I'm currently observing an interesting trend: Every time the market drops 2-3%, we hear the same question everywhere— is this just a correction or the start of a real bubble? It reminds me of a story we should all know: the dot-com bubble of the late 1990s.

Back then, the internet was the hot topic; today, it's AI. But the patterns? They are damn similar.

Let me briefly go back: In the mid-1990s, the internet became accessible to ordinary people. Millions went online for the first time. Companies recognized the potential—and capital flow began. Venture capital firms competed to invest in any startup that had the word "Internet" in its name. A vicious cycle emerged: the more money flowed, the more founders appeared to grab it.

By 1998, it was no longer just enthusiasm—it was pure euphoria. The Nasdaq soared upward. IPOs doubled or tripled their prices on the first day. For investors, this seemed like the sure path to quick wealth. Companies with no revenue, no profits, often no real business model— they reached billion-dollar valuations just because they had ".com" in their name.

The media fueled it all. CNBC celebrated young entrepreneurs who went from dorm rooms to multimillionaires. Day trading became a national obsession. People abandoned diversification and focused everything on speculative tech stocks. Momentum and emotions drove trading, not fundamentals.

But here’s the key point: The numbers made no sense. Many of these companies burned capital at an alarming rate. Their business models constantly required new money for growth, marketing, infrastructure. Profitability? Decades away. Yet instead of warning, rising losses were interpreted as proof of "hypergrowth." The mindset was: size is everything, profits come later.

In early 2000, the environment changed. The Federal Reserve raised interest rates. Liquidity dried up. Established tech companies reported disappointing figures. Suddenly, something broke the aura of inevitability. Sentiment shifted from euphoria to doubt.

And then it happened: After the peak in March 2000, the Nasdaq lost almost 78% of its value in two years. Companies that embodied limitless possibilities lost their entire market capitalization in a few months. Thousands of startups went bankrupt. Office complexes in Silicon Valley emptied out. Trillions of wealth vaporized.

A fascinating detail: Cisco Systems was briefly the most valuable company in the world at that time. Its all-time high of $82 during the dot-com bubble was not reached again until December 2025—more than 25 years later. Cisco survived, but that shows how brutal the overvaluation was.

But here’s the most important part: Not everything disappeared. Amazon, eBay, and some others adapted their models. They focused on operational efficiency, real profitability, practical utility. These survivors showed us that transformative technologies can indeed endure—but only if built on solid fundamentals.

Now, today: AI is real. The demand for computing power is real. But valuations? They resemble the late 1990s. And I keep hearing the same phrase: "But this time, it’s different." Those were the same words during the dot-com bubble. "The internet’s impact is too revolutionary for traditional metrics." Now we say the same about AI.

The question is: How much of this enthusiasm is genuine long-term potential, and how much is speculative hype?

Nvidia is often compared to Cisco. But there’s an important difference: Nvidia generates massive cash flows, has pricing power, and benefits from real demand. That wasn’t always the case with Cisco during its peak. But even strong fundamentals can be overshadowed by extreme speculation.

The timeless lesson remains: Cash flow, sustainability, operational efficiency, and practical utility beat stories. Markets reward short-term rapid growth and good storytelling. But sustainable value only comes from companies that turn innovation into repeatable, profitable results.

Investor psychology hardly changes. FOMO, herd behavior, narrative distortions—they repeatedly push asset prices beyond reasonable limits. The dot-com bubble remains the prime example of how even world-changing technologies can experience world-changing corrections.

So: Next time someone says "this time it’s different"—think back to the 1990s. Discipline, skepticism, and focus on real business models aren’t sexy. But they are what truly matter.