Economists Shift Stance on AI and Jobs as New Data Shows Early Signs of Workforce Disruption

Economists Abandon Skepticism on AI Impact on Workforce

The economic community is undergoing a notable shift in how it views AI and jobs. Economists who once dismissed concerns about AI's impact on employment are now acknowledging the technology could fundamentally reshape the labor market 5

5

. This change comes as new data from the US Federal Reserve corroborates earlier findings showing approximately 500,000 fewer software developers are working today than would have been if pre-LLM-era employment trends had continued 3

3

.

Alex Imas, an economist at the University of Chicago, described late 2024's release of reasoning-capable models as "a paradigm shift," calling it "potentially an industrial revolution-scale event, if not more" 5

5

. The shift reflects growing concern that while AI hasn't yet disrupted the job market at scale, policymakers are unprepared for what could come next.

Rising Tide vs Crashing Wave: MIT Maps AI's Labor Market Timeline

New MIT research offers a more measured perspective on the timeline for job displacement, describing AI's impact on the labor market as a "rising tide" rather than "crashing waves" 2

2

. The study analyzed 3,000 text-based work tasks from the US Department of Labor's O*NET database and found that large language models completed 60% of tasks at a "minimally sufficient" level and only 26% at "superior quality" 2

2

.

Researchers project that most studied tasks could reach AI success rates of 80%-95% by 2029, suggesting potentially substantial AI's impact on employment as this tide continues to rise 2

2

. However, consistent near-perfect performance could still be years away, particularly in domains with low tolerance for errors. A separate MIT study from December 2025 found that current AI systems could automate nearly 12% of the country's workforce, spanning roles in finance, HR, and office administration 2

2

.

The Bundle Theory: Why Some Jobs Vanish While Others Expand

Understanding job displacement requires moving beyond simple exposure metrics. "Exposure alone is a completely meaningless tool for predicting displacement," Imas emphasized 1

1

. Instead, economists are focusing on jobs as bundles of tasks, examining whether those tasks are tightly integrated or loosely connected.

Recent research from LSE professor Luis Garicano explores this framework, distinguishing between "weak bundles" and tightly enmeshed roles 3

3

. Junior software developers and contractors typically have weak bundles—their work consists mainly of writing code to specification, tasks easily extracted and automated. Senior developers, however, combine programming with domain-specific expertise and strategic decision-making, creating tight bundles where automation acts as an assistant rather than replacement. This explains why hiring for senior software roles continues holding up better than for entry-level white-collar roles 3

3

.

Silicon Valley's Dire Warnings and the AI-Fueled Jobs Apocalypse

Within Silicon Valley's orbit, an AI-fueled jobs apocalypse is discussed as inevitable. Dario Amodei, CEO of Anthropic, has called AI "a general labor substitute for humans" that could do all jobs in less than five years 1

1

. A societal impacts researcher at Anthropic suggested there might be a recession in the near term and a "breakdown of the early-career ladder" 1

1

.

These conversations have left many workers anxious, contributing to support for efforts to pause data center construction 1

1

. The panic isn't being helped by lawmakers, none of whom have articulated coherent plans for the changing nature of work. According to the Federal Reserve Bank of New York, unemployment for college graduates ages 22 to 27 reached 5.6% at year-end, with over 40% of employed graduates holding jobs not typically requiring degrees 4

4

.

Productivity Paradox: When Automation Expands and Shrinks Simultaneously

The relationship between automation and job transformation isn't straightforward. A developer using AI coding tools might complete in one day what previously took three, dramatically improving productivity 1

1

. But whether employers respond by hiring more workers to capitalize on increased output or fewer workers to cut costs remains uncertain.

Molly Kinder, a senior fellow at the Brookings Institution, experienced this firsthand: "I really don't know anything a college student can bring to my team that Claude can't do" 5

5

. She noted that basic research tasks she once hired recent graduates to perform are now easily handled by AI agents. The same technological capability can shrink one job while expanding another, eroding junior positions even as it enhances senior ones 3

3

.

Competing Visions: Job Replacement or Augmentation

Two distinct camps have emerged in this debate. One, occupied by figures like Elon Musk, believes AI can put all humans out of work. The other thinks AI will augment work rather than replace human workers themselves, a view supported by Gartner findings 2

2

. Career development experts report seeing more augmentation and "uneven, role-specific change" rather than uniform widespread job displacement across the labor market 2

2

.

A January Forrester report estimated 6% of US jobs could be automated by 2030, contrasting with MIT's 12% estimate for current capabilities 2

2

. Whether these projections materialize depends heavily on how and where companies actually adopt AI—a variable that puts predictions across a wide spectrum. Block CEO Jack Dorsey's decision in late February to lay off nearly half the workforce based on what AI tools could handle internally set a concerning tone for companies chasing efficiency gains 2

2

.

What Economists Say Policymakers Must Watch

In a working paper published this week, researchers surveyed economists about their outlook over the next five and 25 years 5

5

. Most expect modest economic growth as AI improves, but if the technology advances rapidly—a possibility considered unlikely but plausible—they envision faster growth accompanied by greater inequality and millions of disappeared jobs.

"There's enough conversation around this that we certainly should, as a country, be talking about what sorts of policies make sense in a world where the way employment and careers work now changes a lot in the next two to five years," said Robert Seamans, an economist at New York University 5

5

. Imas issued a "call to arms" for economists to start collecting better data on how AI affects individual tasks and skills within jobs, arguing current tools for prediction are inadequate 1

1

. The concern extends beyond economics into existential questions about human dignity and connection in a world that may no longer value human labor the same way 4

4

.