The job title survived. The job itself became something else.
A senior Google executive, drawing on decades of research and a doctorate in artificial intelligence, has argued publicly that fears of mass job elimination by AI are overstated — that history shows work transforms rather than disappears. Yet this reassurance arrives against a backdrop of more than 85,000 technology sector layoffs in the first months of 2026, a 33 percent rise over the prior year, raising an old and unresolved question: when those in power speak of gradual metamorphosis, who bears the cost of the interval between what was and what will be?
- Tech leadership is actively working to calm public anxiety about AI, but the soothing language of 'transformation' is colliding with the blunt reality of tens of thousands of people losing their jobs right now.
- Layoffs across the technology sector surpassed 85,000 through April 2026 — a 33% year-over-year increase — making optimistic executive messaging increasingly difficult to reconcile with the numbers.
- The historical analogies being offered, bank tellers and radiologists who adapted over decades, describe transitions that unfolded slowly enough for workers to adjust, a luxury the current pace of AI deployment may not allow.
- Displaced workers are not navigating a theoretical future labor market; they are searching for jobs today in a sector that is contracting, not expanding, while retraining pipelines and new roles remain largely unbuilt.
- The credibility of tech industry leadership is quietly fracturing — companies are simultaneously promising transformation and executing reduction, and the gap between those two positions is becoming impossible to ignore.
James Manyika, a Google vice president with a doctorate in artificial intelligence from Oxford, recently made a measured public case against apocalyptic predictions about AI and employment. His argument: technology transforms work rather than erases it. Bank tellers still exist in 2026, he noted — the title endured even as the job itself became something unrecognizable compared to 1970. Disruption, in his framing, looks less like an ending and more like a slow metamorphosis.
Manyika is not speaking casually. He co-authored influential research at McKinsey in 2017 examining automation's mixed effects on labor markets. His optimism is grounded in study, not wishful thinking. And in principle, he isn't wrong — roles do evolve, and history does offer examples of survival through technological change.
But the gap between that principle and the present moment is difficult to overlook. Through April 2026, the technology sector announced more than 85,000 layoffs — a 33 percent increase over the same period the year before. The companies delivering reassuring messages about transformation are, simultaneously, cutting faster and deeper than they were twelve months ago.
What separates this wave from previous ones is speed. When ATMs spread across the country, the displacement of bank tellers happened gradually enough that some workers could retrain, relocate, or simply age out. The current pace of AI deployment offers no such buffer. The job losses are arriving now. The new roles, the retraining programs, the reshaped economy — those remain largely theoretical.
The 85,000 people who lost work in early 2026 are not waiting for their positions to evolve. They are searching for new ones in a market that is shrinking. The distance between an executive's long view and a displaced worker's immediate reality has grown wide enough that it can no longer be talked away.
James Manyika, a vice president at Google, sat down recently to make a case that sounds almost reasonable on its surface: artificial intelligence won't destroy jobs the way people fear. The real story, he argued, is subtler. AI will transform what work looks like, not eliminate it wholesale. When tech executives talk about wiping out half the jobs in a sector, they're being hyperbolic. The work will change. The role will persist, just differently.
He reached for history to make the point stick. Bank tellers still exist. Radiologists still practice medicine. But ask yourself what a bank teller actually does in 2026 compared to 1970, Manyika said. The job title survived. The job itself became something else entirely. That's the real story of technological disruption, he suggested—not apocalypse, but metamorphosis.
Manyika brings credentials to the conversation. He holds a doctorate in artificial intelligence and robotics from Oxford. In 2017, he co-authored a report for McKinsey that examined the mixed effects of automation on employment. He's not speaking from ideology or fear. He's speaking from years spent studying exactly this problem.
Yet there's a widening gap between what Google's leadership is saying and what Google and its peers are actually doing. Through April 2026, the technology sector announced more than 85,000 job cuts. That's a 33 percent increase compared to the same stretch the year before. The layoffs keep coming. Manyika's optimism about transformation rings hollow when set against the actual numbers of people losing work.
The contradiction is stark enough to notice. Tech companies are simultaneously telling workers and investors that AI will reshape rather than destroy employment, while those same companies are cutting deeper and faster than they did a year ago. The messaging from leadership suggests patience and adaptation. The actions suggest urgency and reduction. One of these things is true. Both cannot be.
What makes this moment different from previous waves of technological change is the speed and the scale. When ATMs arrived, they didn't eliminate bank tellers overnight—but they did reduce the number of branches needed, and the number of tellers with them. The transition happened gradually enough that some workers could retrain, move, or retire naturally. The current pace of AI deployment doesn't offer that luxury. The cuts are happening now. The retraining programs, the new roles, the transformed economy—those are still theoretical.
Manyika's argument isn't wrong in principle. Jobs do evolve. Roles do transform. But there's a difference between saying that change will happen and saying it will happen painlessly, or that it's happening right now. The 85,000 people who lost work in the first four months of 2026 aren't waiting for their jobs to transform. They're looking for new ones in a market that's contracting, not expanding. The gap between the executive's optimistic framing and the worker's lived reality has become impossible to ignore.
Notable Quotes
The main effect of AI will be to change how people work, not eliminate jobs wholesale— James Manyika, Google VP
I guarantee that what a bank teller does today is not what they did in 1970— James Manyika, Google VP
The Hearth Conversation Another angle on the story
When Manyika says the job will transform rather than disappear, what does he actually mean?
He means the function changes but the category persists. A bank teller in 1970 counted cash and processed deposits manually. Now they handle customer relationships, sell financial products, troubleshoot digital systems. Same title, completely different work.
That's a fair historical observation. But does it apply to what's happening right now?
That's where the argument breaks down. Historical transformation happened over decades. Workers had time to adapt, retire, or move into new roles. What we're seeing now is 85,000 cuts in four months. The pace is different.
So Manyika isn't wrong about the long arc?
No. He's probably right that some roles will evolve rather than vanish. But he's describing a process that takes time, and the companies he represents are acting like time is something they don't have.
Why would they cut so aggressively if they believe transformation is coming?
That's the question nobody's answering. Either they don't actually believe their own rhetoric, or they're optimizing for quarterly earnings while hoping the transformation happens fast enough to rehire later.
Which seems more likely?
Look at the numbers. 33 percent increase year-over-year. That's not the pace of gradual transformation. That's the pace of cost reduction.