Key Takeaways
- • Apollo Global Management's chief economist Torsten Slok says AI has yet to deliver a financial payoff to US businesses outside the tech sector, with revenue per worker and profit margins rising only for the "Magnificent Seven" (hyperscalers plus Apple, Nvidia and Tesla).
- • Apollo argues corporate America is increasingly treating AI token spending as a cost center that weighs on margins rather than a growth driver.
- • Apollo cites a slowdown in token spending growth as evidence that demand is highly price-sensitive, casting doubt on the economics of building ever more datacenters.
- • The claims align with prior research from PwC, Deloitte and McKinsey, which report that most companies have not seen earnings gains from AI, though McKinsey sees a turning point toward returns.
- • Critics on social media disputed Apollo's picture, and some noted that PC productivity gains also lagged adoption by a decade or more.
Summary
Apollo Global Management has argued that artificial intelligence has yet to produce a financial payoff for US businesses outside the technology sector, according to a post on the firm's Daily Sparkblog reported by The Register. The firm said corporate America is beginning to view AI token spending as a cost center that weighs on company margins rather than a source of growth.
Apollo's chief economist, Torsten Slok, said revenue per worker appears to be rising for the so-called "Magnificent Seven" technology firms, which include the hyperscalers as well as Apple, Nvidia and Tesla. He said the same is not true for the rest of the S&P 500. Slok also said profit margins and AI-driven gains in corporate productivity growth show the same pattern, and he supported these claims with charts.
Apollo's assessment drew pushback. Some commentators on social media said they did not recognize the picture Slok painted. Others pointed to the PC industry, where productivity growth lagged widespread business adoption by a decade or more.
The Register noted that Apollo's observations are consistent with several recent reports. PwC found that most chief executives had seen neither higher revenue nor lower costs from AI despite heavy investment. Deloitte research showed that adopting AI tools has not improved the bottom line for most companies. McKinsey said the corporate world is approaching a turning point toward returns on pilots and early deployments, but acknowledged that the share of businesses reporting an earnings boost from AI has remained flat.
Slok questioned whether mainstream corporations will convert AI projects into higher revenue growth and margins, even as cheaper and open-source models emerge. He argued that business decisions depend largely on judgment and context, and that AI does not grasp accountability, trust, culture, ethical trade-offs or fairness, which shape reputational risk and ultimately revenue and margins. Apollo also said the slowdown in token spending growth shows demand falls away or shifts to cheaper open-source alternatives when costs rise even modestly. That, it said, casts doubt on the economics of large datacenter investment if the resulting compute is likely to earn only marginal profits.
Apollo's broader role in the chip sector was also noted: the firm previously invested $11 billion for a 49 percent stake in Intel's Fab 34 manufacturing plant at Leixlip, Ireland, and Intel repurchased that stake earlier in the year for $14.2 billion.
Apollo's chief economist, Torsten Slok, said revenue per worker appears to be rising for the so-called "Magnificent Seven" technology firms, which include the hyperscalers as well as Apple, Nvidia and Tesla. He said the same is not true for the rest of the S&P 500. Slok also said profit margins and AI-driven gains in corporate productivity growth show the same pattern, and he supported these claims with charts.
Apollo's assessment drew pushback. Some commentators on social media said they did not recognize the picture Slok painted. Others pointed to the PC industry, where productivity growth lagged widespread business adoption by a decade or more.
The Register noted that Apollo's observations are consistent with several recent reports. PwC found that most chief executives had seen neither higher revenue nor lower costs from AI despite heavy investment. Deloitte research showed that adopting AI tools has not improved the bottom line for most companies. McKinsey said the corporate world is approaching a turning point toward returns on pilots and early deployments, but acknowledged that the share of businesses reporting an earnings boost from AI has remained flat.
Slok questioned whether mainstream corporations will convert AI projects into higher revenue growth and margins, even as cheaper and open-source models emerge. He argued that business decisions depend largely on judgment and context, and that AI does not grasp accountability, trust, culture, ethical trade-offs or fairness, which shape reputational risk and ultimately revenue and margins. Apollo also said the slowdown in token spending growth shows demand falls away or shifts to cheaper open-source alternatives when costs rise even modestly. That, it said, casts doubt on the economics of large datacenter investment if the resulting compute is likely to earn only marginal profits.
Apollo's broader role in the chip sector was also noted: the firm previously invested $11 billion for a 49 percent stake in Intel's Fab 34 manufacturing plant at Leixlip, Ireland, and Intel repurchased that stake earlier in the year for $14.2 billion.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.