Torsten Slok, the chief economist at Apollo, has sparked a wave of concern across Wall Street by arguing that the artificial intelligence boom is built on a fundamental economic contradiction. According to Slok, the current surge in investment ignores a simple reality: the cost of building massive data centers and generating computing power is currently outweighing the actual financial benefits these tools bring to businesses. He suggests that while investors have bet on unlimited demand for AI infrastructure, most companies are finding that AI acts more like an expensive cost center than a catalyst for growth.
The warning comes as market sentiment shifts away from AI chips and hardware, leaving traders questioning whether the industry can sustain its staggering spending habits. With some giants planning expenditures reaching toward a trillion dollars annually, Slok points out a troubling circularity in how this growth is funded. He argues that many AI startups aren’t paying for their computing needs through real profits, but are instead recycling venture capital funds to keep the lights on, creating an illusion of demand that lacks organic stability.
Beyond the balance sheets, Slok believes there is a deeper conceptual failure in assuming AI can automatically drive corporate revenue. He contends that while AI excels as a sophisticated calculator, it cannot replace human judgment, which relies on nuance, ethics, and cultural context. Because these intangible elements are what truly drive business success and manage reputational risk, mainstream corporations may never find a reason to pay premium prices for increasingly commoditized AI models.
This skepticism is coinciding with visible cracks in the financial foundation of several key players. From widening credit default swaps at Meta to crashing valuations for neocloud providers like CoreWeave, the cost of borrowing is rising just as doubt grows. Adding to the anxiety, analysts from S&P Global warn that there may be hidden debts tied to interconnected financing deals between tech titans and smaller data center projects, suggesting that the true level of risk in the sector remains obscured from public view.