Is the AI Boom a Bubble? Tech Leaders Say No

The rapid expansion of artificial intelligence has prompted warnings from some that the industry might be in a speculative “bubble” that could burst. But several technology leaders argue the current AI boom differs fundamentally from the dot‑com era of the late 1990s and early 2000s, and therefore is less likely to collapse in the same way. Nvidia senior vice‑president of networking Kevin Deierling noted a key distinction: during the internet boom, much of the infrastructure and many use cases still had to be developed from scratch, leaving a gap between investment in capacity and actual consumer demand.

“Back then there was a lag,” Deierling said at a Toronto event on the sidelines of the Cisco Connect conference. “You could build bandwidth for the internet, but you still needed Amazon, Uber, Netflix and so many other services to make use of that capacity.”

AI tech is ready, no wait required

Unlike the internet build‑out, Deierling argued AI does not require decades for demand to catch up with supply. He said software applications that leverage AI are already widely available, allowing organizations to put new infrastructure to immediate use.

“In the dot‑com era, by the late 1990s and early 2000s you saw inventory build up and companies shipping products that weren’t selling through,” he explained in an interview. “That’s not what we’re seeing today. AI capabilities get used as soon as they’re deployed.”

Deierling’s optimistic comments preceded Nvidia’s latest quarterly report, released Wednesday, which may have eased some market concerns. The company reported net income of US$31.9 billion for the third quarter, up from US$19.3 billion a year earlier, while revenue increased 62%. Sales of Nvidia’s graphics processing units—critical hardware for training large AI models that power services like advanced chat assistants and image generators—surpassed analysts’ expectations.

Nvidia, once the largest stock on Wall Street and briefly valued at around US$5 trillion, has experienced volatility this month, losing more than 10% on the S&P 500 as of Tuesday. By late Thursday morning the share price was trading about 6% higher. Analysts are watching Nvidia closely because many companies depend on its chips to accelerate their AI initiatives, and the company’s performance can signal broader industry momentum.

AI demand strong despite profit concerns

Even as AI‑related stocks have rallied for several years, concerns persist that the industry’s heavy investment may not translate into proportionate profits for every company involved. Executives at the conference pushed back against that pessimism, describing steady and genuine demand for AI solutions. Francois Chadwick, chief financial officer at Toronto‑based AI company Cohere, compared the market’s appetite to “a constant drumbeat.”

“There is a real need,” Chadwick said. “In the early internet days some firms built products nobody wanted. Today the demand is clear—companies, enterprises and governments are asking for these capabilities.”

Still, not every AI bet will succeed. Tom Gillis, Cisco’s senior vice‑president and general manager for infrastructure and security, cautioned that disruption of this scale inevitably produces winners and losers. “Someone will place a bet that doesn’t work out,” he said, emphasizing that failures are part of major technological shifts.

But Gillis was firm that AI’s practical value is already evident. “Is there going to be a retraction where people decide AI isn’t useful? Try your chat interface and you’ll see it’s highly valuable. That utility justifies substantial investment to drive the next wave of change.”

Canada strong in research, slow in deployment

A recent study found only 8% of Canadian organizations qualify as “AI‑ready.” According to the Cisco AI Readiness Index, nearly three‑quarters of Canadian respondents plan to deploy AI agents, and 34% expect those agents to complement employees within a year. Yet most lack the secure infrastructure to sustain broad adoption, and organizations that are fully prepared are about 50% more likely to capture measurable value from AI.

Deierling characterized Canada as “ahead on research and behind on deployment.” He said Canada possesses the expertise and fundamental capacity to adopt AI more broadly, and that the obstacle is largely a matter of will: priorities, investment and a readiness to pilot projects.

Acknowledging that many companies remain cautious, he recommended starting small—deploying AI first for internal use cases rather than betting the entire business on unfamiliar systems. “Every company is ready to use AI, even if they don’t realize it,” he said. “The initial risk is manageable. Once organizations deploy solutions and see the productivity gains, demand will accelerate and drive the next generation of adoption.”

Read more news:

• AI‑fuelled scams on the rise as holiday shopping begins
• Stock news for investors: Barrick leads earnings gains as major Canadian companies report mixed Q3 results
• Holiday job postings rise, but competition remains fierce