The White-Collar Reckoning Is Here — And It's Moving Slower Than You Think

The Numbers Behind the Anxiety

The anecdotes are alarming. The data is worse. In early 2026 alone, the US tech sector shed 32,000 jobs [3]. Zoom out to 2025 and you're looking at 1.17 million total US layoffs, with a growing share explicitly tied to AI replacement: Amazon eliminated 14,000 corporate roles, citing AI directly. Microsoft cut 15,000 positions as Copilot integration deepened. Salesforce replaced 4,000 support roles with AI agents. IBM handed HR functions to chatbots across multiple departments. Dario Amodei, CEO of Anthropic — the company building the AI doing some of this displacing — put his forecast in public: 50% of entry-level white-collar jobs gone within five years, with US unemployment potentially reaching 10-20% in that window [3]. Goldman Sachs is more measured, estimating 6-7% net displacement across the broader labor market. Neither number is comfortable.

Claudia Sahm's More Frightening Scenario

Here's where it gets interesting — and where economist Claudia Sahm, creator of the recession indicator that bears her name, departs from the doomsday crowd in a counterintuitive direction. Sahm doesn't think the Citrini-style catastrophic scenario is the most likely one. And she's not saying that to be reassuring. Her concern is the slow version [2]. In a rapid-displacement scenario — mass layoffs, unemployment spiking quickly — policymakers notice. They respond. Congress acts. Fiscal stimulus, tax code rewrites, maybe something like pandemic-era direct payments. The system has shock absorbers for visible crises. The scenario that keeps Sahm up at night is the slow-moving one: a steady, quiet displacement over years. Enough attrition that individuals feel it painfully but the aggregate numbers don't cross the political alarm threshold. Policymakers hesitate. The stimulus is late. The pain accumulates without the relief. "I worry more about that scenario," she said. "A slow-moving, slowly building crisis." That's not a headline that gets 85 million views. But it might be the one that matters more.

I've Seen This Movie Before — In Medicine

I spent ten years as an ER physician before moving into tech. I watched the medical profession go through its own automation anxiety decades before white-collar workers are living through theirs now. In the 1980s and 90s, diagnostic imaging began automating what radiologists did. AI-assisted ECG interpretation threatened cardiologists. Lab automation threatened pathologists. Telemedicine threatened everyone. The anxiety was real, and some of it was warranted — certain tasks did get automated, certain roles did shrink. But medicine adapted. The roles that survived were the ones that leaned into judgment under uncertainty, human relationships, and integrating complex contextual information — exactly what machines struggled with. Radiologists became interpreters and consultants, not just image-readers. Clinicians became orchestrators of care, not just diagnosticians. The white-collar world is now living through a compressed, faster version of that transition. The difference is the pace. Healthcare had decades. Software engineers, financial analysts, and middle managers may have years — maybe less. The workers who will come out the other side aren't the ones denying the shift. They're the ones asking: what does the machine still need me for?

The New Shape of the Threat

What makes this wave different from previous automation cycles is the breadth and the target. Past automation hit physical labor and routine cognitive work — assembly lines, data entry, call centers. Those workers had less political capital, less ability to shape the narrative, and got less policy attention as a result. This wave is hitting knowledge workers — the people who write the op-eds, fund the campaigns, and run the companies [1]. The Citrini essay went viral not just because it was scary but because it was written by and for the kind of people who normally watch disruption happen to someone else. That's what makes this political moment unusual. The threat is finally touching people with the platform to demand a response. Whether that response comes in time — and whether it's the right kind — is the question policymakers should be asking right now, before the slow-moving crisis Claudia Sahm described becomes undeniable.

The Bottom Line

This isn't a doomsday piece. The Citrini scenario is speculative. Goldman Sachs' economists are right that macroeconomics 101 pushes back on the most extreme projections. But Jack Dorsey firing 40% of his company and watching his stock go up 14% is not a drill [4]. That is a signal — a clear, public, financially rewarded signal — that eliminating human labor with AI is now shareholder-value-positive. That signal will be received. It will be repeated. The question isn't whether the white-collar reckoning is happening. It's whether the policy, the education systems, and the individual workers are ready for the pace at which it's arriving. Based on everything I've seen in medicine and in tech — my honest read is: not yet. But not never. The workers who start asking the right questions today will have a very different 2030 than those who wait for certainty before acting.