Is "jobs" a four-letter word?

The invisible id.

An office, 2026.

When three of us started a startup thirteen years ago, our first office was in a sublet of a sublet. The full office has space for about 70 people; we took a back room off to the side, among a scrum of unused desks and Aeron chairs. It was a good temporary spot, we told ourselves, until we hired more people and moved into something bigger.

Our direct landlords offered us the room because they only had about ten employees themselves. But they wouldn’t be that small for long. They recently raised $15 million, were making bold plans for what to do with it, and had a warning for us: They would be hiring aggressively, and would soon need the room back.

Similarly, the office’s original tenants—the grandparents of our lease—moved out a few months ago, into a new “crib” down the street, so that their company could grow from 68 employees to more than 450. To celebrate the upgrade, their CEO gave a tour of the new office to TechCrunch. The video is an eight-minute time capsule of 2010s’ tech culture: An open bullpen of desks, game rooms, beer fridges, office birthday parties,¹ corporate rap songs, and unburdened optimism about the world-eating potential for enterprise cloud software. At one point in the interview, the CEO looks out at a construction site across the street: “And as we expand and fill up this building,” he says, “maybe we’ll expand over there as well.”

A decade ago, that was what ambition meant in Silicon Valley: Hiring. Companies wanted to fill twinkling offices; CEOs wanted to command bustling swarms of eager employees. “How big is your startup?,” people might ask. You couldn’t tell people how much money you made because most of us didn’t make any, so most of us talked about how much we’d raised and how many people we planned to hire with it. In every compressed social ecosystem, there are always social scoreboards and seating charts. In Silicon Valley, headcount was one of ours.

It was, obviously, a bad measuring stick. It inverted our ends—make good software? Delight customers? Make the world a better place?—with our means; it amplified the boom and bust cycles that technology companies are already prone to follow. And when a founder or CEO flew their legions too close to the sun, it was often the laid-off employees who fell the hardest.

Still, the compulsion to hire wasn’t all bad. For every CEO or VP that wanted to hire an army, an army of people got hired. It was, in an odd and indirect way, distributive and democratic. Yes, the capital C-Capital—the young Stanford founders raising millions, and the venture capitalists who gave it to them—needed the Labor to build their widgets, but they also needed the Labor to impress their friends. They needed the Labor to get invited on the best podcasts. They needed the Labor to fuel the economic machinery of their enterprises, and they needed it for their egos. It was a form of social symbiosis: Even in cases when the invisible hand might want to turn a company into a gutted husk like those that private equity firms often leave behind, the invisible id gives it pause—big layoffs are embarrassing.²

In other words, headcount might be our bleakest scoreboard, except for all the others.

---

Artificial intelligence is remaking the world, but nobody knows how. AI will annihilate millions of jobs, some say; no, it won’t, argue others. It will launch a handful of people into a tiny cloister of obscene wealth and bury everyone else in a permanent underclass. No, that theory is a fallacy, and doomerism is wrong. Recent computer science graduates can’t find jobs and are working at Chipotle; demand for tech talent is rising. AI-native startups are smaller and hire fewer people; or, companies that spend more on AI actually hire more people.³ Oracle, like many other tech firms, says the “adoption and deployment of AI technologies across our operations have resulted, and may continue to result, in reductions to our workforce;” the Wall Street Journal wonders if the cuts are “efforts to correct for past overhiring simply to come into line with industry norms.” Pick your narrative, and you can find evidence for it. The Jevons paradox comes up a lot.

But at least one thing has clearly changed: The scoreboard. Tech startups no longer brag about how many employees that they have. Instead, they now celebrate how few they need.

According to the Wall Street Journal, staying small is the tech industry’s newest flex. Business Insider reports that revenue per employee has become the most popular metric in Silicon Valley; Garry Tan, the CEO of Y Combinator, the powerful startup incubator, says that it’s the most important metric for building a hardcore [laudatory] culture. And engineer who works at Anthropic built a literal scoreboard. I’ve built successful companies “the old fashioned way by hiring hundreds of people,” it says. “I’m now building in AI and many of us in Silicon Valley are now big believers in the idea that [OpenAI CEO] Sam Altman put forward.”

That idea? Two years ago, Altman famously predicted that someone would create a company that was worth a billion dollars with one employee. For many entrepreneurs, it wasn’t seen as a prediction; it was seen as a challenge.

Because that is the new ambition: Not a bustling office, but an empty one. Stripe, the payments processor, reports one-employee businesses are booming; citing that report, Derek Thompson—formerly of The Atlantic, and now, fittingly, writing independently on Substack—declared this as “a golden age for solo acts.” And because this is Silicon Valley, there are also startups building “infrastructure for the one person billion dollar company.”⁴

Even layoffs—long considered the blackest of eyes for an executive—are now, almost, celebrated. In the last six months, companies from Block, Coinbase, and ClickUp all laid off between 14 and 40 percent of their staff. The layoffs were all announced through posts on Twitter, and all followed the same rhythm: Business is booming, the company is healthy, but the world is changing and we need to change faster, and great days are ahead. You could argue that this is also misdirection and marketing—as some have—but that’s partly the point: Layoffs are marketing. And the highest status job in tech is getting rich while running the smallest team.⁵

Of course, this could just be another fad; San Francisco is, after all, the same town that invented Google Glass and NFTs. But it’s also the same place that incubated the internet, social media apps, and the modern office.⁶ What happens in tech sometimes becomes a joke, and sometimes becomes ubiquitous.

And if AI is the next thing that escapes Silicon Valley, is it so hard to imagine a world where hiring is stigmatized? “Oh, you run a company with a 1,000 workers? How 2010s of you.” “You lead a department with a hundred employees? How did you let that happen?” And then, a judgmental look, and a recommendation to listen to some popular podcast: “It’s an interview with the CEO of a company that makes twice as much money, and is a fifth the size.”

Sure, there are reams of research to be done on the economic effects of AI on the corporate world. But what of cultural ones?

1. I’m not sure it’s worth lighting the building on fire for a chocolate sheet cake—chocolate cake, vanilla icing; the worst combination—but for funfetti one? I’m burning it down. (The full album! Out today! Stop reading this! Go! Go celebrate! Go listen! Louder!)

2. You could argue that companies that have been “optimized” by private equity firms are economically ideal, and that anything else is inefficient. Which, maybe, though surely there are degrees. Companies are already overwhelmingly economically motivated, especially in the United States. The social desire to run a big company is probably a minor counterweight. Is that counterweight currently too big? Or, put differently, which world would we rather live in—this one, or one that has, in effect, been even more optimized by private equity?

3. This study produced a buzzy headline—that companies that spend more on AI also hire more—but I’m not at all sure it means what all the press about it said it means. To qualify as a high spender, a company had to spend an average of $34 per person per month on AI stuff, but $34 is the cost of a ChatGPT Plus plan and a basic subscription to Granola. I don’t think people are generally worried about a simple chatbot and a meeting transcription service dramatically altering the labor market; the worry (or hope) is about jobs being replaced by much more complex and expensive tools. Though some companies surely spent more than $34—which is both the average (though, given the likely skew of the distribution, also means a lot of companies probably spent much less) and the initial amount they spent, so spending could’ve gone up over time (or, similarly, down)—that feels like an awfully low threshold from which to draw any meaningful conclusions.

4. Wherever there’s a gold rush, there are always shovel salesmen.

5. Most of the framing here is negative—because bah humbug—but it’s also very possible that all of this is positive. According to a recent survey of thousands of tech workers, people who work at small companies are the happiest and most optimistic. So maybe the shift towards small teams is aligning what is fashionable with what is good.

6. Like bullpens of open desks and game rooms, though someone else is to blame for the corporate rap videos.