ClickUp just laid off 22% of its workforce and framed it not as cost-cutting, but as a bet on AI. That’s the headline from TechCrunch AI, and it’s worth sitting with. CEO Zeb Evans announced the cuts on X, then promised that most of the savings would flow back to the people who stayed. He’s even floating million-dollar salary bands for anyone who creates outsized impact with AI.
The framing matters here. Layoffs usually come wrapped in apology and balance-sheet language. Evans went the other way. He’s calling this a “radical embrace of AI” meant to turn ClickUp into what he describes as a “100x org.”
What actually changed inside ClickUp
According to TechCrunch AI, ClickUp recently rolled out roughly 3,000 internal AI agents to handle complex tasks. The job of the remaining staff has shifted. They no longer do the work directly. They direct the agents and review the output to make sure it clears the company’s bar.
That’s the real story under the layoff number. The unit of work is changing from “person doing a task” to “person managing a fleet of agents.” Evans put it bluntly: “The people that automate their jobs with AI will always have a job.”
The catch nobody’s saying out loud
Here’s the tension. If agents keep absorbing more tasks, ClickUp needs fewer and fewer people over time. The ones who don’t automate well get cut. Evans is honest about that logic, even if it’s uncomfortable.
And the broader data isn’t fully on his side. TechCrunch AI cites a recent Gartner survey showing about 80% of companies using autonomous tech have already cut jobs. The kicker: those cuts aren’t reliably producing real financial returns. Gartner’s read is that some firms are using unproven AI as cover to downsize.
ClickUp says it’s not in that bucket. Evans told TechCrunch the company is measuring genuine productivity gains from its agents, and it’s building those efficiencies into a product for customers.
Tokenmaxxing vs. value created
There’s a quieter trend worth flagging. More companies are now tracking how many tokens each employee burns through, using it as a proxy for who’s actually adopting AI. The practice has a name: “tokenmaxxing.”
Critics say it’s the wrong number. High token use just means high AI spend, not high output. Evans agrees. “Instead of gamifying token cost, we gamify value created and time saved,” he wrote. That distinction is going to matter a lot in the next year as managers figure out what “AI adoption” even means on a scorecard.
The extreme version already exists
To see where this road leads, look at Polsia. TechCrunch AI points to the one-year-old startup that claims to run all software operations for solopreneurs. It’s staffed by exactly one person, founder Ben Broca. Polsia just raised $30 million at a $250 million valuation.
That’s the proof-of-concept for the “one human, many agents” model. Whether it scales past a niche is the open question.
What this means for you
The next 1-3 years are going to reward a specific skill: directing AI, not racing it. A few practical moves while the dust settles.
- If you’re an operator or employee: Get fluent at managing agent output, not just prompting. The valued role is becoming reviewer and director, not doer.
- If you run a company: Measure value created and time saved, not token spend. Gartner’s data says cuts without real gains are common, so don’t confuse activity with results.
- If you’re hiring: The Evans thesis is that automators keep their jobs. Screen for people who already restructure their own workflows around AI.
My take? ClickUp is testing a thesis in public, and that’s brave. But the Gartner numbers are the warning label. Plenty of firms are dressing up layoffs as AI strategy, and the returns aren’t showing up yet. ClickUp’s million-dollar salary bands will be the tell. If the people who stay actually get paid that way, the bet was real. If not, it was a layoff with better marketing.
Full details are at the original TechCrunch AI report.