Henry Blodget, a veteran of the dot-com boom and bust, sees striking parallels between the current AI boom and the internet frenzy of the late 1990s — but he doesn’t view that as a bad thing. In a recent conversation, Blodget argued that financial bubbles, while often overhyped and prone to collapse, can serve a vital role in driving innovation by channeling massive amounts of capital and attention into transformative technologies. Blodget, who was once a prominent Wall Street analyst before facing SEC charges over misleading stock recommendations — which he settled without admitting guilt — now runs his own media venture. His perspective is shaped by firsthand experience with both the rise and fall of a technological revolution. He believes the AI boom, like the dot-com era, is a kind of massive, market-driven R&D experiment. Billions, possibly trillions, are being poured into AI not because everyone believes every startup will succeed, but because the potential is so vast that even a fraction of success could reshape the world. He acknowledges that most AI ventures will fail, and that the current enthusiasm may be outpacing real-world progress. Adoption of AI tools like ChatGPT and Claude is strong, but many companies still struggle to integrate AI meaningfully into their operations — much like businesses in the late 1990s that built websites without clear strategies for using them. Valuations, meanwhile, are pricing in decades of sustained growth, which makes them vulnerable to setbacks — similar to how AOL’s model faltered when broadband emerged. Blodget warns that the current AI investment frenzy is distorting the broader economy. Major tech firms are spending hundreds of billions on data centers, fueling short-term growth. But unlike railroads or fiber-optic networks, which last decades, today’s AI infrastructure — especially chips from companies like Nvidia — becomes obsolete in just three to four years. This creates a risk of massive overinvestment followed by rapid devaluation. Still, Blodget remains confident that AI is real and here to stay. The key question isn’t whether AI will succeed, but where we are in the cycle. Are we in 1996, early in the wave? Or in late 1999, on the brink of a crash? He admits no one can be certain. For the average person, his advice is simple: don’t bet on individual AI stocks. Instead, stick to a balanced portfolio of index funds, cash, and bonds. The real winners, he suggests, won’t be the companies making headlines today, but the ones that emerge after the inevitable shakeout — much like Google, which didn’t dominate the 1990s but rose to power in the early 2000s by getting the technology right. The AI era, like the internet before it, may be messy, inflated, and full of hype — but it could also be the foundation of the next great leap in human progress.