The “Cancer” Eating Away at Artificial Intelligence Growth
Nearly three years after ChatGPT burst onto the scene—reaching 100 million users faster than any application in internet history—the Artificial Intelligence (AI) boom that once promised to revolutionize industries now faces a critical reckoning. Despite record-breaking hype, AI’s business model remains elusive, weighed down by soaring energy costs,
infrastructure bottlenecks, and investor fatigue. Mounting Fears of an AI Bubble The past week underscored growing instability across the AI sector. Shares of major AI-driven firms slumped, with Nvidia dropping 7% shortly after celebrating a historic $5 trillion valuation milestone. Meta Platforms lost 17% of its market value in just two weeks, while AI software firm Palantir fell nearly 8%, despite strong quarterly earnings. The Wall Street Journal recently described the AI surge as “increasingly fragile,” as concerns rise over whether the industry’s growth is sustainable—or simply another tech bubble in the making. SoftBank’s Sudden Exit Sparks Concern Market jitters deepened after SoftBank abruptly sold its entire stake in Nvidia, reaping $5.83 billion . Officially, the Japanese investment giant cited a desire to maintain a strong financial base. However, CNBC reported that the sale may be tied to SoftBank’s potential $22.5 billion investment in OpenAI and plans to acquire the robotics division of Swiss tech company ABB. This move revived memories of SoftBank’s earlier exit from Nvidia in 2019, when its Vision Fund cashed out a $4 billion stake. A Costly Race Without Profit Behind the market volatility lies a fundamental issue: the vast gap between AI investments and actual revenue . OpenAI, the company behind ChatGPT, reportedly plans to spend $1.4 trillion over the next eight years , yet its current annual revenue is just $20 billion . The company expects losses to skyrocket to $74 billion by 2028 , while CEO Sam Altman has publicly acknowledged investor unease over spending. OpenAI’s promises of new devices, cloud services, and robotics projects remain largely conceptual. Borrowed Billions to Power AI Dreams To fund the astronomical infrastructure costs of AI, tech giants are taking on record levels of debt. Oracle signed a $300 billion agreement with OpenAI in September to boost computing capacity, raising $18 billion through bond sales. According to Goldman Sachs , AI firms issued $139 billion in bonds between January and October 2025—up 23% from the previous year. Meta alone secured $27.3 billion in private financing for its new data center complex in Louisiana, one of the largest such deals ever made. The Energy Crunch AI’s relentless expansion also faces a physical constraint: a lack of electricity . The massive data centers that power AI systems require enormous amounts of energy, with operational costs far exceeding those of traditional internet services. To cope, tech giants are increasingly turning to nuclear energy . Microsoft has joined the World Nuclear Association , while Google and Amazon are investing in small modular reactor (SMR) technology to power their AI infrastructure. But until those projects become operational, energy shortages continue to cause delays and inefficiencies across AI services. Investor Doubts and Market Realities Prominent investors are starting to question the accounting practices behind AI profits. Michael Burry , the hedge fund manager famous for predicting the 2008 housing crash, recently shorted several AI-related stocks. He accused major cloud providers of overstating profits by extending the estimated lifespan of Nvidia chips used in their data centers—a move he described as “one of the most common forms of modern financial manipulation.” A Costly Future Despite mounting skepticism, the industry shows no sign of slowing down. AI leaders are projected to spend a combined $400 billion this year alone. The lingering question for investors and industry insiders alike remains: when—or if—these massive bets will ever pay off.
