No products in the cart.
Founder Interview – Stella S.
February 9, 2026
Seed Five
What We Can Learn From the 90’s
Jide Adebowale
An economic analyst and writer based in New York City. He spent six years working as an economist for the U.S. Bureau of Labor Statistics and now writes about urban life, regional economies, and the future of work. His work blends cultural commentary with a data-driven perspective.
Is the internet actually useful, or just a gimmick? Is it a legitimate business tool, or a money pit? And even if it is powerful, won’t it just wipe out millions of jobs?
Those questions might seem silly today, but they were genuine concerns in the 1990s through the early 21st century. Politicians asked them. CEOs asked them. Every day, Americans ask them.
If you replace “internet” with “AI,” you get the exact same questions being asked now. Whether AI is going to fulfill its sweeping promises is still debated. Hopping in a time machine and thinking about the trajectory of the internet offers a pretty good roadmap for what might be next.
Why might AI be a bubble? The logic is straightforward. Billions of dollars have poured into AI globally. Expectations are sky-high. At some point, those investments need a return, a return that is far from guaranteed. The sheer volume of capital thrown at AI is staggering. Over the next two years, Amazon, Google, Meta, and Microsoft alone are expected to spend $750 billion on AI. OpenAI, creator of the world-famous ChatGPT, reported a net loss of $9 billion dollars this year, an amount that they expect will keep increasing.
AI VALUATION BOOM
While the AI craze seems unprecedented, the Internet mania was nothing to sneeze at.
The late 90s internet boom that became known as the dot com bubble was eerily similar to the current moment: valuations soaring, companies going public without revenue, VCs throwing money at any startup with “.com” in the name. And that bubble busted. Companies evaporated. Investors got wiped out. Stock markets cratered. After the stock market peaked in March 2000, it took 15 years for the NASDAQ to reach those levels again.
The pattern of overinvestment in a new technology and a collapse under those expectations isn’t new. That could spell trouble for certain AI companies. However, that alone misses the main lesson from the dot-com boom. The hype might collapse, the stock market might drop, but the underlying technology may still win out in the long run, as it certainly did in the case of the internet.
Despite the dot-com crash wiping out fortunes and vaporizing entire companies, the internet still won the long game. Internet adoption grew slowly—and then exponentially. In 2000, roughly half of American adults said they used the internet. By 2025, that number reached 96%.
This is the quiet truth beneath many bubbles. If the underlying technology is powerful enough, it outlives the hype cycle. It survives the boom, the bust, the think-pieces, and the panic. It keeps expanding because people keep adopting it.
Technologies like the internet, and potentially AI, move through predictable stages. First come the early adopters. Then the hype. Then the real utility. Then, integration into the workplace. Then, the moment when a new generation grows up, never knowing life without it. Once schools use it, employers require it, and entire industries build on top of it, the technology stops being optional. It becomes infrastructure.
2005
Broadband Begins Growth
68% use the internet & 29% have home broadband.
2025
Current State
96% use the internet, 78% have home broadband, and 16% are smartphone-only.
AI may be headed down that same path. Long before ChatGPT, Gemini, or Claude, people were already living with low-level AI. Siri on their phones, autocorrect in their messages, automated customer-service systems, recommendation engines. Today, advanced models are becoming just as commonplace; plenty of people rely on these tools daily at work, school, and at home.
Even if the AI market faces a valuation correction or a slowdown in investment, the broader trajectory seems clear: the models will get better, the use cases will multiply, and the technology will seep into every layer of daily life. A bubble may burst, but the underlying movement doesn’t disappear.
Just like the internet, AI may very well be an unstoppable rocket ship. Even with delays, fumes, and turbulence along the way, it still looks like a vessel headed for the moon.
You may now be willing to accept the premise that even if AI is surrounded by hype, exaggeration, and too much venture cash, it still feels destined to be transformative, just as the internet was before it. But that raises the next big question: what happens to jobs?
If AI becomes as efficient as some claim, if AGI, Artificial General Intelligence, truly is “around the corner”, will there even be jobs left? Will companies simply be filled with AI instead of people? Who will buy anything? Will society need a universal basic income just to function?
All fair questions. And once again, the past gives us a blueprint.
The internet absolutely eliminated jobs. Take travel agents. Once upon a time, booking a vacation required calling a human being. Now you can plan an entire international trip from your phone while sitting on your couch. Travel agencies shrank, and yet, the travel industry exploded. Cheaper access + greater usability = a much bigger market overall. That pattern repeats across sectors.
There are countless non-obvious examples of industries that only exist because the internet existed first. Amazon started as an online bookstore, and the internet allowed it to scale into a platform selling every product imaginable. Then, years later, it morphed again, using its infrastructural advantage to become one of the world’s dominant cloud-computing providers. Entire divisions and hundreds of thousands of jobs emerged from a technological foundation that didn’t even exist a generation prior.
Will AI follow the exact same trajectory? Not necessarily, but the precedent is strong.
History suggests a simple truth: every major technological leap destroys certain jobs and then creates entirely new categories that were previously unimaginable. AI might shrink some roles, but it will almost certainly create new industries, new workflows, and new forms of work in its wake.
That’s the pattern, disruption followed by reinvention. AI is unlikely to be the exception.
One last reason not to dismiss AI as a bubble: we haven’t seen its final form yet. Realistically, we never will. Technologies of this magnitude don’t finish. They evolve. The internet of 2025 is unrecognizable compared to the dial-up era, and AI is on that same trajectory.
Only a few years ago, AI models struggled to produce realistic images. Now they can generate any face, object, or scenario you can imagine with eerie accuracy. Yes, AI models hallucinate, but that’s improving fast. And the arms race between ChatGPT, Gemini, Claude, and others is accelerating that improvement by the day.
Competition matters. Competition forces innovation. It forces efficiency. It forces companies to build tools that are actually useful, not just flashy. And because the stakes are so high, the pressure is relentless. Billions of dollars, global prestige, and technological dominance are on the line. When the incentives are that powerful, progress tends to compound. That same competitive foundation helped the internet succeed, and it may be the exact thing that propels AI into its next chapter.
It’s almost guaranteed that not every AI company riding today’s hype wave will survive. They shouldn’t. They’re not supposed to. Only the best use cases deserve to endure, just as only the strongest internet companies survived the dot-com collapse. The core internet functions that truly mattered, the ones that connected people and powered businesses, became so indispensable that we now keep the entire network in our pockets. AI is in the early stages of that same exploration.
How will it be used?
How will it reshape industries?
How will it transform daily life?
We don’t know yet, but the uncertainty is the point.
If history is any guide, we are not living through the end of a bubble; we are living through the 1990s all over again. We’re in the early days of a messy, chaotic, thrilling technology that will change how the world works, and alter the trajectory of history forever.
Enter the Lounge
Subscribe to the newsletter everyone in Manhattan is talking about. No filler. No fluff. Just what matters, when it matters.
