Oracle risks 50 billion while Elon Musk lacks pocket money for xAI

Fifty billion dollars. This is not just a figure from an annual report, it's the budget of an average European state or the cost of several dozen aircraft carriers. This is exactly how much Oracle plans to raise in 2026 through debt issuance and stock sales.

Why? To simply keep running in this insane arms race called "infrastructure for artificial intelligence." If Oracle was previously perceived as a reliable but slightly inflexible giant from the database world, today Larry Ellison's company is transforming into the main architect of the digital metropolises of the future.

The need for computing power is growing exponentially, and those who don't manage to build their data centers now will find themselves on the sidelines of progress in two years.

The context here is extremely simple: cloud giants like Microsoft and Google have been spending billions for a long time, and Oracle has to raise the stakes to avoid losing market share. Building AI infrastructure is not just about buying Nvidia chips; it also involves colossal spending on energy, cooling, and logistics. Ellison understands that this is the moment of "all or nothing." The cloud computing market is overheated, but the demand for training large language models (LLMs) is so great that investors are willing to swallow Oracle's debt obligations hoping for future superprofits. It's a long game, where the entry ticket gets more expensive with each quarter.

Meanwhile, at another end of the technological front, Elon Musk once again confirms his reputation as a master of corporate acrobatics. According to Bloomberg, he is negotiating a merger of SpaceX and his neural network company xAI. This decision looks like an acknowledgment of an obvious fact: even the resources of the richest person on the planet and his numerous foundations are not enough to single-handedly feed the voracious AI. xAI, which created Grok, needs gigantic computing clusters, and SpaceX possesses not only money but also a unique engineering base, as well as the Starlink satellite network, which could become the perfect nervous system for a future global intelligence.

Why does SpaceX need this? Beyond the obvious technological synergy, the merger allows Musk to redistribute cash flows within his empire. SpaceX is a profitable and stable business with government contracts, while xAI is a venture project with high risks and enormous costs. Combining these entities creates a structure that is practically impossible to sink financially. However, for SpaceX investors, this is an alarming signal: their investments in rockets will now support Musk's AI ambitions, which so far have not generated direct profit. This is classic Elon—burning bridges and consolidating all resources for one grand goal, whether it's Mars or AGI.

We are witnessing the end of the era of "garage" innovation in AI. If in 2022 you could surprise the world with a smart chatbot assembled on the fly, then in 2026 without 50 billion dollars in your pocket you won't even get a place in the queue for graphics cards. The consolidation of resources in Oracle's hands and the merger of Musk's assets are symptoms of one disease: AI has become too expensive for private capital.

Now it's a game of states and state-corporations. The barrier to entry into the top league of AI has risen to the stratosphere, and those who cannot find the leverage for financing at this scale will inevitably become just customers of those who dared to risk everything to build the "hardware" foundation.

The bottom line: The AI industry has definitively entered the phase of capital-intensive gigantism. Will Musk manage to turn SpaceX into the world's leading AI company, or will Oracle overtake everyone thanks to unlimited credit leverage?