画像クレジット: Image via source article. Used under fair use for news commentary. · source
Is AI coming for your job this quarter, or is the whole thing overblown?
If you listened to Wall Street this week, you'd get whiplash trying to figure it out. Two major voices, both with serious credentials, both looking at the same economy, and they can't agree on whether artificial intelligence is reshaping the labor market right now or whether we're all getting ahead of ourselves. Again.
I've seen this movie before. The script hasn't changed much since the dot-com boom, or the mobile revolution, or (call me old-fashioned) the introduction of ATMs that were supposed to eliminate bank tellers by 1990. Spoiler: there are still bank tellers.
Ozan Tarman, global macro vice chair at Deutsche Bank, went on Bloomberg Television this week to argue that AI isn't just another tech trend, it's THE narrative driving stock markets right now. Not Iran, not interest rates, not whatever the Fed chair said at breakfast. AI.
Tarman's position is that the technology is already having real effects on employment patterns, that companies are making hiring decisions today based on what they think AI will be able to do in 18 months. This is the anticipatory displacement theory, if you want to get academic about it. Businesses aren't waiting to see if the robots can actually do the job, they're cutting headcount now on the assumption that they will.
Now, there's something to this. I've talked to enough manufacturing executives and logistics managers over the years to know that capital expenditure decisions get made on vibes as much as spreadsheets. If every CEO at every conference is talking about AI efficiency gains, the pressure to "get lean" becomes enormous even if nobody can quite explain what tasks the AI will actually take over.
関連記事
More in AI Models
I spent a week parsing the claims around Google's new 'always-on' AI agent, and the answer is more complicated than the marketing suggests.
Aisha Patel · 5 hours ago · 7 min
The AI company is now officially the world's most valuable startup, and it's moving fast toward public markets.
James Chen · 6 hours ago · 3 min
The Claude maker beat OpenAI to the SEC paperwork, but I've seen enough tech IPO races to know this is really about runway, not rivalry.
Mark Kowalski · 6 hours ago · 5 min
Everyone's writing about the $200B CPU market grab. The actual story is how Nvidia is quietly becoming the landlord of global AI compute.
But here's where I get skeptical. The stock market narrative and the labor market reality are two very different animals. Stocks can run on stories for years! The actual economy, where people need paychecks and pay rent, tends to be more stubborn.
Enter Torsten Slok, chief economist at Apollo, who basically said everyone needs to calm down. In his Bloomberg appearance, Slok argued that the May jobs report won't show any meaningful AI impact, and that people are fundamentally misunderstanding how these technologies actually change employment.
Slok's argument, if I'm reading between the lines correctly, is that AI displacement is a slow-motion phenomenon. It doesn't show up in monthly payroll numbers. It shows up over years, decades even, as job categories gradually shift and new roles emerge to replace old ones. The ATM didn't kill bank tellers, it just meant banks could open more branches with fewer tellers per location. Net employment in banking actually grew for decades after ATMs became ubiquitous.
This is the more historically literate position, I think. Every major technological shift has followed roughly the same pattern: enormous hype about imminent job destruction, followed by a much messier reality where some jobs disappear, others transform, and entirely new categories emerge that nobody predicted. The kids building AI startups today probably can't imagine what the job market will actually look like in 2035, and honestly, neither can I.
Here's the uncomfortable truth: both Tarman and Slok might be right, just about different timeframes and different parts of the economy.
The white-collar knowledge work that AI is best positioned to automate (writing, coding, data analysis, customer service) tends to be concentrated in certain industries and certain income brackets. A paralegal in Manhattan faces very different AI exposure than a plumber in Phoenix. The aggregate jobs numbers might look fine even if specific sectors are getting hammered.
We also don't have great data on this yet. The Bureau of Labor Statistics wasn't designed to track "jobs eliminated because management bought ChatGPT Enterprise." Companies don't report that. They say they're "restructuring" or "improving efficiency" or "right-sizing the organization." Good luck separating AI-driven cuts from normal business cycle layoffs or plain old bad management.
I only found two serious economic analyses trying to quantify near-term AI job displacement, and they disagreed by a factor of five. That's not science, that's astrology with spreadsheets.
Look, I've been covering technology long enough to know that the truth usually lands somewhere boring. AI will probably eliminate some jobs, create others, and transform a whole bunch more in ways that are hard to predict. The timeline will be longer than the enthusiasts claim and shorter than the skeptics assume. Some workers will get screwed, others will thrive, and policy will lag behind reality by about a decade, as it always does.
What bothers me about the current conversation is how binary it's become. You're either a doomer convinced that 40% of jobs will vanish by 2030, or you're a skeptic who thinks the whole thing is overhyped nonsense. Neither position is particularly useful for, say, a 45-year-old accountant trying to figure out whether to invest in learning new skills or ride out the next 20 years until retirement.
The honest answer is: we don't know yet. We don't know which specific tasks AI will master next, we don't know how quickly companies will actually deploy these tools at scale (as opposed to just talking about it), and we don't know how workers and institutions will adapt. Anyone who tells you they have the answer is selling something.
The May jobs report will come out. It probably won't show dramatic AI effects, Slok is likely right about that. And the stock market will continue treating AI as the dominant narrative, Tarman is probably right about that too. These two things can coexist because markets and economies operate on different timescales and different logics.
What I'll be watching is the sector-specific data. Are legal services firms hiring fewer paralegals? Are customer service centers shrinking faster than normal attrition would explain? Are coding bootcamps seeing enrollment drops because kids are worried the jobs won't exist? That's where the early signals will show up, not in the aggregate payroll numbers.
In the meantime, I'd suggest ignoring anyone who speaks with certainty about AI and jobs. This includes most tech CEOs, most economists, and definitely most LinkedIn influencers. The self-driving car hype cycle taught us that timelines slip, capabilities plateau, and the future arrives slower and weirder than anyone expected.
But what do I know. I still prefer email to Slack, and I've been wrong about technology timelines more times than I can count. If you want to argue about any of this, my email's on the about page.
