Apollo Economist Says AI Has Yet to Pay Off for Most Companies

Highlights

Apollo Global Management chief economist Torsten Slok says major companies outside the technology sector have yet to see measurable financial benefits from artificial intelligence. His analysis examines the “S&P 493,” defined as the S&P 500 excluding Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla. He finds that profit margins and revenue per worker appear flat, with no clear sign of improved productivity. Slok questions whether cheaper machine intelligence can translate into stronger growth, emphasizing that business outcomes depend on judgment and human considerations. The central point is that widespread AI investment has not yet produced a demonstrated financial payoff for corporate America outside tech. Other studies also indicate that significant productivity gains remain elusive.

Sentiment Analysis

• Overall tone: Cautious and skeptical. The article presents AI’s business impact as unproven, particularly among large companies outside the technology sector.


• Basis for the assessment: Slok’s analysis points to apparently flat profit margins and revenue per worker within the S&P 493, and the article notes that recent studies have not established significant productivity gains.


• Positive dimension: The article recognizes that AI may offer useful computational capabilities and that substantial investment and interest in the technology could support future applications.


• Concerns: It stresses that business success involves judgment, accountability, trust, culture, ethics and fairness—factors that may not be captured by treating AI as inexpensive intelligence alone.


• Assessment: The outlook is mixed but leans negative regarding near-term financial returns. The argument does not establish that AI can never create value; rather, it says current evidence has not shown that value in broad corporate financial measures outside tech.

Article Text

Artificial intelligence has attracted substantial investment and public attention, but its financial effects have not yet appeared clearly in the performance of large companies outside the technology industry, according to Torsten Slok, chief economist at Apollo Global Management. In a recently published note, Slok argues that the available indicators do not show a broad improvement in corporate results attributable to AI. His assessment challenges the assumption that rapid adoption and widespread enthusiasm have already translated into measurable business gains.

Slok’s analysis focuses on what he calls the “S&P 493.” The term refers to the S&P 500 after excluding Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla. These companies are closely associated with the technology sector and with the development or deployment of AI. By examining the remaining companies, the analysis aims to consider whether businesses outside the major technology names are benefiting financially from the technology.

According to Slok, profit margins and revenue per worker among this group appear to be flat. Those measures are relevant to the debate because companies often expect new technologies to help employees produce more, increase sales, reduce costs or improve profitability. If AI were already generating broad financial benefits, changes in productivity-related measures or margins might provide evidence of that effect. Slok says the indicators he reviewed do not currently show such a pattern.

He summarized his conclusion by saying that, despite large investments and widespread excitement about artificial intelligence, the technology had delivered “zero financial payoff” to corporate America outside the technology sector. This is a statement about the observed financial evidence in the analysis, rather than a claim that no company has used AI or that the technology has no practical capabilities. The distinction matters: adoption and investment can occur before their effects become visible in company-wide financial results.

A key question is whether non-tech companies simply need more time to catch up with technology leaders. Slok expresses skepticism about that explanation. He argues that making intelligence increasingly inexpensive may not, by itself, lead companies to achieve faster revenue growth or higher profit margins. Businesses must integrate tools into real operations, and the existence of a capable system does not automatically mean that it can improve the decisions and relationships on which commercial outcomes depend.

Slok’s argument distinguishes computational ability from business judgment. He describes AI as potentially a “great calculator,” while noting that intelligence in business often involves judgment. In his account, commercial decisions are shaped not only by processing information but also by responsibility, organizational norms and circumstances that may be difficult to encode. The article’s central distinction is between AI’s ability to perform tasks and the broader human judgment needed to turn those tasks into reliable business value.

He also points to accountability, trust, culture, unwritten rules, ethical trade-offs and fairness. These considerations can influence how an organization manages risk and how customers, workers and other stakeholders respond to its actions. Slok links them to reputational risk and, ultimately, to revenue growth and profit margins. His position is that AI tools do not automatically understand or manage these dimensions simply because they can process information or generate outputs.

The discussion does not establish that AI will never produce productivity gains or financial returns. Instead, it highlights a gap between high expectations and the evidence cited so far. The article notes that recent studies have also found that AI has yet to deliver significant productivity gains. Taken together, the observations support a cautious reading: companies may be investing in the technology, but measurable improvements across broad financial and productivity indicators remain limited or unproven in the evidence described.

For businesses, the implications are analytical rather than definitive. Evaluating AI’s impact requires looking beyond the scale of investment, the number of tools adopted or the enthusiasm surrounding them. Companies and researchers must assess whether AI changes work output, costs, sales, margins and decision quality, while accounting for organizational and human factors. Slok’s analysis suggests that the financial payoff should be judged by demonstrated outcomes, not assumed from the technology’s potential or its prominence in corporate strategy.

Key Insights Table