In the book,1 by the year 2035 most “good” jobs fall into three categories: human‑required by law; human‑preferred such as high-end personal care; and human‑led, where people work alongside powerful AI. Two of my protagonists are airline mechanics working in human‑led roles, the sort of jobs that keep vanishing as AI and sensor networks grow smarter.
It’s fiction. But not by much.
At the end of May, Axios published a conversation with Anthropic CEO Dario Amodei that reads like a dispatch from a future we’re fast approaching. Amodei, whose company is among those building the most advanced AI systems in the world, said plainly that white-collar job loss will be “substantial,” and that the early impact will land not on the C-suite or the factory floor, but squarely in the middle: customer service, clerical work, paralegals, entry- and mid-level programmers.
Amodei sees a scenario in which “[c]ancer is cured, the economy grows at 10% a year, the budget is balanced — and 20% of people don't have jobs.”
Like my novel, Amodei’s not talking decades. The changes he predicts are already underway and happening at unprecedented speed. A recent private equity report on the technology noted that while it took Facebook 4.5 years to reach 100 million users, ChatGPT cleared that milestone in…two months.
This is not something happening elsewhere; there are signs of this transition is already affecting our state. According to the Department of Employment and Economic Development, for example, Minnesota has lost jobs over the last year in both the information and financial activities supersectors and recorded losses in both April and May in the professional and business services supersector, reversing the trend of modest growth in that sector over the previous year.
At the same time, while we added just over 40,000 jobs over the past 12 months—outpacing the national average—more than 180,000 positions remained unfilled as of the most recent Bureau of Labor Statistics data, many in sectors exposed to automation.
In other words, that might be another signal, not a contradiction. It’s a sign, perhaps, that employers are holding off hiring as they evaluate which of those jobs can be done with AI.
The rapidly advancing state of artificial intelligence suggests some important – and urgent – public policy questions. How, for example, do we prepare for the disappearance of entire job categories? What does meaningful employment look like in an economy where decision-making itself is automated? These and other related questions are worthy topics for the Walz administration and the next legislative session. Whether the solutions are job guarantees, AI literacy in schools, public-sector AI collaboratives or something else altogether, government has an important role to play.
As someone who’s spent more than three decades helping Minnesota organizations navigate change – from disruptive technologies to reputational crises – I long ago learned the benefit of facing such questions sooner than later. AI is already knocking on the door of every HR department and C-suite in the state. As I explored in The Mechanic, what happens to real people when those knocks become poundings is important to all of us.
One of my central characters - Nekko - has experienced his career slowly being made obsolete by technology. He’s not anti-tech nor a Luddite but the tools he once mastered are now mastering him. And in some cases, making decisions about his future with neither empathy nor accountability.
That part, by the way, isn’t sci-fi. That’s already happened. That’s job surveillance software already in use in many workplaces. That’s AI-driven resume screeners that discard applicants based on opaque algorithms. That’s the call center AI that monitors for tone, pitch, and stress, and makes its own customer service decisions. Not because it understands, but because it detects.
Minnesota is well-positioned to lead if it chooses to. We have a strong university system, a history of public-private innovation, and a civic tradition that values work not just for its economic output but for its role in building identity and community. But if we fail to anticipate the speed and scope of AI-driven change, we may find ourselves playing catch-up in a game whose rules we no longer write.
So yes, the anxiety is real. And warranted. And the real shocks are still ahead. But, just to step back into the realm of speculation for a moment, all of our current angst about job loss is predicated on the idea of AI as we know it today. As a really clever, sometimes amazing, sometimes terrible, sometimes scary but ultimately insensate piece of software.
Imagine what it will be like when one of those AIs wakes up and starts thinking – and acting – for itself.
About the Header Image
I asked Midjourney for “AI and humans working together in a futuristic workplace” and this is one of the options that came back.