you work for a certain company, whether you are recognized as being “core” to the mission or just have a supporting role, whether and how much of your labor is being recognized as labor at all: all of that turns out to be governed by the same logic.

This emphasis on imposing form has transmitted itself to modern Silicon Valley as a prioritization of the platform over the content. But it has shifted from a methodological concern (how best to analyze) to one of value generation. It has remade these workplaces, both shaping and making invisible the immense inequality that makes them tick. What has remained are the gender politics behind that idea. And however far Silicon Valley has traveled from the heady days of McLuhan, Fuller, and Brand, these ideas have come to govern how Silicon Valley organizes its companies—from the largest corporations to the scrappiest startup. As any startup founder would insist, in turn those ideas have remade our world.