As artificial intelligence continues its meteoric rise, a familiar chorus of concern has emerged: Are we living through another dot-com bubble? The comparison is tempting soaring valuations, explosive hype, and a rush of companies trying to stake their claim in a transformative technology.
But this time, the story is different.
Speaking at the Cisco Connect conference in Toronto, Nvidia’s Kevin Deierling offered a perspective that deserves more attention. Unlike the internet boom of the late 1990s, which largely outpaced real-world use cases at the time, the AI revolution is riding on demand that already exists and is growing by the day.
Back then, infrastructure was built ahead of its purpose. Companies were scrambling to explain why they needed that much bandwidth or computing power. Everyone was waiting for the Amazons, Netflixes, and Ubers of the world to eventually materialize. And they did but not before the market corrected itself with a dramatic crash.
Today’s AI wave, by contrast, is being absorbed as quickly as it’s built.
Nvidia’s latest earnings back that up: US$31.9 billion in quarterly profit, revenue up 62%, and GPU sales soaring beyond analysts’ forecasts. These aren’t numbers of a market disconnected from reality they’re the output of infrastructure being deployed and used immediately.
Deierling put it plainly: “This stuff gets used as soon as it gets built.”
Even as Nvidia’s stock faces short-term turbulence losing over 10% earlier this month before rebounding the long-term demand remains undeniable. Every major sector, from finance to healthcare to government, is racing to adopt AI tools. Cohere’s CFO Francois Chadwick described this momentum as a “constant drumbeat,” a persistent push driven not by hype but by necessity.
And necessity is the key difference between this era and the dot-com frenzy.
In the early internet days, companies built products no one asked for. Today, organizations are practically begging for AI solutions even if most aren’t yet ready to deploy them. Cisco’s recent AI Readiness Index revealed that only 8% of Canadian companies are truly “AI-ready,” yet a staggering 74% plan to deploy AI agents soon.
The demand is real; the infrastructure simply needs to catch up.
Canada’s paradox strong research leadership but slow deployment highlights a broader truth: hesitation, not lack of capability, is the biggest barrier. Companies fear the unknown, but as Deierling emphasized, they don’t need to dive into the deep end. Starting small, focusing on internal productivity, and gradually scaling is the smarter path.
And in doing so, many organizations will realize the surprising truth: they’re more ready than they think.
The risk isn’t over-investing in AI it’s being left behind by those who embrace it early.
The dot-com crash was born from inflated expectations without real demand. Today’s AI boom is driven by companies urgently seeking productivity, efficiency, and competitive advantage. Unlike the internet era, AI’s most valuable use cases aren’t decades away. They’re here, and they’re multiplying.
If there’s a bubble, it isn’t in AI. It’s in the fear that holds companies back from using it.
This time, the future isn’t waiting to be invented it’s already here, and the world is scrambling to keep up.
