The moral injury of having your work enshittified
If you’d like an essay-formatted version of this post to read or share, here’s a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
This week, I wrote about how the Great Enshittening — in which all the digital services we rely on become unusable, extractive piles of shit — did not result from the decay of the morals of tech company leadership, but rather, from the collapse of the forces that discipline corporate wrongdoing:
The failure to enforce competition law allowed a few companies to buy out their rivals, or sell goods below cost until their rivals collapsed, or bribe key parts of their supply chain not to allow rivals to participate:
The resulting concentration of the tech sector meant that the surviving firms were stupendously wealthy, and cozy enough that they could agree on a common legislative agenda. That regulatory capture has allowed tech companies to violate labor, privacy and consumer protection laws by arguing that the law doesn’t apply when you use an app to violate it:
But the regulatory capture isn’t just about preventing regulation: it’s also about creating regulation — laws that make it illegal to reverse-engineer, scrape, and otherwise mod, hack or reconfigure existing services to claw back value that has been taken away from users and business customers. This gives rise to Jay Freeman’s perfectly named doctrine of “felony contempt of business-model,” in which it is illegal to use your own property in ways that anger the shareholders of the company that sold it to you:
Undisciplined by the threat of competition, regulation, or unilateral modification by users, companies are free to enshittify their products. But what does that actually look like? I say that enshittification is always precipitated by a lost argument.
It starts when someone around a board-room table proposes doing something that’s bad for users but good for the company. If the company faces the discipline of competition, regulation or self-help measures, then the workers who are disgusted by this course of action can say, “I think doing this would be gross, and what’s more, it’s going to make the company poorer,” and so they win the argument.
But when you take away that discipline, the argument gets reduced to, “Don’t do this because it would make me ashamed to work here, even though it will make the company richer.” Money talks, bullshit walks. Let the enshittification begin!
But why do workers care at all? That’s where phrases like “don’t be evil” come into the picture. Until very recently, tech workers participated in one of history’s tightest labor markets, in which multiple companies with gigantic war-chests bid on their labor. Even low-level employees routinely fielded calls from recruiters who dangled offers of higher salaries and larger stock grants if they would jump ship for a company’s rival.
Employers built “campuses” filled with lavish perks: massages, sports facilities, daycare, gourmet cafeterias. They offered workers generous benefit packages, including exotic health benefits like having your eggs frozen so you could delay fertility while offsetting the risks normally associated with conceiving at a later age.
But all of this was a transparent ruse: the business-case for free meals, gyms, dry-cleaning, catering and massages was to keep workers at their laptops for 10, 12, or even 16 hours per day. That egg-freezing perk wasn’t about helping workers plan their families: it was about thumbing the scales in favor of working through your entire twenties and thirties without taking any parental leave.
In other words, tech employers valued their employees as a means to an end: they wanted to get the best geeks on the payroll and then work them like government mules. The perks and pay weren’t the result of comradeship between management and labor: they were the result of the discipline of competition for labor.
This wasn’t really a secret, of course. Big Tech workers are split into two camps: blue badges (salaried employees) and green badges (contractors). Whenever there is a slack labor market for a specific job or skill, it is converted from a blue badge job to a green badge job. Green badges don’t get the food or the massages or the kombucha. They don’t get stock or daycare. They don’t get to freeze their eggs. They also work long hours, but they are incentivized by the fear of poverty.
Tech giants went to great lengths to shield blue badges from green badges — at some Google campuses, these workforces actually used different entrances and worked in different facilities or on different floors. Sometimes, green badge working hours would be staggered so that the armies of ragged clickworkers would not be lined up to badge in when their social betters swanned off the luxury bus and into their airy adult kindergartens.
But Big Tech worked hard to convince those blue badges that they were truly valued. Companies hosted regular town halls where employees could ask impertinent questions of their CEOs. They maintained freewheeling internal social media sites where techies could rail against corporate foolishness and make Dilbert references.
And they came up with mottoes.
Apple told its employees it was a sound environmental steward that cared about privacy. Apple also deliberately turned old devices into e-waste by shredding them to ensure that they wouldn’t be repaired and compete with new devices:
And even as they were blocking Facebook’s surveillance tools, they quietly built their own nonconsensual mass surveillance program and lied to customers about it:
Facebook told employees they were on a “mission to connect every person in the world,” but instead deliberately sowed discontent among its users and trapped them in silos that meant that anyone who left Facebook lost all their friends:
And Google promised its employees that they would not “be evil” if they worked at Google. For many googlers, that mattered. They wanted to do something good with their lives, and they had a choice about who they would work for. What’s more, they did make things that were good. At their high points, Google Maps, Google Mail, and of course, Google Search were incredible.
My own life was totally transformed by Maps: I have very poor spatial sense, need to actually stop and think to tell my right from my left, and I spent more of my life at least a little lost and often very lost. Google Maps is the cognitive prosthesis I needed to become someone who can go anywhere. I’m profoundly grateful to the people who built that service.
There’s a name for phenomenon in which you care so much about your job that you endure poor conditions and abuse: it’s called “vocational awe,” as coined by Fobazi Ettarh:
Ettarh uses the term to apply to traditionally low-waged workers like librarians, teachers and nurses. In our book Chokepoint Capitalism, Rebecca Giblin and I talked about how it applies to artists and other creative workers, too:
But vocational awe is also omnipresent in tech. The grandiose claims to be on a mission to make the world a better place are not just puffery — they’re a vital means of motivating workers who can easily quit their jobs and find a new one to put in 16-hour days. The massages and kombucha and egg-freezing are not framed as perks, but as logistical supports, provided so that techies on an important mission can pursue a shared social goal without being distracted by their balky, inconvenient meatsuits.
Steve Jobs was a master of instilling vocational awe. He was full of aphorisms like “we’re here to make a dent in the universe, otherwise why even be here?” Or his infamous line to John Sculley, whom he lured away from Pepsi: “Do you want to sell sugar water for the rest of your life or come with me and change the world?”
Vocational awe cuts both ways. If your workforce actually believes in all that high-minded stuff, if they actually sacrifice their health, family lives and self-care to further the mission, they will defend it. That brings me back to enshittification, and the argument: “If we do this bad thing to the product I work on, it will make me hate myself.”
The decline in market discipline for large tech companies has been accompanied by a decline in labor discipline, as the market for technical work grew less and less competitive. Since the dotcom collapse, the ability of tech giants to starve new entrants of market oxygen has shrunk techies’ dreams.
Tech workers once dreamed of working for a big, unwieldy firm for a few years before setting out on their own to topple it with a startup. Then, the dream shrank: work for that big, clumsy firm for a few years, then do a fake startup that makes a fake product that is acquihired by your old employer, as an incredibly inefficient and roundabout way to get a raise and a bonus.
Then the dream shrank again: work for a big, ugly firm for life, but get those perks, the massages and the kombucha and the stock options and the gourmet cafeteria and the egg-freezing. Then it shrank again: work for Google for a while, but then get laid off along with 12,000 co-workers, just months after the company does a stock buyback that would cover all those salaries for the next 27 years:
Tech workers’ power was fundamentally individual. In a tight labor market, tech workers could personally stand up to their bosses. They got “workplace democracy” by mouthing off at town hall meetings. They didn’t have a union, and they thought they didn’t need one. Of course, they did need one, because there were limits to individual power, even for the most in-demand workers, especially when it came to ghastly, long-running sexual abuse from high-ranking executives:
Today, atomized tech workers who are ordered to enshittify the products they take pride in are losing the argument. Workers who put in long hours, missed funerals and school plays and little league games and anniversaries and family vacations are being ordered to flush that sacrifice down the toilet to grind out a few basis points towards a KPI.
It’s a form of moral injury, and it’s palpable in the first-person accounts of former workers who’ve exited these large firms or the entire field. The viral “Reflecting on 18 years at Google,” written by Ian Hixie, vibrates with it:
Hixie describes the sense of mission he brought to his job, the workplace democracy he experienced as employees’ views were both solicited and heeded. He describes the positive contributions he was able to make to a commons of technical standards that rippled out beyond Google — and then, he says, “Google’s culture eroded”:
Decisions went from being made for the benefit of users, to the benefit of Google, to the benefit of whoever was making the decision.
In other words, techies started losing the argument. Layoffs weakened worker power — not just to defend their own interest, but to defend the users interests. Worker power is always about more than workers — think of how the 2019 LA teachers’ strike won greenspace for every school, a ban on immigration sweeps of students’ parents at the school gates and other community benefits:
Hixie attributes the changes to a change in leadership, but I respectfully disagree. Hixie points to the original shareholder letter from the Google founders, in which they informed investors contemplating their IPO that they were retaining a controlling interest in the company’s governance so that they could ignore their shareholders’ priorities in favor of a vision of Google as a positive force in the world:
Hixie says that the leadership that succeeded the founders lost sight of this vision — but the whole point of that letter is that the founders never fully ceded control to subsequent executive teams. Yes, those executive teams were accountable to the shareholders, but the largest block of voting shares were retained by the founders.
I don’t think the enshittification of Google was due to a change in leadership — I think it was due to a change in discipline, the discipline imposed by competition, regulation and the threat of self-help measures. Take ads: when Google had to contend with one-click adblocker installation, it had to constantly balance the risk of making users so fed up that they googled “how do I block ads?” and then never saw another ad ever again.
But once Google seized the majority of the mobile market, it was able to funnel users into apps, and reverse-engineering an app is a felony (felony contempt of business-model) under Section 1201 of the Digital Millennium Copyright Act. An app is just a web-page wrapped in enough IP to make it a crime to install an ad-blocker.
And as Google acquired control over the browser market, it was likewise able to reduce the self-help measures available to browser users who found ads sufficiently obnoxious to trigger googling “how do I block ads?” The apotheosis of this is the yearslong campaign to block adblockers in Chrome, which the company has sworn it will finally do this coming June:
My contention here is not that Google’s enshittification was precipitated by a change in personnel via the promotion of managers who have shitty ideas. Google’s enshittification was precipitated by a change in discipline, as the negative consequences of heeding those shitty ideas were abolished thanks to monopoly.
This is bad news for people like me, who rely on services like Google Maps as cognitive prostheses. Elizabeth Laraki, one of the original Google Maps designers, has published a scorching critique of the latest GMaps design:
Laraki calls out numerous enshittificatory design-choices that have left Maps screens covered in “crud” — multiple revenue-maximizing elements that come at the expense of usability, shifting value from users to Google.
What Laraki doesn’t say is that these UI elements are auctioned off to merchants, which means that the business that gives Google the most money gets the greatest prominence in Maps, even if it’s not the best merchant. That’s a recurring motif in enshittified tech platforms, most notoriously Amazon, which makes $31b/year auctioning off top search placement to companies whose products aren’t relevant enough to your query to command that position on their own:
Enshittification begets enshittification. To succeed on Amazon, you must divert funds from product quality to auction placement, which means that the top results are the worst products:
The exception is searches for Apple products: Apple and Amazon have a cozy arrangement that means that searches for Apple products are a timewarp back to the pre-enshittification Amazon, when the company worried enough about losing your business to heed the employees who objected to sacrificing search quality as part of a merchant extortion racket:
Not every tech worker is a tech bro, in other words. Many workers care deeply about making your life better. But the microeconomics of the boardroom in a monopolized tech sector rewards the worst people and continuously promotes them. Forget the Peter Principle: tech is ruled by the Sam Principle.
As OpenAI went through four CEOs in a single week, lots of commentators remarked on Sam Altman’s rise and fall and rise, but I only found one commentator who really had Altman’s number. Writing in Today in Tabs, Rusty Foster nailed Altman to the wall:
Altman’s history goes like this: first, he founded a useless startup that raised $30m, only to be acquired and shuttered. Then Altman got a job running Y Combinator, where he somehow failed at taking huge tranches of equity from “every Stanford dropout with an idea for software to replace something Mommy used to do.” After that, he founded OpenAI, a company that he claims to believe presents an existential risk to the entire human risk — which he structured so incompetently that he was then forced out of it.
His reward for this string of farcical, mounting failures? He was put back in charge of the company he mis-structured despite his claimed belief that it will destroy the human race if not properly managed.
Altman’s been around for a long time. He founded his startup in 2005. There’ve always been Sams — of both the Bankman-Fried varietal and the Altman genus — in tech. But they didn’t get to run amok. They were disciplined by their competitors, regulators, users and workers. The collapse of competition led to an across-the-board collapse in all of those forms of discipline, revealing the executives for the mediocre sociopaths they always were, and exposing tech workers’ vocational awe for the shabby trick it was from the start.