NVIDIA Corp. CEO Jensen Huang strongly pushed back against concerns about an artificial intelligence (AI) spending bubble, telling attendees at the World Economic Forum in Davos on Wednesday that massive AI investments represent the foundation of “the largest infrastructure buildout in human history.”
Speaking with BlackRock CEO Larry Fink, Huang outlined an industrial transformation creating immediate demand for skilled tradespeople like plumbers, electricians, construction workers, and steelworkers to build the data centers and chip factories necessary to power the AI revolution. Salaries for these positions, he said, have nearly doubled in some cases, with workers now commanding six-figure incomes.
“It’s wonderful that the jobs are related to tradecraft,” Huang said, noting the industry faces a “great shortage” of these workers. The observation echoed warnings from Ford Motor Co. CEO Jim Farley, who has estimated shortages of 600,000 workers in factories and a similar number in construction.
Huang directly challenged the bubble narrative that has gripped markets since mid-2025, following OpenAI’s underwhelming GPT-5 release and an MIT study showing 95% of generative AI pilots failing to generate returns. High-profile executives including Amazon.com Inc. founder Jeff Bezos, Goldman Sachs CEO David Solomon, and Microsoft Corp. CEO Satya Nadella have all raised concerns about unsustainable spending levels.
Huang argued such fears misdiagnose the situation. He described AI as a “five-layer cake” requiring foundational investments in energy, chips, cloud infrastructure, AI models, and applications to create tangible assets rather than speculative ventures.
As evidence, Huang pointed to surging demand for GPU rentals. “If you try to rent an NVIDIA GPU these days, it’s so incredibly hard, and the spot price of GPU rentals is going up,” he said, noting that even two-generation-old chips command premium prices. The scarcity demonstrates that established companies are reallocating research budgets. Pharmaceutical giant Eli Lilly and Company, for example, is shifting funds from wet labs to AI supercomputing rather than simply burning venture capital.
The NVIDIA chief also addressed broader employment anxieties, asserting that AI transforms job tasks rather than eliminate job purposes. Despite AI diffusing throughout the field over the past decade, the number of radiologists has increased. By handling scan analysis faster, AI allows doctors to focus on patient diagnosis and care, improving hospital throughput and driving more hiring.
The same dynamic applies to nursing, where AI tools help with charting and documentation, he said. Those tasks currently consume nearly half of nurses’ time despite a nationwide shortage of 5 million nurses.
Fink reframed Huang’s argument into a provocative question: “Are we investing enough?” He suggested current spending levels might be insufficient for broadening the global economy, particularly as 2025 saw over $100 billion in venture capital deployed worldwide, mostly to AI-native startups.
Still, not everyone shares Huang’s optimism. International Monetary Fund managing director Kristalina Georgieva warned that AI represents a “tsunami” hitting labor markets, with 40% of jobs “either enhanced or scrapped or changed quite significantly.” Even in well-prepared countries, she said, “We are not prepared enough.”
Growing concerns about AI investment, meanwhile, are surfacing across North America’s technology industry, as companies increasingly rely on borrowed money to fund ambitious AI infrastructure projects while racing to achieve technological advances.
S&P Global Ratings released two reports today examining the evolving financial landscape of AI development. In “Where Are AI Investment Risks Hiding?”, the credit rating agency explores the transition from equity to debt financing for AI projects and outlines warning signs that could indicate an impending market downturn. A companion report, “AI Infrastructure Buildout Weighs Credit Risks and Rewards”, assesses how massive infrastructure investments are affecting the creditworthiness of companies in the sector.
Both reports, published on RatingsDirect, reflect what analysts describe as a necessary skepticism toward the sustainability of current AI spending patterns, particularly as companies take on significant debt obligations to remain competitive in the rapidly evolving field.
“To this point, investors remain skeptical on the trajectory of spending in the AI Revolution this year as seen so far by the underperformance of major tech stocks year to date,” Wedbush Securities analyst Dan Ives said in a note to clients on Thursday. “We believe there is a disconnect between the recent treadmill tech stocks vs. the eye-popping AI enterprise spending intentions, massive tech cap-ex buildouts by Big Tech, and unprecedented global demand on the horizon.”
