This wave of AI enthusiasm is starting to feel a lot like the pre-internet bubble era. Many believe AI will boost efficiency and reduce costs, and therefore naturally benefit the economy. But reality may be the opposite. You’ll notice that global tech giants are spending money recklessly— building data centers, buying chips, securing power, expanding servers. Microsoft, Google, Amazon, Meta—each one is investing more aggressively than the last. AI hasn’t yet fully unleashed productivity gains, but the infrastructure costs have already been poured out. The problem is, none of these investments are free. Electricity is getting more expensive, computing power is rising in cost, and data center construction is becoming increasingly costly. If these expenses ultimately get passed on to the broader economy, inflation may not ease as easily as many assume. The most dangerous aspect of today’s market is this: Everyone believes in AI’s future, yet very few are calculating just how much money this frenzy will burn through. Over the past year, many AI-themed stocks on the U.S. market have seen wildly inflated gains. Some companies’ valuations have already priced in growth expectations for years ahead. This has happened with every major technological revolution in history— railroads, electrification, the internet. In the end, technology transforms the world—but along the way, someone always overstates the story and inflates prices beyond reason. So I’m increasingly convinced: An AI bubble doesn’t mean AI is fake. On the contrary— the things most likely to produce bubbles are often the truly great innovations. The difference lies only in this: Technology can ultimately deliver value. Stock prices? Not necessarily.