Federal Reserve Chair Jerome Powell has drawn a sharp line between the current artificial intelligence boom and the dotcom bubble of the early 2000s, arguing that today’s surge in AI spending is anchored in profitable, established businesses and tangible economic activity — not speculative mania.

“I won’t go into particular names,” Powell told reporters after the Fed’s policy meeting on Wednesday, “but they actually have earnings. These companies … actually have business models and profits and that kind of thing. So it’s really a different thing” from the dotcom bubble, he said.

According to Fortune, it was perhaps Powell’s most direct acknowledgment yet that artificial intelligence is now a fundamental driver of U.S. growth. Over the past two years, the AI industry has spurred hundreds of billions of dollars in data center construction, semiconductor manufacturing, and cloud infrastructure investment — a scale that rivals some of the largest industrial expansions in modern history.

A Structural Shift, Not a Monetary Bubble

Powell was clear that the AI boom isn’t being fueled by easy money or low interest rates. “I don’t think interest rates are an important part of the AI or data center story,” he said. “It’s based on longer-run assessments that this is an area where there’s going to be a lot of investment, and that’s going to drive higher productivity.”

That stance runs counter to one of Wall Street’s popular narratives — that loose financial conditions could be inflating another tech bubble. Instead, Powell described a structural transformation: a long-term bet by corporations that AI will reshape how work is done.

Major tech companies are spending at a historic pace to seize that future. Nvidia, whose chips power much of the global AI infrastructure, is on track to generate over half a trillion dollars in annual revenue. Microsoft and Alphabet have committed hundreds of billions of dollars in capital expenditure for new data centers, networking systems, and AI research. But unlike the speculative surge of the late 1990s, Powell noted, these firms are already profitable and reinvesting cash flow — not borrowing aggressively — to fund the buildout.

Goldman Sachs has backed that interpretation. In a recent note titled “The AI Spending Boom Is Not Too Big,” the bank’s chief U.S. economist Joseph Briggs argued that “anticipated investment levels are sustainable, although the ultimate AI winners remain less clear.” The Goldman team estimated that AI’s productivity potential could add between $8 trillion and $19 trillion to the U.S. economy in present value terms, depending on how quickly the technology spreads across industries.

“We are not concerned about the total amount of AI investment,” Briggs wrote. “AI investment as a share of U.S. GDP is smaller today — less than 1% — than in prior large technology cycles, which peaked between 2% and 5%.” In other words, despite the hype, the AI buildout is still in its early innings.

Grounded in the Real Economy

Powell pointed out that AI’s footprint is already visible beyond balance sheets and earnings reports.

“It’s the investment we’re getting in equipment and all those things that go into creating data centers and feeding the AI,” he said. “It’s clearly one of the big sources of growth in the economy.”

His comments align with private-sector forecasts suggesting that AI infrastructure spending could soon rival the economic boost once delivered by the U.S. shale revolution. JPMorgan economists recently estimated that the buildout could add roughly 0.2 percentage points to annual GDP growth over the next year — a meaningful lift for an economy of America’s scale.

The AI-driven investment wave has also reshaped industrial and energy dynamics. Utilities across the U.S. are racing to expand grid capacity as power demand surges from data centers. Energy executives have described the current load growth as the steepest in decades. Meanwhile, construction firms are reporting record backlogs for AI-related projects, from silicon fabrication plants to cooling systems and fiber networks.

In short, Powell is talking about cranes, concrete, and capital goods — not just code.

Although being optimistic, Powell warned that the long-term impact of AI on productivity and jobs remains uncertain.

“I don’t know how those investments will work out,” he said, acknowledging that it’s too early to tell how much of that productivity story will come through.

Economists have cautioned that the benefits of AI are likely to be unevenly concentrated among a small number of large firms that can afford the technology, while many others lag behind. Moreover, the short-term effect of automation could be disinflationary for wages and potentially negative for employment.

Powell acknowledged that tension, noting that a lot of companies that are making layoff announcements are talking about AI and what it can do, which underscores the paradox that the same technology boosting output might also slow job creation.

The Fed chief also referenced recent labor market adjustments, observing that job growth, once adjusted for statistical overcounting, is now pretty close to zero.

The Fed’s Balancing Act

Powell’s comments highlight a broader challenge of balancing the near-term slowdown in job growth against the long-term promise of an AI-led productivity surge. The Fed’s dual mandate, maximum employment and stable prices, could be complicated by the uneven rollout of a technology that simultaneously drives efficiency and disrupts labor demand.

Yet Powell’s remarks show that the Fed sees the AI boom as a real, durable driver of investment and growth — not a speculative mirage.

With corporate America pouring unprecedented sums into AI infrastructure and applications, Powell’s distinction between bubble and transformation may prove critical in shaping how both investors and policymakers navigate what could be the defining economic trend of the decade.

Nvidia CEO Jensen Huang has dismissed national security fears surrounding the sale of advanced semiconductors to China, insisting that collaboration between the two superpowers is not only safe but essential for the growth of artificial intelligence and global technological advancement.

Speaking to reporters in Seoul, South Korea, on Friday, Huang emphasized that China remains a crucial part of the global semiconductor market, calling it a “singular, vital, important, dynamic market” that no other economy can replace.

“The way to think about the China market is, it’s a singular, vital, important, dynamic market, and nobody can replace that,” he said. “It’s in the best interest of America to serve that China market. It’s in the best interest of China to have the American technology company bring … technology to the China market … It’s in the best interest of both countries, and I hope that policymakers will ultimately come to that conclusion.”

Huang, whose company sits at the center of the world’s artificial intelligence revolution, is navigating an increasingly tense geopolitical environment defined by Washington’s efforts to curb China’s access to advanced computing technologies. The United States has repeatedly argued that such technologies could be used to enhance Beijing’s military capabilities, leading to a series of export restrictions on Nvidia’s most powerful AI chips.

President Donald Trump confirmed Thursday that he and Chinese President Xi Jinping discussed semiconductor exports during their meeting in South Korea. While the leaders did touch on Nvidia’s operations in China, Trump clarified that their talks did not cover the company’s most advanced Blackwell graphics processing units, which are currently restricted under U.S. export rules.

“I said that’s really between [China] and Nvidia, but we’re sort of the arbitrator,” Trump told reporters aboard Air Force One after the meeting.

Washington has maintained that the restrictions are aimed at limiting both China’s access to U.S.-designed chips and its capacity to develop homegrown equivalents capable of supporting AI-driven applications. But Huang suggested that these measures may no longer be relevant, pointing out that China has made significant progress in semiconductor development and no longer depends solely on U.S. technology.

“China makes plenty of AI chips themselves, and the Chinese military surely have plenty of access to chips that are created in China,” Huang said in an interview with CNBC’s Eunice Yoon. “So, whatever national security concerns, have to take into consideration the fact that China has blocked H20 [an Nvidia chip] and, so, in a lot of ways, China is saying that, ‘listen, we have plenty of AI technology ourselves’.”

He added that the logic behind national security concerns has weakened, noting that Beijing’s decision to restrict purchases of Nvidia’s own chips demonstrates that China is confident in its domestic capabilities.

“The national security concern, from that perspective, I think, is really answered by the fact that China doesn’t want H20 or any American chips,” Huang said.

The H20 is a modified AI chip Nvidia designed specifically for the Chinese market to comply with U.S. export controls. While the chip was intended to balance performance with regulatory compliance, Chinese authorities have reportedly discouraged its purchase, favoring locally produced chips from domestic firms such as Huawei Technologies.

Respect for Huawei’s Technological Strength

In his remarks, Huang also addressed the rapid rise of Huawei, describing it as a formidable competitor with deep technological expertise. Despite the U.S. ban on federal use of Huawei products and restrictions preventing American companies from doing business with the telecom giant, Huang said it would be “foolish” to underestimate the company’s innovation and resilience.

“It is foolish to underestimate the might of China and the incredible, competitive spirit of Huawei,” he said. “This is a company with extraordinary technology. They dominate the world’s 5G telecommunication standards and technology. They build amazing smartphones, they build amazing chips, they’re incredible at networking and so when they announced CloudMatrix, I was not surprised that they were able to create such an amazing thing.”

Huawei’s CloudMatrix supercomputing system, unveiled last month, is designed to rival Nvidia’s AI hardware and marks a major step in China’s push for self-reliance in semiconductor and cloud computing technologies. Huang said Nvidia’s leadership takes such competition seriously and uses it as motivation to innovate faster.

“It’s deeply uninformed to think that Huawei can’t build systems,” he said. “We take competition very seriously. We respect the competition, we respect deeply the capabilities of China. That’s why we run so fast, and that’s why we dedicate ourselves to inventing the future so we can get there before anybody else.”

Huang’s comments underline a growing divide between business leaders and policymakers on the future of U.S.-China technological cooperation. While Washington views the semiconductor industry through the lens of national security, companies like Nvidia, which derive as much as 20 percent of their revenue from China, continue to see the country as indispensable for innovation and growth.

Analysts say Huang’s stance highlights a pragmatic reality: despite trade restrictions, China remains the world’s largest market for computing hardware, and access to that market could shape the trajectory of AI development globally.

Hello,

Greetings! We are happy to announce that the next edition of Tekedia AI Technical Lab is scheduled to begin on Saturday, November 15, 2025, and run for four Saturdays. The program webpage is here with pricing, syllabus and other details provided.

Co-learning with us, we will provide guidance on how to set up environment to build AI agents on your local computer and virtual web servers. When you are done, AI agents will be working on your machine or virtual private server.

Tekedia AI lab is not a coding or programming program; our goal is to help our co-learners connect pieces and components required to create AI agents, using DeepSeek, Google Gemma, Ollama, etc foundational models in both local and web servers, and not to teach programming. Basic secondary school education is the minimum requirement. We will provide required codes and sub-VPS location to co-learners.

The cost is $500 or N350,000 and you can pay at the program website here. This also covers bonus access to Tekedia AI in Business Masterclass. We support Naira bank transfer, PayPal, Stripe, Zelle, etc. Classes will begin on Saturday, Nov 15, 2025.

Upon completion, we award Advanced Diploma in AI Technical Design and Deployment, and Advanced Diploma in Artificial Intelligence (AI) in Business certificates. Join us.

Meanwhile, we have opened registration for the 19th edition of Tekedia Mini-MBA  (Feb 9 – May 2, 2026) here.

Regards,

Tekedia Institute Team