We Are in an AI Bubble — And That’s Not All Bad

The Catchy Question

We’ve heard it before: from venture capitalists, senior executives, journalists. “Are we in an AI bubble?” Bret Taylor, board chair at OpenAI and CEO of Sierra, doesn’t dodge the question. In a recent interview, he answered yes — but with a caveat: that doesn’t mean the sky is falling. In fact, the bubble might be part and parcel of transformative change.

Bubble Parallels: Dot‑Com and AI

Taylor draws a direct line from today’s AI boom to the dot‑com bubble of the late 1990s. Back then, many companies were wildly overvalued and crashed spectacularly. Many others failed. And yet, all the while, much of what people believed about the promise of the internet turned out to be true.

He suggests that, similarly, with AI, there is an enormous opportunity. The transformative economic value, the shift in how businesses operate, how we integrate AI into products and daily life — these aren’t speculative fancy; they are very real potential. But yes, there will also be many losers. Investments, startups, projects that seemed promising but couldn’t deliver, or couldn’t scale, or misread the market, will inevitably collapse.

“Someone Is Going to Lose a Phenomenal Amount of Money”

This striking line, originally from OpenAI CEO Sam Altman, echoes throughout the current discourse on AI investment. Bret Taylor affirms Altman’s warning. Some will back projects that don’t survive; others will misjudge timing or technology. However, Taylor argues, this is not a reason to pull the brakes entirely. The potential upside is so large that even after the dust settles from the bubble, the gains will outweigh the losses.

It’s a stance that seems paradoxical on the surface — embracing the volatility of the moment while believing in the long-term trajectory. But Taylor is not alone in this thinking. He joins a chorus of tech leaders who believe that bubbles, while painful, often accompany the emergence of epoch-defining technologies.

Why the Bubble Might Be Necessary

Taylor’s stance — shared by many in the industry — is that bubbles aren’t always bad. They can accelerate innovation, push boundaries, mobilize capital, and force hard choices. New infrastructure, new business models, and new tools often come out of eras of over‑enthusiasm. Even if many speculators get burned, the long‑term winners — those with substance, with balanced risk, with capability — could reshape industries.

He also points out that excess tends to expose weak ideas. The burst of a bubble forces clearer judgment about what works, what doesn’t, and what is sustainable. In other words, some amount of chaos might be healthy. It separates the wheat from the chaff. Without the heat of hype, some transformative ideas might never gain the attention or backing they need to reach escape velocity.

What to Watch Out For

If Taylor’s prediction holds, a few outcomes appear likely. Many AI startups will flood the market. A large portion of them will struggle to survive. Some will shut down, some will be absorbed by larger firms, and a few will scale into enduring giants.

At the same time, the capital flowing into AI may not always be well-placed. Investors could prioritize buzzwords and superficial innovation over technical rigor and long-term viability. This chase for quick returns can result in overinvestment in hype and underinvestment in the fundamentals — engineering robustness, ethical guardrails, data efficiency, and real-world usability.

There is also rising pressure on the regulatory, ethical, and social fronts. As AI becomes more powerful, questions around misuse, bias, privacy, and labor disruption will intensify. Companies that ignore these challenges may build fast but fail under scrutiny.

Meanwhile, some players — those that combine technical excellence with real product-market fit and responsible deployment — will likely capture a disproportionate share of the long-term value.

A Balanced Take

Bret Taylor’s perspective is not one of doom, but of realism infused with optimism. A bubble doesn’t negate the promise of AI; it just intensifies the risk. His view suggests we should acknowledge both: the excitement and the overreach, the possibility of big breakthroughs, and the inevitability of failure for many.

In the end, what matters is who builds well, who adapts, who invests wisely, and who can survive the shake‑out. If history is any guide, even after the bubble pops, a new generation of value creators will emerge — leaner, smarter, and perhaps more grounded in what AI can actually do.