Top economist warns that the AI math doesn’t make sense: ‘Profits are currently being funded by investors rather than earned from customers’

Apollo Chief Economist Torsten Slok has issued a stark warning regarding the sustainability of the current artificial intelligence (AI) boom, suggesting that its apparent profitability is not rooted in genuine customer demand but is instead being artificially inflated by investor capital. In a blog post released on Friday, Slok detailed a significant and counterintuitive divergence in profit margins across the various segments of the AI value chain. This structure deviates sharply from traditional business models, where companies that successfully sell end products directly to consumers typically command the highest profit margins.

Slok meticulously broke down the AI ecosystem into four distinct categories: "models and applications," "cloud and compute," "energy and grid," and "silicon and equipment." To support his analysis, he utilized data compiled from Pitchbook and Bloomberg, incorporating financial information from prominent companies deeply involved in the AI landscape. These included generative AI pioneers like OpenAI and Anthropic, major cloud providers and AI integrators such as Microsoft and Amazon, energy infrastructure providers like Constellation Energy, and key hardware manufacturers including Nvidia, AMD, and Micron.

Slok's findings revealed a striking inversion of expected profitability. The "silicon and equipment" segment, which encompasses companies designing and manufacturing the foundational hardware like advanced semiconductors essential for AI processing, demonstrated the highest profit margin at an impressive 41%. This segment is home to chip giants like Nvidia and AMD, whose advanced GPUs are critical for training and running AI models. In stark contrast, the "models and applications" segment, which includes companies developing the AI software, algorithms, and user-facing applications, such as OpenAI and Anthropic, exhibited a deeply negative operating margin of -59%. This indicates that these entities are currently losing substantial amounts of money on their operations.

Slok attributes this pronounced disparity to the fundamental nature of the current AI investment landscape. He argues that the financial success observed within the AI sector is not being driven by organic, widespread demand for AI-powered products and services from end-users. Instead, the sector is being propelled by a surge of investment from shareholders who are eager to capitalize on the potential of what they perceive as the next transformative technological revolution. Slok explicitly articulated this concern, stating, "AI boom’s profits are currently being funded by investors rather than earned from customers." He elaborated that while the profit margins in the upstream segments (like silicon and equipment) are indeed real, they are being financed by the substantial capital being raised by the loss-making layers (models and applications), rather than by sustainable cash flow generated from actual customer adoption and usage.

This situation is particularly concerning given the projected scale of AI investment. Goldman Sachs, a leading global financial institution, has forecast that investments in AI could surge beyond a staggering $1 trillion by the year 2026. However, despite this immense financial influx, the tangible economic impact of AI remains limited. There have been no significant, economy-wide improvements in overall productivity or broad-based growth in profit margins observed across the general corporate landscape, with the exception of a handful of large-cap technology companies often referred to as the "Magnificent Seven." The current lopsided profit margin structure, where upstream suppliers profit from the investment in downstream developers who are not yet generating revenue, poses a significant risk of market instability. Should the flow of investor financing begin to decelerate or dry up, the entire AI ecosystem could face a severe correction, potentially leading to a market "toppl" as Slok alluded to, with significant implications for the broader economy.