Most people are captivated by the huge amount of money raised in OpenAI's recent $110 billion funding announcement, but the truly important point lies elsewhere.

In this round, where Amazon invested $50 billion, NVIDIA $30 billion, and SoftBank $30 billion, the order in which Sam Altman expressed his thanks is quite interesting. It was Amazon, Microsoft, NVIDIA, then SoftBank, and hidden within this order is a strategic insight.

As pointed out by overseas blogger Aakash Gupta, the real key terms are "Stateless API" and "Stateful Runtime Environment." These are the technological concepts that are dividing the present from the future.

Stateless API is the current mainstream. It’s used when integrating AI into existing systems in finance, retail, healthcare, etc. It answers questions, summarizes documents, enhances search, and so on. For companies, it’s convenient because it allows adding AI capabilities without changing the organization. However, there’s a problem: as models become similar, computational costs decrease, and price competition intensifies, Stateless API becomes easier to commoditize, which compresses profit margins.

On the other hand, the "Stateful Runtime Environment" is still limited at a commercial scale, but it signifies a fundamental shift in business paradigms, not just a feature improvement. It can actually perform tasks as digital labor, not just answer questions. The budget expands from mere API call costs to automation, process management, and labor cost reduction. In other words, the market size for the Stateful Runtime Environment could far exceed current expectations.

What Microsoft and Amazon have acquired makes this picture clear.

Microsoft has secured the current Stateless API traffic through a $250 billion contract and exclusive service rights. All Stateless API calls to OpenAI go through Azure. It’s a highly predictable cash flow, but the profit margins for Stateless API are shrinking, which is a concern.

Amazon, with $50 billion in actual funds and a $100 billion expansion contract, has secured infrastructure hosting rights for the era of the Stateful Runtime Environment. If agents become central to enterprise productivity, then computing power, storage, scheduling, workflow integration—all will be accumulated within AWS.

They’ve secured current cash flow and bet on the future productivity structure. Their strategies are completely different.

But here’s the interesting part: from OpenAI’s perspective, thanks to this clear separation and profit guarantee in the cooperation agreements, their bargaining power has significantly increased. In the past, OpenAI depended heavily on Microsoft, holding 27% of the shares and managing the platform, which tilted negotiation power toward Microsoft. But Amazon’s entry has changed this dynamic.

OpenAI is adopting a typical diversified investment strategy. It doesn’t rely heavily on a single cloud provider, doesn’t entrust future growth to one side, and uses future business potential as leverage to negotiate better terms. Neither Microsoft nor Amazon can afford to give up OpenAI at this point. As long as both remain at the table, negotiation power will naturally revert to OpenAI.

I believe this is the core of this funding round.