The AI sector right now mirrors the late 1990s internet boom almost to the letter. Back then, startups raised absurd amounts of capital on little more than a domain name and a pitch deck. Companies with no revenue—and in some cases no real product—were suddenly valued in the billions. The bubble burst hard in 2000–2001, wiping out most of those players. Pets.com is the poster child, but dozens of others vanished almost overnight.
Yet the internet wasn’t a fad. The survivors—Amazon, eBay, Google—redefined commerce, media, and communication. Infrastructure built during the boom (fiber optic cables, data centers, early hosting) became the backbone of the next two decades of innovation. The speculation phase laid down the rails that enabled the digital economy we take for granted today.
AI is in that same speculative stage. Every venture capitalist wants exposure, every startup calls itself “AI-first,” and overspending is rampant. The froth guarantees a crash. Weak players—those chasing hype without defensible IP or sustainable business models—will fold. But the underlying technology is real. Generative AI, multimodal models, and agent-based systems are as transformative as HTTP and broadband were in the ’90s.
The Shakeout Will Be Brutal but Productive
When the AI bubble contracts—and it will—it won’t be the end of AI. It will be the beginning of its real economy. Just as the dot-com bust paved the way for Web 2.0, the AI shakeout will clear the noise and leave behind infrastructure and platforms that fuel lasting growth. Think: cloud-hosted inference engines, data pipelines, trust frameworks, and domain-specific vertical models.
In other words, today’s “AI gold rush” is financing the future roads, bridges, and power plants of the AI economy. The investors writing reckless checks are subsidizing infrastructure that, in five years, incumbents and survivors will exploit for massive advantage.
What This Means for Business Leaders
- Don’t confuse hype for value. Most AI startups won’t make it. Focus on fundamentals—real customer problems, scalable delivery, and durable moats.
- Use the frenzy to your advantage. Talent, compute infrastructure, and datasets are all being developed at pace. Position yourself to capitalize once prices normalize.
- Think long game. Just as “dot-com” eventually became “the economy,” AI won’t stay in the novelty bucket. It will simply be business.