It smelled of burnt coffee and old carpet; rain tapped the window, and a single coffee ring stained a notebook beside a stack of recruiter emails. The office felt smaller than the job posting still open on the screen.
Small signs tell you a lot. You notice the quiet first — fewer people, fewer offers — then the rest follows.
What “enshittification” looks like at work
There’s a tidy logic behind a phrase Cory Doctorow has been using: platforms or industries start by courting workers or users, then squeeze them once they control the market. He’s turned that lens on tech employment, arguing that the very bargaining power engineers once enjoyed has been hollowed out as firms weaponize automation and layoffs to justify lower pay and longer hours. The piece lays out a stark arc: scarcity of talent created leverage; now AI, mass cuts and brittle hiring markets are eroding it. (doctorow.medium.com)
That’s not just rhetoric. Big corporate moves are reshaping whole labor markets. Reuters flagged dramatic cuts at major outsourcing houses this week — Tata Consultancy Services announced its largest reductions ever, a move many industry observers linked to companies pushing AI-driven efficiency. The story is not limited to one company or one country; it’s part of a broader rearrangement in which firms claim AI does the work formerly done by teams of humans. (reuters.com)
Workers are uneasy. Pew Research polling from early 2025 finds most employees don’t expect AI to open new opportunities for them; many think it will reduce job prospects. In short: confidence is low, and anxiety is high — even in sectors that once felt insulated. (pewresearch.org)
What’s changing, and why it matters
For two decades, the tech labor market was a seller’s market. Recruiters chased engineers by the dozen; stock grants and glitzy campuses (kombucha on tap, laundry service) were the rule. That abundance created a practical, if informal, leverage that boosted wages and opened career ladders.
Now the leverage is evaporating. Companies point to models that can write boilerplate code, triage issues or synthesize reports, then cut teams and say they’ll rely on AI to keep pace. The math is seductive for executives: fewer people, similar outputs, higher margins. The reality for employees is different. Workloads compress. Career-step roles — junior jobs that used to be on-ramps — are the first to disappear, choking the ladder for entry-level hires.
Aisha Patel, 33, a former senior engineer at a mid-sized payments startup, picks at a loose thread on her sleeve as she talks. “You get this text from HR that a job’s been ‘restructured,’ and you’re like — okay — but then they post the same role at lower comp. I mean, it’s brutal,” she says. “I had a mentor who used to walk in with a giant thermos of tea and a worn golf glove on the desk — little rituals, y’know? They’re gone now. And yeah, AI’s real, but it’s being used to set expectations. They’re negotiating with a machine and we get the bill.”
Mark Delaney, 47, a Silicon Valley recruiter who’s worked both in-house and for boutique headhunters, taps his fingers on a notebook with coffee rings. “Clients will say, ‘We have this model that can do 60% of the work,’” he says with a weary laugh. “So they feel comfortable offering lower salary bands. And, uh, candidates get scared. You can’t bargain when the other side believes they’ve already got a bot.” His voice softens. “It’s like watching the slow deflation of a balloon.”
Wider implications: not all doom, but real friction
There are contradictory trends. On one hand, AI tools are producing new product categories, and some firms are hiring for roles tied to model governance, data labeling or safety. On the other, many of the steady, time-tested entry and middle roles — the ones that let people build careers — are shrinking. Sources remain conflicted about net job creation versus displacement, and the reality is likely more complicated than simple “AI will kill jobs” headlines suggest.
This matters politically and socially. If the tech sector stops functioning as a ladder — if the baseline for an early-career software job becomes a precarious contractor position or is replaced by an AI-assisted hire — then inequality widens. Universities and policymakers will squabble over training and safety nets, but businesses are already reshaping the rules of workplace bargaining without much public debate.
How companies use AI in negotiation and culture
It’s not only about automation. The mere presence of an AI claim becomes a bargaining chip. Firms will tout “AI augmentation” in earnings calls, then trim budgets for raises, or ramp up performance metrics while insisting productivity is higher. Executives frame longer hours as maximizing “innovation windows” in a race that — conveniently — justifies fewer hires.
There’s a cultural shift too: the language of “productivity” replaces the language of stewardship. Where once leaders might have defended headcount for strategic reasons, many now default to forecasts of model-driven gains. That shift affects morale, retention and the incentive to mentor newcomers — a quiet erosion of institutional knowledge.
What can be done — and what I’ve seen
Organizing is one obvious route. Doctorow argues that scarcity alone won’t save workers anymore; collective bargaining may be the only durable counterbalance. The idea has traction: a handful of tech unions and worker groups have been testing that model, with mixed success.
From my own reporting and past newsroom days (I remember covering labor beats when dial-up was still a novelty and Seinfeld’s reruns felt edgy), the texture of change matters. Small, prosaic things — a scrappy coffee station, a whiteboard with someone’s half-drawn idea — often signal a workplace’s health. When those things vanish, the loss is practical and cultural.
An odd aside: at a recent meetup I attended, someone passed around a battered Game Boy as a joke about “legacy tech.” It was funny, then quietly sad — nostalgia for earlier eras when tech felt a little more playful, a little less transactional.
Where uncertainty lingers
There are unanswered questions. Will regulation step in to mandate hiring or retraining? Will firms that over-rely on AI find innovation suffers because human, cross-disciplinary conversation shrinks? The evidence on long-term productivity gains from AI is mixed, and investors’ short-term desires can push companies toward leaner staffing even when that’s not optimal for product quality.
Practical advice for readers
If you’re in tech or hiring into it, think about the ladder. Look for companies that still staff intentionally — that keep junior roles and mentorship on the books. If you’re a policymaker or community leader, consider creating real, funded pathways for mid-career reskilling rather than hoping market forces alone will sort things out. Pew’s polling suggests workers aren’t optimistic; that’s not a number you can wish away. (pewresearch.org)
Final thought — and a personal note
Call me old-fashioned, but I prefer a market where bargaining happens between people, not between people and marketed algorithms. The phrase Doctorow uses is blunt, but useful: when platforms or employers start extracting value through engineered scarcity and opacity, the job of journalism is to name it, examine it, and push for accountability. I’ve been covering this beat long enough to have seen cycles repeat. The tools change. The stakes do, too. We shouldn’t pretend the machines are neutral referees in what’s effectively a human negotiation.
Sources cited in the reporting: a recent essay by Cory Doctorow on shifts in tech labor markets; Reuters’ reporting on large outsourcing layoffs tied to AI shifts; and Pew Research Center polling on workers’ views of AI in the workplace. (doctorow.medium.com, reuters.com, pewresearch.org)
About the Author
