What I Learned from the Dot-Com Bust
That Applies to AI Today

I was there in the Bay Area during what felt like the real start of the tech boom for many of us on the ground. Right out of high school in the 1980s, I went to work for Fairchild Semiconductor, manufacturing silicon wafers. Tech was already deeply embedded in the Bay Area environment—early semiconductor work, transistors turning into integrated circuits, companies like Fairchild laying the foundation for what would become Silicon Valley. For us insiders, it was the beginning: the excitement of building the hardware that would power everything to come. While the rest of the world barely noticed, we were in the thick of it, sensing something big was underway.

History tends to spotlight the late 1990s “dot-com boom”—the period from roughly 1995 to the Nasdaq peak on March 10, 2000, followed by the brutal crash. That wave gets the headlines: the internet startup frenzy, wild IPOs, and the bust that erased trillions. But for those of us who lived it from the silicon roots in the 1980s, the boom felt continuous—a steady build from chips to computers to the web.

Later in the 1990s, I moved into the emerging world of the internet, designing web pages and sites under the nickname “gooey.” While coders handled the backend, I focused on the front-end experience. One of our early innovations came when we embedded gradients into website backgrounds—a technique that was considered impossible at the time because of painfully slow baud-rate dial-up connections. Full background images were too large to load quickly, so we sliced the gradient into a tiny two-pixel-wide vertical strip, embedded the gradient into that small file, and set it to repeat horizontally across the screen. The result was a seamless, visually rich background that loaded fast and looked modern. It was a small technical win, but it taught me how creative problem-solving could make the impossible feel effortless.

That experience led me into bigger projects: helping startups build their brands during the height of the dot-com rush. I watched new companies emerge almost overnight—some with solid ideas, others riding pure hype. I worked closely with one called Global VR, helping it navigate the chaos and become a sustainable company that’s still around today. Being part of that era gave me a front-row seat to both the creativity and the recklessness.

Now, in 2026, with AI dominating conversations, it feels familiar—but accelerated and more intense. The pace is relentless: new models, tools, and announcements drop almost daily. Tech hubs are buzzing again with that sense of being on the bleeding edge. The possibilities are real and significant, but so is the pressure to adopt everything immediately. Don’t fall for the new hype. I’ve seen what happens when excitement races ahead of practical reality.

The Familiar Frenzy: Speed Over Substance

In the dot-com days (especially the late ’90s peak), almost any idea with a website could attract funding. Companies burned through cash chasing growth, assuming the market would reward them eventually. When reality hit, the fallout was severe—massive index drops, years to recover, countless failures.

AI moves even faster today. Startups pivot in days, enterprises scramble to integrate “AI everywhere,” and investment pours in hoping to back the next big winner. What took months back then happens weekly now. That velocity makes it tough to distinguish real progress from overpromising. Many tools sound revolutionary but offer limited gains, require heavy customization, or come with hidden costs in data, reliability, or integration.

Small businesses bear the brunt. They’re bombarded with messages that they must embrace AI immediately—generative tools, automation, analytics—or risk obsolescence. Without careful vetting, they invest in solutions that underdeliver, create more work than they save, or expose them to risks like data privacy issues or unpredictable outputs.

The Bubble Mentality: Superiority and Isolation

From the semiconductor years through the web-design boom, the Bay Area felt like its own world. Conversations revolved around breakthroughs; there was a growing sense of being far ahead. People developed an aura of superiority—dismissing non-tech sectors or outsiders who didn’t get it. I grew tired of that insularity and eventually left to escape the echo chamber. What seemed cutting-edge inside the Valley often felt abstract or irrelevant elsewhere.

AI brings back that dynamic, perhaps amplified. The “prima donnas”—as I call the most fervent tech enthusiasts with a smile—know exactly what I mean when I point it out; they grin because it’s part of the culture. Brilliant and driven, but sometimes caught up in hype that blinds them. Outside those circles, small business owners just want reliable tools without the drama or endless upgrades.

Practical Takeaways: Implement Wisely, Not Blindly

The work I did in the 1980s on silicon wafers, the gradient trick in the ’90s, and the brand-building for startups like Global VR weren’t total losses. They built foundations—steady improvements in hardware, user experience, and sustainable business models. AI has similar long-term potential: meaningful automation, better insights, reduced drudgery.

But rushing in blindly repeats old mistakes. The key lesson from those cycles: prioritize caution and fit over speed. Ask the tough questions:

• Does this address a specific, immediate need in my operations?
• Are the real costs—in time, money, training, or data—outweighed by tangible benefits?
• Can I pilot it modestly before scaling?

For small businesses, start simple: use AI for low-risk tasks like drafting emails, basic data sorting, or routine customer responses where results are easy to check and adjust. Avoid betting big on unproven “game-changers” until they’ve matured and proven consistent value over time.

Hype cycles come and go; grounded, practical application endures. The path from 1980s wafer fabs to 1990s web design to today’s AI boom shows tech evolves in waves, with bursts of excitement followed by corrections. AI’s promise is genuine, but so are its current limitations. Stay measured, evaluate honestly, and build incrementally. That’s the way to navigate the frenzy and come out stronger when things settle.