On May 21, 2026, Zeb Evans did something that CEOs usually announce with grim faces and vague language about "restructuring." The CEO of ClickUp, a collaboration software startup last valued at $4 billion, announced on X that he had just laid off 22% of his workforce. But the tone was not apologetic. It was triumphant.
"Most savings from this change will flow directly back into the people who stay. We'll be introducing million-dollar salary bands," Evans wrote. "If you create outsized impact using AI, you'll be paid outside of traditional bands."
The message was clear: this was not a layoff. This was an upgrade.
The 3,000 AI Agents Taking Over ClickUp
According to Fortune, ClickUp has deployed roughly 3,000 internal AI agents to handle complex tasks across the company. Instead of writing code, designing interfaces, or drafting documentation themselves, employees now direct these agents and review their output. The humans have become managers of a digital workforce that never sleeps, never complains, and never asks for a raise.
Evans's stated goal is to transform ClickUp into what he calls a "100x org" — a company that produces 100 times the output per employee. It's a seductive vision for investors and a terrifying one for workers. And it's becoming less theoretical by the day.
"The people that automate their jobs with AI will always have a job," Evans claimed. But the corollary is implicit and brutal: those who fail to automate their functions will be eliminated. Not demoted. Not reassigned. Eliminated.
The Gartner Data Nobody Wants to Talk About
ClickUp's announcement lands in the middle of a broader trend that should worry anyone who earns a living with their mind rather than their hands. A recent Gartner survey found that approximately 80% of companies using autonomous AI technology have cut jobs. But here's the twist: those workforce reductions are not delivering the financial returns executives promised.
"While Gartner's findings suggest some companies use unproven AI as an excuse to downsize, ClickUp maintains it is not one of them," TechCrunch reported. Evans told the publication that ClickUp is indeed seeing productivity gains from its AI agents and is measuring those efficiencies internally. The company is apparently preparing to productize its agent-management approach for customers — turning its internal experiment into a commercial platform.
The phrase "gamifying value created and time saved" appears in Evans's communications. It replaces the older metric of "tokenmaxxing" — monitoring how many AI tokens employees consume — which critics like Reid Hoffman have argued is the wrong measure because it simply racks up AI expenses without guaranteeing results.
The Polsia Precedent: One Human, $250 Million
If ClickUp's vision sounds extreme, consider Polsia. The one-year-old startup claims to handle all software operations for solopreneurs. It has exactly one employee: founder and CEO Ben Broca. That one-person team just raised $30 million at a $250 million valuation.
Polsia is not a rounding error or a novelty. It is a proof of concept. If one person with AI agents can build and run a software company that investors value at a quarter-billion dollars, the traditional organizational chart starts to look like a historical artifact rather than a business necessity.
The implications cascade. If a Series A startup can operate with one person, how many people does a Series C startup need? If a growth-stage company can function with 50 employees instead of 500, what happens to the other 450? And if public companies adopt the same model, the job losses will make the dot-com bust look like a mild correction.
What "Million-Dollar Salary Bands" Actually Means
Evans's promise of million-dollar salary bands for AI-augmented workers deserves scrutiny. It is simultaneously generous and grotesque. Generous because it offers a path to extraordinary wealth for those who master AI tools. Grotesque because it accepts as given that most of the workforce will be discarded, and the survivors will be paid extravagantly to compensate for the guilt and workload of managing machines that replaced their colleagues.
This is not a new economic model. It is the Hollywood model applied to software. A blockbuster film employs a handful of stars who earn millions, a slightly larger crew that earns decent wages, and thousands of aspiring workers who never get hired. ClickUp is proposing the same structure for tech: AI agents as the crew, a small number of AI whisperers as the stars, and everyone else as unemployed spectators.
The mathematics are compelling for the survivors. If a company produces the same output with 20% of the staff, it can theoretically pay each remaining employee five times as much. But the psychological cost of knowing your paycheck was funded by your friends' termination letters is not factored into the spreadsheet.
The Industry Context: Not Just ClickUp
ClickUp is not operating in a vacuum. The past 18 months have seen a wave of AI-driven or AI-justified layoffs across the tech sector.
Cloudflare cut jobs and explicitly blamed AI. Snap did the same. Meta, Cisco, and Coinbase have all announced workforce reductions amid strategic pivots toward AI-driven efficiency. The pattern is consistent: company invests heavily in AI, company reduces headcount, company promises that the remaining workers will be more productive and better compensated.
What remains unproven is whether the productivity gains are real or imagined. Gartner's finding that job cuts are not translating into financial returns suggests that many companies are chasing a mirage — replacing experienced workers with AI systems that require constant supervision, produce inconsistent output, and create new categories of technical debt that no one knows how to manage.
ClickUp claims to be different. Evans insists his company is measuring real productivity gains. But the broader industry data suggests skepticism is warranted. AI agents can handle routine tasks impressively. They struggle with ambiguity, context, and the kind of cross-functional judgment that makes experienced employees valuable. A company that replaces 100 senior engineers with 10 AI whisperers may discover, six months later, that it has lost the institutional knowledge required to debug the systems the agents built.
The Ethical and Regulatory Dimension
Beyond the economics sits a question that policymakers have barely begun to address. If companies can use AI to eliminate most of their workforce, what obligation do they have to the people they replace?
The current answer, in most jurisdictions, is: none. Employment law does not require companies to retain workers when technology makes them redundant. Unemployment insurance provides temporary support but not retraining. And the pace of AI advancement is outstripping the ability of educational institutions to produce workers with skills that machines cannot mimic.
Some countries are experimenting with responses. The European Union's AI Act includes provisions for workplace impact assessments. South Korea has discussed AI displacement taxes. But these are early-stage ideas, not implemented frameworks. In the United States, where ClickUp operates, the political will to address AI-driven unemployment is fragmented at best.
Pope Leo XIV's recent encyclical on AI — which explicitly warned against weakening human agency and deepening inequality — adds a moral dimension to the policy debate. Whether religious doctrine influences corporate behavior remains to be seen. But the Vatican's framing of AI ethics as a religious imperative creates pressure that extends beyond secular regulation.
The Worker Response: Adapt or Die
For individual workers, the ClickUp announcement is a signal with painful clarity. The era of learning a skill, practicing it for decades, and earning a stable living is ending. The new reality requires constant adaptation, relentless self-education, and a willingness to treat one's job as a perpetually provisional arrangement rather than a career.
"The people that automate their jobs with AI will always have a job," Evans wrote. The statement is simultaneously a promise and a threat. It promises security for those who embrace AI. It threatens obsolescence for those who do not. And it ignores a critical question: what happens when the AI agents become so capable that they no longer need human direction?
Evans's million-dollar salary bands assume that human judgment will remain the scarce resource. But AI development is moving fast enough that this assumption may not hold for long. Today's AI whisperers are managing agents that handle routine tasks. Tomorrow's AI systems may handle strategic decisions, creative direction, and executive judgment with minimal human oversight. When that happens, even the survivors may discover they have automated themselves into irrelevance.
What Comes Next
ClickUp's experiment will be watched closely by every startup and enterprise considering AI-driven restructuring. If the company can demonstrate genuine productivity gains — not just cost savings, but superior output — it will become a template. If it struggles with quality, reliability, or the hidden costs of managing an AI workforce, it will serve as a cautionary tale.
The early indicators are mixed. Evans is confident enough to bet his company's culture and his personal reputation on the transformation. Gartner's data suggests most companies pursuing similar strategies are not seeing financial returns. And the human cost — 22% of a workforce suddenly unemployed, their skills now competing against machines that work 24/7 for the cost of electricity — is real and immediate.
The broader question is whether this is the beginning of a new economic model or the peak of a bubble. If AI agents genuinely enable 100x productivity gains, the companies that adopt them early will dominate their markets. If the gains are illusory — if the agents produce buggy code, misleading analysis, and fractured customer experiences — the companies that bet big on AI will find themselves with depleted workforces and broken products.
Either way, the workforce transformation is underway. ClickUp has made its bet. The rest of the industry is watching. And millions of knowledge workers are wondering whether their job will be next.