Alternative investing
Aug 8, 2025
Jobs in the AI revolution: Disruption today, growth tomorrow
Artificial intelligence (AI) could already be disrupting the labor market.
There are signs that recent college graduates are having a harder time finding jobs because of AI.1 Microsoft announced this year that it will cut around 15,000 jobs as it invests more money into developing its AI capabilities.2
The impact may only spread from there: Venture capitalists and growth equity managers view the 71 million knowledge workers in the United States (with average annual salaries of $85,000) as a $6 trillion market opportunity. The International Monetary Fund estimates that over 60% of jobs in the developed world are AI-exposed.3
This could come with major downsides. Dario Amodei, the CEO of AI startup Anthropic, has said that AI could result in unemployment rates as high as 20%.4 This raises obvious questions: If AI does all the jobs, who will earn the money to spend on goods and services? How can AI benefit the economy and stock market if it causes mass unemployment?
While the conflict appears worrying, a closer inspection of technological innovation throughout history reveals a much more nuanced pattern, and provides reason for optimism. From the steam engine to electricity, to the mainframe computer, technological breakthroughs didn’t lead to mass unemployment. Rather, they slashed the cost of transportation, industrial power and information processing, which led to transformed demand patterns in the economy. New functions became possible, and over time, these more than offset the losses from obsolete tasks.
We think AI could follow the same trajectory: Violent task churn, then broad productivity growth. Modern macroeconomic management, influenced by Keynesian economics and sharpened through crises such as the global financial crisis and COVID, could further soften the blow to the labor market.
Finding the pattern
The steam engine revolutionized the textile industry and transportation in the early 1800s. Hand-loom weavers were the first to feel the effects of disruption from mechanized power looms, as their real wages were cut in half between 1806 and 1820.
Steam-powered trains made canal boatmen and cart drivers obsolete. On the other hand, cloth output and textile consumption soared, and inland freight rates fell. These shifts created new demand for labor in expanding industries such as coal mining, rail maintenance and urban retail.
Electricity improved on steam’s promise. In 1899, less than 5% of factories in the United States were powered by electricity. By 1929, over 75% of factories were electrified. Gas lamps were replaced, decimating employment for lamplighters. However, factory output surged.
Electricity led to entirely new industries, innovations and occupations as the marginal cost of light, sound and motion fell to nearly zero. Retail stores could now stay open past sunset. Electric projectors and safe indoor lighting enabled movie theaters and the cinema industry. Elevators transformed the sky into real estate. Household appliances decreased the burden of tasks such as laundry, cleaning and food preservation.
No futurist could have foreseen these layers of electricity’s impact when Thomas Edison lit Pearl Street for the first time.
Electricity, transistors and the integrated circuit enabled the mainframe computer in the 1950s and 1960s. Information processing—the art of collecting, organizing, storing, retrieving and communicating data to make decisions—was transformed. Banks were able to automate the tasks performed by thousands of accountants and bookkeepers, and after the introduction of the ATM, the number of bank tellers per branch fell from 20 to 13.
However, these innovations allowed banks to open far more branches, increasing employment overall. Low costs of data organization and arithmetic allowed credit cards and airlines to scale national networks. Stenographers became obsolete, but demand for programmers and financial analysts soared. Businesses also become more productive. In the 1980s, it took eight employees to generate $1 million of revenues. By the 2000s, it took only six.
This highlights another long-term trend: New innovations are contributing more quickly to overall productivity growth.
Even the optimists may underestimate the speed of AI’s productivity gains
Years from innovation to productivity growth
Source: U.S. Census Bureau, Business Trends and Outlook Survey, J.P. Morgan. Data as of November 2023.
What we can learn
With hindsight, the benefits of these technological feats are clear. But the AI revolution could have similarities to two other secular trends that are worth considering.
The first is the mechanization of agriculture through internal combustion tractors, self-propelled combine harvesters, mechanical cotton pickers, hybrid seeds, synthetic nitrogen fertilizers, pesticides and center-pivot irrigation. Farm output and productivity soared as employment in the United States collapsed from around 10 million in 1930 to only 3.5 million in 1970. However, food quality and supply increased, while prices dropped. Cities and suburbs boomed.
Globalization, which allowed corporations to increase productivity by accessing lower-cost workers, disrupted manufacturing jobs in the United States and the broader developed world. The United States shed over 6.5 million manufacturing jobs between 1979 and 2019. This disruption led to acute pain in the Rust Belt, and had social and political ramifications that are still being felt today. It also resulted in higher corporate earnings, wider margins, lower goods inflation and a less cyclical economy.
Those transitions took place over decades. Early adoption of AI has been much more rapid: ChatGPT is rapidly approaching 1 billion users in just over two years. However, deep corporate adoption will take time.
Recent research has shown that 60% of all job-related queries to Claude (a leading AI model) were related to job augmentation, not automation, suggesting that AI will augment work long before replacing it. Deep AI integration will require years of data architecture reorganization and process redesign. The stock market may begin to price in the benefits now, but it could take years for the labor market impact to be realized.
Recent research has suggested that at most only 2.5% of all jobs are currently at risk of automation. Humans also have some comparative advantages that ought to endure: common sense, causal inference, dexterity, emotional intelligence, high stakes accountability, adaptive learning and intrinsic motivation, to name a few.
The importance of policy
These secular labor market transitions were not without cyclical hardships. Especially in the pre-Keynesian economic era, rudimentary macroeconomic management—the inflexibility of the gold standard, lack of fiscal stabilization, nonexistent central bank liquidity—resulted in painful booms and busts. Some of these caused widespread job losses. During the 1970s, oil price and dollar purchasing power shocks sparked a long period of stagflation, even while firms were adopting the computer, reallocating labor and improving productivity.
Today, employers and policymakers have the necessary tools to help cushion future fallout in some tasks, especially while AI is in its nascent stages. The Federal Reserve seems to have plenty of room to ease monetary policy to stimulate demand in interest rate–sensitive sectors such as housing, especially if labor market slack builds due to AI disruption.
Meanwhile, employers can build trust by targeting low-value, “no-joy” tasks for replacement by AI, and by retraining existing employees for new roles. Policymakers can also help cushion shocks by incentivizing apprenticeship programs and encouraging transparent tracking of sectors with increasing labor redundancy.
What to expect from the AI revolution
Despite these cyclical shocks, we can see that historically, technological breakthroughs begat price declines, which begat demand surges, which begat productivity booms. Tasks became obsolete, but aggregate demand for labor grew. The steam engine increased demand for coal miners, electricity necessitated powerhouse crews, and the mainframe led to the Wall Street analyst.
Today, generative AI is driving down the cost of cognition, which means that more people will have more access to expertise. High-touch legal advice will likely become increasingly attainable for households and small businesses. Small-budget real estate development projects may be able to afford premium architectural design services. AI-driven diagnostic triage could lower hospitals’ cost per patient and increase caseloads. From a macroeconomic perspective, AI technology could help offset the risks to the working-age population posed by demographics and stricter immigration policies.
Businesses will probably not pass the cost savings from enterprise AI use cases entirely to shareholders; they are more likely to reinvest in new growth opportunities. AI will likely lead to employment gains in the companies that build software applications and data infrastructure, as well as those that assemble and embed the new AI-driven tools into existing workflows and data systems. Some tech industry workers are already branding themselves as “AI architects”—those with expertise in building AI applications.
Demand for non-work leisure activities and services may increase as knowledge workers spend less time at the office. AI will also likely do work that is not being done today: Think about use cases such as outbound customer service or prospecting calls.
The value proposition of AI is that it reduces costs and increases output for enterprises by making all workers more productive and specific tasks obsolete. Said differently, the total addressable market for AI is human labor.
But the pattern we’ve shown, across generations of technological innovations, strongly suggests that AI will increase productivity and economic growth, and will create new channels for aggregate demand without lasting labor market damage. Disruption, not destruction. We think there is opportunity for outsized market returns in the companies that engineer the AI boom, and the enterprises that effectively adopt the technology.
We can help
For more about investing in artificial intelligence in ways that serve your needs and goals, contact your J.P. Morgan team.
1 Derek Thompson, “Something Alarming Is Happening to the Job Market,” The Atlantic, April 30, 2025, https://www.theatlantic.com/economy/archive/2025/04/job-market-youth/682641.
2 Alex Halverson, “Microsoft to Lay Off About 9,000 Employees in Latest Round, Seattle Times, July 2, 2025, https://www.seattletimes.com/business/microsoft/microsoft-to-lay-off-as-many-as-9000-employees-in-latest-round.
3 Kristalina Georgieva, “AI Will Transform the Global Economy. Let’s Make Sure It Benefits Humanity,” IMF.org, January 14, 2024, https://www.imf.org/en/Blogs/Articles/2024/01/14/ai-will-transform-the-global-economy-lets-make-sure-it-benefits-humanity.
4 Jim Vandenhei and Mike Allen, “Behind the Curtain: A White Collar Bloodbath,” Axios, May 28, 2025, https://www.axios.com/2025/05/28/ai-jobs-white-collar-unemployment-anthropic.
Important Information
The rise of AI is set to disrupt traditional jobs, but as with past innovations, it should pave the way for new opportunities and increased productivity.
Jacob Manoukian
U.S. Head of Investment Strategy
Published Aug 15, 2025
you may also like
Experience the full possibility of your wealth
We can help you navigate a complex financial landscape. Reach out today to learn how.
Contact us
