OpenAI, Nvidia, and the bubble debate: Lessons from 30 years of tech surges
Understanding which players will endure, and which may vanish, is key to navigating the current market frenzy.
By around 2005, it would become clear that the impact of the Internet on the economy was no greater than that of the fax machine. This prophecy was written in 1998 by renowned economist and Nobel laureate Paul Krugman. Even then, a heated debate raged globally, though not on social networks, which did not yet exist, about whether the Internet revolution had created a giant bubble. It took another two years for the stock market to burst, but any comparison between the Internet and the fax today seems, at best, amusing, and at worst, absurd. Both can transmit information quickly, but that is largely all they have in common.
Yet, thirty years since the Internet bubble began to inflate, it is difficult to ignore the similarities between it and the current AI bubble. The parallels are striking, but the question that preoccupies investors and analysts alike is: “What year are we in today?” We may only be in 1996, when investors celebrated the IPOs of the early Internet stars Netscape and Yahoo, a generation that younger readers today must look up on Wikipedia.
By the end of that year, Alan Greenspan, the revered chairman of the Federal Reserve, famously warned against “irrational exuberance.” Perhaps we are in 1998, at the peak of the stock market run, following Amazon’s IPO in 1997. The shares of what was then an online bookstore were issued at $18 and, a little over a year later, were already trading at $250. By the end of 1998, Henry Blodget, the most famous Internet analyst of the time, projected that Amazon’s stock would reach $400 within a year. It did, fueling the frenzied market surge throughout 1999. Or maybe we are already in 2000, when the Nasdaq index peaked at 5,048 points before plunging to 1,114 by October 2002.
The Explosion Will Come When No One Expects It
The good news is that we are probably still far from the “2000 moment.” The AI revolution, along with the stock market surges fueled by enthusiasm for it, is likely still in its early stages, somewhere between 1996 and 1998. This does not mean there will not be hiccups along the way. Technology developments and earnings multiples support this view: while multiples seem high, they are still below the triple-digit valuations reached by the largest Internet companies at their peak. The S&P 500’s multiple is historically high, but most companies have already reported strong financial results, and profitability is increasing.
The bad news is that a crash will come, of course, when no one expects it, just as in previous bubbles, when the last skeptic enters the market. Last week, investors were startled by billionaire Michael Burry’s billion-dollar short positions against Nvidia and Palantir. Burry famously bet against the real estate bubble before the 2008 subprime crisis, profiting greatly. Yet, it is often forgotten that he opened his major short position in 2005 and endured losses for three years before the housing market collapsed.
Claude’s Anthropic Chases Behind OpenAI
The more pressing question is not the exact timing of a potential crash, since market corrections are inevitable over time. For example, the Nasdaq took 15 years to return to its 2000 peak, though other indices recovered more quickly. The more important issue is which companies will dominate the AI revolution. Just as we now struggle to remember companies like AOL or Nortel from the Internet era, history may soon judge today’s AI leaders.
Investors eyeing Nvidia or the anticipated OpenAI IPO, which recent reports value at around $1 trillion, should remember the lessons of the Internet bubble. Cisco was the “Nvidia” of its day, providing infrastructure for the Internet boom, while Yahoo was the “OpenAI.” Google, founded in 1998 and going public in 2004, became a tech giant only after the Nasdaq slump. Cisco still exists today but is far from its former glory; Yahoo, AOL, and Nortel have largely disappeared. Meanwhile, Anthropic’s Claude is quietly catching up to OpenAI, which garners the most attention, and scrutiny.
According to a recent forecast by Anthropic, the company is expected to be profitable within three years, with projected profits of $17 billion on $70 billion in revenue by 2028. Data from last month suggests that Claude leads ChatGPT in enterprise usage, while ChatGPT remains more popular with home users. Microsoft, Amazon, and Apple have reinvented themselves, and to a large extent also the business models that allowed the Internet to become a real thing that changes the entire corporate world. If the Internet had remained just a means to communicate with friends or buy pet food, Krugman’s prophecy might have been correct.
