Torsten Sløk has a blunt message for executives and policy makers rattled by headlines of AI-driven layoffs. Zero evidence exists that the technology destroys net employment. The data tell a different story.
The chief economist at Apollo Global Management examined weekly ADP private payroll figures through late May 2026. They show steady gains. No dip appears that one could tie to artificial intelligence. Instead companies hire specialists to implement the new tools. Data centers rise. That construction lifts demand for semiconductors, equipment and power. Salaries for AI talent climb. Prices follow.
“There is zero evidence of job losses because of AI,” Sløk wrote in his May 29 note on the Apollo site. “The bottom line is that the AI spending boom is stoking both employment and inflation.” He expects May nonfarm payrolls to beat the consensus forecast of 95,000 by a noticeable margin. (Apollo Global Management)
The argument rests on a 160-year-old observation. When a resource or process becomes cheaper, consumption of it rises. William Stanley Jevons noticed this with coal in the 1860s. Efficiency improvements did not shrink the coal industry. They expanded it. Sløk calls the modern version the Jevons employment effect. Cheaper cognitive work through AI does not shrink the market for human effort. It enlarges it. (Fortune)
But the narrative in corporate earnings calls and on social media runs the other way. Block’s Jack Dorsey said AI efficiencies would allow smaller, flatter teams and announced cuts affecting roughly 4,000 people. Klarna let natural attrition shrink its staff after adopting the technology for marketing, legal work and customer communications, saving about $10 million a year. Amazon’s Andy Jassy told employees that generative AI would reduce the corporate head count over time even as new roles emerge. JPMorgan’s Jamie Dimon urged leaders not to ignore the shifts ahead.
Some of those claims look like repositioning. OpenAI’s Sam Altman has spoken of “AI washing,” where executives blame the technology for reductions they would have pursued anyway to lift margins or trim bureaucracy after pandemic-era hiring binges. Anthropic’s Dario Amodei once warned of sweeping white-collar displacement but later emphasized expansion of work. Jensen Huang of Nvidia dismissed the rush to credit AI for every layoff as a “lazy narrative.”
Numbers back the skeptics. Roughly 55,000 U.S. jobs carried an explicit AI link in 2025, according to one tally from the National Bureau of Economic Research. That figure equals only 4.5 percent of the 1.2 million total layoffs recorded that year. A Yale Budget Lab analysis through early 2026 found no statistically significant negative effect on employment or wages in AI-exposed occupations when compared with synthetic controls. Unemployment rates for roles most open to automation sit below those for less-exposed work. (Business Insider)
Offshore centers offer another data point. Philippine and Indian call-center employment kept rising through 2025. The Philippines nearly doubled its head count over a decade to two million. “As AI makes call center work cheaper and faster, companies are buying more of it, not less,” Sløk observed. The pattern matches the Jevons logic exactly. Lower unit costs spur higher volume.
Yet the discussion remains heated. An EY survey found 60 percent of financial-services CEOs expect AI to maintain or increase staff levels in 2026. Aaron Levie, Michael Dell, David Sacks and Goldman Sachs’ David Solomon have echoed Sløk’s productivity-and-employment view. At the same time, surveys of HR leaders show 89 percent anticipate visible changes to job content next year. BCG estimates that 50 to 55 percent of U.S. roles will be reshaped within two to three years, with only 10 to 15 percent truly at risk of elimination. Most workers stay but face new expectations around output and skills. (BCG)
The labor market has absorbed earlier technology waves. Personal computers raised productivity and employment growth by roughly 1.7 percent annually in following decades. AI appears on track to repeat the pattern, only faster. Demand for implementation experts, data-center operators, energy traders and domain specialists who pair judgment with machine output grows immediately. Entry-level tasks that once trained young professionals may compress. That creates a separate policy question around reskilling and career ladders, but it does not signal aggregate job destruction.
Sløk’s read carries weight on Wall Street because Apollo manages hundreds of billions and his forecasts have often aligned with realized data. He does not dismiss displacement risk in specific tasks. He simply notes that the macro evidence so far points the opposite direction from the popular story. Hiring of AI talent plus capital spending on infrastructure outweighs any efficiency-driven head-count reductions.
Inflationary side effects deserve attention too. Higher semiconductor and electricity demand from data centers already shows in price data. Wage pressure for scarce technical skills adds another channel. Central bankers may face a growth-plus-inflation mix rather than the deflationary productivity surprise some predicted. Stronger employment supports consumer spending. That further buoys demand.
Executives face a practical choice. They can treat AI as a pure cost cutter and risk missing the demand expansion Sløk describes. Or they can invest in people who amplify the machines and capture new revenue streams. History favors the second path. So does the current ADP print.
Markets will test the thesis in coming months. If May and June payrolls exceed expectations while AI-related capital expenditure stays hot, Sløk’s Jevons employment effect gains converts. If unemployment ticks up sharply in white-collar categories, the doomers regain the microphone. Early returns favor the optimists. But the sample remains short. The technology itself continues to improve.
One fact stands clear today. The labor market has not cracked under AI’s weight. It shows signs of tightening in precisely the areas the new tools touch. Firms chase talent instead of shedding it. That reality should temper panic and focus attention on adaptation rather than resistance. And the data keep coming.
WebProNews is an iEntry Publication