Firms like OpenAI and Meta claim a compute shortage while also exploring selling compute capacity. This isn't a contradiction but a strategic evolution. They are buying all available supply to secure their own needs and then arbitraging the excess, effectively becoming smaller-scale cloud providers for AI.

While investors fear "Chipflation" (rapidly rising memory prices) could end the AI investment boom, the reality is more nuanced. Morgan Stanley argues higher costs will primarily reprice and ration access to AI infrastructure, favoring large hyperscalers, rather than halting the overall cycle.

AI companies with the foresight to sign long-term, multi-year compute contracts gain a significant margin advantage. They lock in prices based on past valuations, while competitors are forced to buy capacity at much higher current market rates driven up by the increasing value of new AI models.

To finance AI infrastructure without massive equity dilution, firms use debt collateralized by guaranteed, long-term purchase contracts from investment-grade customers. The rapidly depreciating GPUs are only secondary collateral, making the financing far less risky than it appears and debunking common criticisms about its speculative nature.

Historically, tech giants spent ~20% of operating cash flow on CapEx. The AI buildout has pushed this to ~100%, fundamentally transforming their financial models. This move from capital-light to capital-intensive means future growth requires external funding, a major shift.

A major shift in behavior among top AI labs is their move from three-year to five-year take-or-pay contracts for GPU infrastructure. They are locking in capacity at massive scale for longer durations, signaling extreme confidence in sustained, long-term demand for compute.

As AI models achieve human-level capabilities in valuable roles like software engineering, they can generate significantly more revenue from the same hardware. This increased monetization potential will cause the rental price of GPUs to skyrocket, potentially by over 15x, to match the economic value they produce.

Unlike the dot-com era where capital built unused "dark fiber," today's AI funding boom is different. Every dollar spent on GPUs is immediately consumed due to insatiable demand. This prevents a supply overhang, making the "circular funding" model more sustainable for now.

Despite reports of falling H100 spot rental prices, contract prices for sustained GPU workloads are rising. This indicates the market is shifting from short-term, experimental use to long-term, committed production deployments, reflecting stronger, not weaker, underlying demand for AI infrastructure.