According to Futurism, OpenAI has signed a $250 billion rental agreement with Microsoft and a $38 billion contract with Amazon for data center capacity. HSBC analysis reveals the company will spend $620 billion annually just to power its AI models, despite only a third of the contracted 36 gigawatts coming online before 2030. The company needs to reach 3 billion ChatGPT users by 2030, but current weekly active users stand at around 800 million with growth already plateauing. Only 5% of users currently pay for subscriptions, despite paying subscribers representing 70% of annual recurring revenue. CEO Sam Altman became visibly angry when questioned about how a company with $13 billion in revenue can justify $1.4 trillion in spending commitments.
The Money Pit Problem
Here’s the thing about OpenAI‘s financial situation – it’s not just burning cash, it’s creating a financial black hole. We’re talking about commitments that make their current revenue look like pocket change. The company needs to grow from 800 million weekly users to 3 billion in just six years while somehow convincing way more people to actually pay for the service. And all this while competitors like Google’s Gemini 3 are gaining ground. It’s basically trying to scale Mount Everest while carrying a backpack full of bricks.
The Bigger Picture
What really worries me is how much of the broader economy might be riding on this AI boom. Experts are saying the AI bubble has grown so large that it’s propping up the entire US economy. Think about that for a second. If OpenAI and similar companies can’t turn these massive investments into actual profits, we could be looking at something much bigger than just a tech company failure. Nvidia’s recent stock performance despite strong earnings suggests investors are already getting nervous about AI valuations. And Nvidia is the one selling the picks and shovels – imagine being the company that has to actually find the gold.
The Data Center Trap
OpenAI is caught in what the Financial Times describes as a vicious cycle. Every new user requires more data center capacity, which costs more money, which requires more paying users, which requires more capacity. It’s like running on a treadmill that keeps speeding up. The company is committing to infrastructure spending that would make even the largest tech giants pause. HSBC’s suggestion that “less capacity would always be better than a liquidity crisis” seems like common sense, but OpenAI appears to be betting everything on unprecedented growth.
Can They Actually Pull This Off?
Look, I want to believe in the AI revolution as much as anyone. But the numbers here are staggering. We’re talking about needing productivity-driven economic growth that would “dwarf what is often seen as unreasonable spending.” That’s a lot of pressure to put on technology that’s still finding its footing in the real world. When even your own investors are asking tough questions about the math, and the CEO responds with “if you want to sell your shares, I’ll find you a buyer,” it suggests the tension is real. The next couple of years will tell us whether this is visionary thinking or the biggest tech bubble we’ve ever seen.
