By: Nicholas Creel, Assistant Professor of Business Law at Georgia State College and University [Georgia College and State University] / Newsweek
Translation: Telegrafi.com
Artificial intelligence [AI] is heading for a market collapse. The math behind this conclusion is simple. Big Tech is expected to spend over $400 billion on AI infrastructure by 2025, a figure that is expected to grow to $2 trillion per year by 2030. This despite the fact that forecasts show that these companies are not expected to close the gap between revenues and spending anytime soon.
Wall Street likes to tell itself that this gap will gradually narrow through the steady adoption of this transformative technology. This narrative claims that companies will slowly integrate AI, that productivity will gradually improve, and within a few years its business case will become clear. The problem is that early adoption of AI is proving to be financially unsustainable, meaning it could take several years — or longer — for AI to prove itself to be a worthwhile investment for corporate America.
Meanwhile, the capital intensity of AI makes patience a luxury that investors cannot afford. When you spend hundreds of billions on infrastructure every year, you can’t just wait years for organic adoption. The operational costs themselves demand rapid monetization.
That this case could be a stock market bubble is reinforced when you realize that Microsoft-i, Amazon-i, Google and Meta they are not spending the money they have saved in savings accounts. Much of this investment depends on circumstantial financing, a feature of the bubble dot-com which greatly increases the possibility of financial failure in the chain - if any link shows weakness.
While it is often compared to this, a key difference between this market and the telecom boom of the early 2000s is the size and speed of spending. The Bubble dot-com led telecom companies to spend $121 billion a year at its peak in 2000. Today's spending on AI is much greater - and focused on a shorter period.
Also, unlike laying fiber optic cables, AI infrastructure depreciates quickly. GPUs [Graphics Processing Unit] become obsolete as new models require more computing power. Data centers require constant upgrades. Energy prices continue to rise. The countdown clock until AI companies need a new infusion of capital is not just ticking away — it’s accelerating.
The competitive dynamics among the big AI companies only make things worse. No major player can afford to slow down spending, because they risk being left behind if a rival makes any progress. Everyone is locked in an arms race, where slowing down means losing everything.
To avoid a stock market bubble, the gap between spending and revenue for AI companies needs to close quickly. But how?
The only realistic scenario where this could happen is if OpenAI, anthropic and Google achieve AI replacing large parts of high-wage jobs. This means that AI would have to eliminate tens of millions of knowledge worker jobs to capture that economic value for itself. Analysts who cheer for the success of AI are essentially cheering for mass unemployment in most middle-class professions.
The middle ground — where AI adoption is gradual and employees get by smoothly while investors earn reasonable returns — is a game of accounting. The numbers simply don’t allow it. When you spend $400 billion a year, you’re not betting on sustainable progress. You’re gambling on a historic and rapid transformation.
The bubble scenario is much simpler. If the technology doesn't deliver enough results soon, revenue growth will stall and investors will quickly realize that the math is off. The bubble will burst and investors will suffer a blow. The recent massive layoffs in the AI team Meta-s are perhaps an indicator that signals that AI companies may be realizing that the boom is ending.
History offers a grim warning of what is to come. The telecom crash from 2000 to 2002 wiped out hundreds of billions in market value as companies collapsed under the weight of capital spending. The AI collapse will be much bigger. More capital has been invested, more companies are involved, more market share has been capitalized on this story. The contagion will not stop at tech stocks. Spending on AI infrastructure is now such a large component of GDP growth that its collapse would spread to the entire economy, sending us into recession.
The AI boom has never been sustainable at this scale. The only question is: how many quarters will it take before the rest of Wall Street accepts this? /Telegraph/
