OpenAI and the Economics of Superintelligence: How Artificial Reason Became the Most Expensive Infrastructure on the Planet
According to analysis from HSBC Global Research, OpenAI may require at least 207 billion dollars in new capital by 2030 to remain financially stable and sustain the growth of its computational infrastructure. The forecast, published at the end of November, presents a picture where the company’s revenue growth goes hand in hand with exponential growth in expenses—and this contradiction reflects the essence of the artificial intelligence era.
Analysts estimate that OpenAI’s revenue by 2030 may reach 515 billion dollars, yet expenses for leasing, building, and operating data centers will exceed 792 billion. The company has already signed contracts for 36 gigawatts of computing capacity, with a total estimated cost of 1.8 trillion dollars. This scale is comparable to the GDP of entire countries and the energy demands of an industrial power.
The economics of artificial intelligence are taking the shape of a “second industrial revolution”: instead of oil and steel—chips and energy-intensive server farms; instead of workers—algorithms, but with the same appetite for electricity and capital. Even under the most optimistic scenario, HSBC warns, OpenAI will continue subsidizing its users for at least another ten years, as the cost of a single computational request exceeds the revenue from a subscription or API usage.
This raises the question of a new monetization model for AI. Traditional SaaS formats cannot cope with colossal capital expenditures, and in the coming years, infrastructure partnerships and government co-financing—especially in energy and semiconductors—will become crucial. Leading players such as Microsoft, Nvidia, TSMC, and Amazon Web Services are already building a supply network around OpenAI to keep the project afloat.
Against this backdrop, a paradox emerges: OpenAI may be the first company in history whose “intelligence” is an asset rather than a product. But maintaining this asset turns out to be more expensive than any technology of the past century—from nuclear energy to space exploration.
