Contrary to the narrative of burning cash, major AI labs are likely highly profitable on the marginal cost of inference. Their massive reported losses stem from huge capital expenditures on training runs and R&D. This financial structure is more akin to an industrial manufacturer than a traditional software company, with high upfront costs and profitable unit economics.

The market is rewarding companies selling scarce AI resources (power, memory, GPUs) as they can raise prices and expand margins. Conversely, the hyperscalers buying this shortage face multiple compression as their capex soars and ROI on each dollar declines, creating a clear divide between winners and losers.

Unlike past infrastructure booms (railroads, fiber optics), the most costly part of the AI build-out is computer chips that become obsolete in 2-3 years. This creates immense pressure to generate revenue rapidly before the debt-financed hardware becomes worthless, a financial risk often passed to the public.

While immense value is being *created* for end-users via applications like ChatGPT, that value is primarily *accruing* to companies with deep moats in the infrastructure layer—namely hardware providers like NVIDIA and hyperscalers. The long-term defensibility of model-makers remains an open question.

Unlike typical tech cycles where suppliers and customers thrive together, the current AI boom sees semiconductor companies capturing value while their customers (hyperscalers, model builders) incur massive losses. This unsustainable dynamic suggests a future market correction.

The paradoxical financial state of AI labs: individual models can generate healthy gross margins from inference, but the parent company operates at a loss. This is due to the massive, exponentially increasing R&D costs required to train the next, more powerful model.

The market cap lost by software companies being disrupted by AI is not disappearing. It's rotating into investments for the underlying infrastructure—AI chips and data centers—that power the AI agents causing the disruption, effectively "feeding the beast."

Despite record profits driven by AI demand for High-Bandwidth Memory, chip makers are maintaining a "conservative investment approach" and not rapidly expanding capacity. This strategic restraint keeps prices for critical components high, maximizing their profitability and effectively controlling the pace of the entire AI hardware industry.

The AI infrastructure boom is a potential house of cards. A single dollar of end-user revenue paid to a company like OpenAI can become $8 of "seeming revenue" as it cascades through the value chain to Microsoft, CoreWeave, and NVIDIA, supporting an unsustainable $100 of equity market value.