Does the AI Economy Makes Sense?

By 2027, big tech companies are expected to spend a staggering $1 trillion on new data centers worldwide—an amount that equals more than 3% of the US GDP. On top of this, these companies will collectively be in $1.6 trillion of debt, all in pursuit of AI dominance. But is this massive investment sustainable, or are we witnessing a modern-day gold rush with uncertain returns?

Understanding the AI Ecosystem

The AI economy is a complex web, stretching from the bottom of the supply chain all the way to the end user. It starts with the companies building the equipment for chip manufacturing—think ASML and similar firms. Then come the chip designers and makers like Nvidia, Intel, AMD, and ARM. Supporting these are memory and storage manufacturers such as SanDisk, Western Digital, and SK Hynix.

Next in the chain are software creators who develop the AI models, including pure model makers like OpenAI, Anthropic, and xAI, as well as cloud providers like Google, Amazon, and Microsoft who host these models. Finally, the AI ecosystem reaches consumers and businesses who use AI to drive their operations.

The Ideal: A Sustainable System

A truly sustainable economic system is one in which every layer, from hardware to software to end user, can generate profits. Look back at the early 2010s: chip makers built chips, software companies used those chips to create valuable products, and customers purchased software to run their businesses. This created a virtuous cycle of innovation and profit at every level.

What’s Going Wrong Today

Today, however, tech giants are going all in on AI, pouring hundreds of billions into infrastructure and model training. While equipment and chip makers are seeing record orders, there’s growing doubt about whether there are enough end customers to make this ecosystem sustainable.

Many model makers are struggling: OpenAI’s revenue is growing, but expenses are outpacing revenue, and it has only 50 million paid subscribers. Anthropic may have its first profitable quarter, but details are unconfirmed. xAI posted a $6.4 billion loss last year. Meanwhile, the financial results of big tech companies are murky, complicated by circular investments where cloud providers invest in model makers who in turn buy cloud services from their investors.

For end users, adoption is rising, but mainly for tasks like writing, search, and general assistance—not for large-scale automation or transformative business use. Studies also show that the early promise of AI replacing entire workforces has not materialized; some companies that laid off staff for AI are now rehiring humans.

Conclusion: The Real Winners

Much like the California Gold Rush, the biggest winners in the AI boom are those selling the “shovels”—chip makers and cloud providers. Model makers have yet to prove they can be sustainably profitable, especially as open-weight models erode their competitive edge. Meanwhile, end users are integrating AI, but not at a scale that justifies the current spending frenzy.

AI isn’t going anywhere, but unless usage and profits catch up to investments, we may see spending slow to more realistic levels.

Plug

Support this free website by visiting my Amazon affiliate links. Any purchase you make will give me a cut without any extra cost to you

Other Ecosystem Items

• Apple TV (Amazon) / Apple TV 4K (Amazon)
• AirTags (Amazon)
• HomePod Mini (Amazon)

Love this? Consider buying me a coffee!