What future of AI is prophesied by internet diva Mary Meeker, who didn't see the dot-com bubble coming

Mary Meeker, a famous technology analyst and venture capitalist, predicted a brilliant future for a new technology - the Internet - in the mid-90s. And, as we can now see, she was right. But there is a nuance: before coming to prosperity, the market of Internet companies in 2000 experienced the «dot-com crash», on which investors lost trillions of dollars. Now, after a multi-year hiatus, Meeker has published a detailed report on her view of the artificial intelligence technology market. From her point of view, AI technologies are developing at an «unprecedented pace» and their prospects are bright. It makes you wonder about analogies: are we not witnessing the blossoming of a new technological «bubble»?

Fortune magazine calls the Netscape IPO «the spark that ignited the Internet boom» in 1995. The company was only 16 months old and had yet to make a dime of profit. Netscape's stock was priced at $28, but the investor frenzy was such that trading couldn't open for two hours, the magazine said. When it did, the stock soared to $75 that day and closed at $58, giving the startup capitalization of nearly $3 billion. By the end of the year, the stock was trading at $174. 

«Netscape has mesmerized investors and captured America's imagination. More than any other company, it set the technological, social, and financial tone of the Internet age. Its founders, 24-year-old baby-faced Midwestern programmer Mark Andriessen and restless middle-aged technology pioneer Jim Clark - who desperately wanted to find gold again - inspired a generation of entrepreneurs to try to become tech millionaires,» Fortune described the situation.

Microsoft co-founder Bill Gates that same year called the Internet «the most important development since the invention of the IBM PC.»

High expectations of the new technology combined with a surplus of inexpensive money on the market have resulted in a real rush;

«Many investors were willing to invest at any valuation in any company, especially if it had «.com» in its name,» writes Finbold. Hence the name «dot-com bubble.»

New IPOs happened almost every day, shares instantly appreciated 3-4 times, creating new millionaires and billionaires. If we look at the numbers: in 1999, most of the 457 IPOs were related to Internet companies, followed by 91 in the first quarter of 2000 alone. As a result, the Nasdaq Composite stock market index rose 400% between 1995 and 2000, Finbold calculated. In 1990, the value of stocks of companies traded on the Nasdaq was 11% of those traded on the NYSE; in December 1999, it was 80%, wrote Goldman Sachs.

«Tech companies at the time were known for throwing expensive events called 'dot-com parties' to create excitement around a launch (or any other cause for celebration, really). Budgets of up to a million dollars a month were spent on extravagant parties in the San Francisco Bay Area,» Finbold describes the festival atmosphere of the times. 

According to Salon magazine, there were 15-20 entertainment events a week with very questionable PR-effectiveness: many partygoers just traveled from one to another, not really interested in who invited them and why, as long as drinks at the bar were free.

The hangover was heavy.

By March 2000, several factors - from the gradual tightening of monetary policy in the U.S. to the threat of recession in Japan - coincided to force the most cautious investors to lock in profits in dotcoms and start moving into more traditional stocks. The selling caused panic, and the Nasdaq index plunged precipitously.

The market for new IPOs froze, and by October 4, 2002, the Nasdaq Index had fallen to 1,139.90 points, which was 77% off its peak at the time. 

It was able to recover to new highs only after almost 15 years, by April 23, 2015, notes Goldman Sachs. Investor losses were huge: Finbold estimates that about $5 trillion of capitalization (approximately $9.3 trillion at today's prices) quickly «evaporated» from the market.

Dotcoms, many of which grew solely on investors' money because they spent huge sums on advertising and PR without having a solid business model, began to close one after another;

CNN lists a dozen major collapsed companies that were once valued in the hundreds of millions and even billions of dollars. Like for example Internet delivery service Webvan, which raised $375 million in its IPO in November 1999, with shares trading at around $30 and the company valued at $1.2 billion. It closed in July 2001, at which time its stock was worth 6 cents.

Finbold names four main reasons for the crash - the huge overvaluation of dotcoms on the wave of frenzy, the profligacy of startups that spent a lot and made almost no money, the glut of low-cost venture capital, and the media that fueled the bubble with bravura reports of ever-increasing success. These are all common causes, but every disaster, as they say, has a name. At least one of them is Mary Meeker.

Mary Meeker's star rose in the mid-90s when she became chief technology analyst at investment bank Morgan Stanley (which, incidentally, organized the legendary Netscape IPO that launched the Internet era), wrote Fortune magazine in May 2001, fresh off the back of the dot-com crash.

In 1996, Meeker published a huge and now legendary report with the simple title «About the Internet,» in which she detailed in 322 pages which companies could benefit from the new technology. «We have never before seen anything like such rapid growth and scale in the use of online services and the Internet,» she wrote.

«Morgan Stanley's Mary Meeker was no ordinary analyst. Called «Queen of the Net» by Barron's, lovingly described by The New Yorker, equated to Alan Greenspan and Warren Buffett as a market mover by the Wall Street Journal, Meeker was the undisputed diva of the Internet age. Technology companies begged her to cover their activities. Morgan Stanley paid her a stunning $15 million in 1999. Ordinary investors hounded her, demanding autographs. During the dot-com craze, Mary Meeker was by far the most important voice of the Internet and the idea that companies without profits could transform the world and go to the moon,» writes Fortune.

Meeker's phenomenon is that the Internet was a novelty, many people saw the future, but no one understood exactly how to evaluate dotcoms. And Meeker became such a benchmark, she moved away from valuation by classic metrics such as revenue, debt, profit, focusing more on the number of internet page views. However, as it turns out, when the flow of investor money dries up, you can't get enough of views.

In her defense, Meeker told Fortune that Morgan Stanley held only 8% of the dot-com era IPOs, and therefore, to hold her and the bank she worked for entirely responsible for the collapse is, at the very least, unfair.

The dot-com bubble wasn't the end of her career: Meeker worked at Morgan Stanley until 2010, then became a partner at venture capital fund Kleiner Perkins, and now heads her own investment firm, BOND, with a $5.9 billion portfolio and included, by the way, in Forbes' Midas List, a ranking of top tech investors.

She published her reports on internet trends through 2019, anticipating in them, among other things, the rise of Apple, Google and the digital economy, writes Forbes. And now, after a five-year hiatus, Meeker has decided to say a powerful word about the prospects for AI.

In its new 340-page report, titled as simply as the previous ones: «Trends - Artificial Intelligence,» the word «unprecedented» is repeated 64 times. 

«The data show that the speed and scale of change driven by AI is unprecedented,» she writes.

800 million ChatGPT users in 17 months is unprecedented. The rate at which AI costs are falling (99% in two years in terms of 1 million tokens) is also unprecedented. The $212 billion growth in capitalization of the six largest US tech companies in 2024 alone - guess which one? That's right, also unprecedented.

She believes that thanks to satellite internet like Starlink from SpaceX, which already has about 5 million users, about 2.4 billion more people will soon join the global network. And they will start not with browsers, to which we are accustomed, but immediately with the use of AI interface, which is able to understand and fulfill user requests in any language. The main benefit will go to those who own and manage such interfaces.

In the physical world, AI will take over areas such as cab driving, mineral exploration, defense technology and agriculture.

Just as the Internet is now woven into virtually every aspect of our lives (imagine what would happen to your work, recreation, communications, shopping  and more if it disappeared tomorrow) - in a decade or two, AI will similarly penetrate every aspect of human activity, Meeker predicts.

There is one dark spot in this benign picture: while the costs of building and training the latest AI models are rising, the prices for using them are, as noted, «unprecedentedly» falling. Especially thanks to the entry of Chinese companies into the race, which, unlike the same OpenAI, rely on open-source models to help new entrants to the competition offer their solutions quickly and inexpensively, Meeker notes.

The situation is somewhat reminiscent of the dot-com era, with the largest players spending billions of dollars from investors in the hope of capturing leadership and future profits. And how exactly the leadership will be seized and who will be able to seize it is still unclear.

Will AI model developers remain universal platforms? Or will they switch to highly specialized niches? Will 1-2 leaders be able to concentrate their core audience, monetizing it through subscriptions, advertising or digital services? For now, we can only observe;

«One thing is clear in the short term: the economics of universal LLMs increasingly resemble the trading of conventional commodities, but with a venture capital burn rate,» Meeker writes honestly. - As investors, we always assume that things can go wrong, but the exciting part is imagining what could go right. History has proven time and time again: optimism is one of the most profitable strategies over the long term.»

Very inspiring, but tell that to investors who got burned on dotcoms.

«It's a time of great gains and great capital losses,» she said in an interview with Axios. Meeker knows this better than anyone else.