London
CNN
—
When French startup Mistral AI raised 105 million euros ($118 million) last month in one of the largest seed rounds in European history, it didn't have a working product. But Antoine Moiroux, a partner at Lightspeed Venture Partners, one of the startup's biggest backers, was undaunted.
“It may seem like a huge number,” he told CNN, but he said the company has big global ambitions and will need a lot of expensive computing power to pull it off. .
This explosive deal is fueled by enthusiasm surrounding the potential of “generative” artificial intelligence, which can create original text, images and other content in response to prompts from users, to generate tremendous returns for investors. Just one example of the excitement.
But some investors and industry observers are concerned that the funding frenzy is turning into a bubble, with money going to companies without profits, innovative products or the right expertise.
Emad Mostaq, founder and CEO of StabilityAI, a generative AI company whose investors include California-based Lightspeed, says the current wave of investment in AI companies is “the largest in history.” We predict that this will create a bubble.
“I call it the 'dot-i' bubble, and it hasn't even started yet,” Mostak said recently, referring to the 'dot-com' bubble of the late 1990s, when emerging Speculative bets on Internet companies ultimately result in large losses for many investors.
The Mistral AI investment is just one of many investments venture capitalists will be competing to hop aboard the AI rocket this year. In the first six months of 2023, the company funneled $15.2 billion into generative AI companies around the world, according to Pitchbook data.
A large portion of this amount comes from Microsoft's (MSFT) $10 billion investment announced in January in OpenAI, the developer of the popular generative AI chatbot ChatGPT.
But even excluding Microsoft's big deal, VC investments in generative AI increased by almost 58% compared to the same period in 2022.
According to Lightspeed's Moyroud, the current buzz was sparked by the general availability of ChatGPT in November. He has seen more founders mention generative AI in their funding pitches, but he takes some of those pitches with a grain of salt.
“we, [seen] Moileux said some people who haven't necessarily spent a lot of time in the industry are adding a bit of a generative AI shine (if you can call it that) to their pitch, adding: It takes time to do that,” he said. ” This is behind the claims of some founders.
He does not include Mistral AI in that group. Moyrou's venture capital firm (which he would only say contributed to a “significant portion” of the startup's €105 million in revenue) is paying a premium for the “unparalleled” experience of the three founders. had paid the amount. Previously, everyone was talking about “Large Language Models''. Two of them are in his Meta, which is the parent company of his Facebook, and one is in his DeepMind, which is his Google.
“Only a fraction of probably 80 to 100 people in the world have experience training large-scale language models… [and] on a large scale,” Moyroux said.
It's not just big-money retail investors looking to cash in on the AI boom. Flows into the world's top five exchange-traded funds focused on AI have increased an average of 35% since the beginning of the year. And after a painful 2022, shares in the tech-heavy Nasdaq index have soared nearly 42% in that time, outpacing the broader S&P 500 index, which has risen less than 19%.
In May, US-based Nvidia, which makes advanced microchips needed to generate AI, became the sixth company in the world to reach a market capitalization of $1 trillion. The company's stock price has soared 207% since the beginning of the year.
However, NVIDIA stock has traded at a price-to-earnings ratio of 237 times over the past 12 months, a measure of whether a stock is overvalued or undervalued. The higher this ratio, the more likely the stock is overvalued. For comparison, companies in the S&P 500 index have traded at an average multiple of 24x over the same period.
Nvidia is profitable, but AI software company C3.ai, whose stock has soared more than 240% this year, is not. Also, there is no prospect of making a profit this year or next.
Investors told CNN that the situation is strikingly similar to the dot-com bubble. But every time a bubble occurs, there is always a pop sound.
Starting in late 1998, the Nasdaq more than doubled in value in 1999 alone as investors poured money into dot-com companies. However, despite high expectations and huge valuations, most startups failed to generate revenue or profits, according to Goldman Sachs. The Nasdaq stock price plummeted 81% from its peak in March 2000 to the end of September 2002. The bubble has completely burst.
Mike Reynolds, vice president of investment strategy at U.S. asset management firm Glenmede, said the current excitement is “reminiscent of what happened in the past.” [90s] During the tech bubble, many companies were not yet profitable, but people were so optimistic about their prospects that they were willing to bid. [their stock price] It gets higher and higher. ”
“We haven't really seen it yet [the AI hype] It will lead to concrete fundamental results,” he added.
It will be “very difficult” for investors to know whether they are backing the next AI equivalent of Amazon (AMZN) or Google (GOOGL), Reynolds said. Of the top 10 most valuable technology and communications stocks today, Microsoft (MSFT) and Cisco (CSCO) were in the top 10 at the peak of the dot-com bubble in March 2000, according to Glenmede analysis. There were only two companies.
“It's not always clear who the long-term winners of innovative disruption will be,” Reynolds says.
He added that in the late 1990s, companies “just had to add the word 'dotcom' to the end of their name and stock price.” [would go] The next day it rose 10%. ”
Jordan Jacobs, co-founder and managing partner of Radical Ventures, a Toronto-based venture capital firm specializing in AI, sees a similar impulse among today's technology founders.
“If you buy a 'dot-ai' domain and claim to be an AI company, that doesn't actually make you an AI company,” Jacobs told CNN. “Part of our job as investors is to figure out who's real and who's not.”
Jacobs, who has founded two AI companies in the past 13 years, said he believes there is “no understanding” of how valuable the technology will be in the future.
He predicts that within the next decade, AI will be integrated into or completely replace all software, creating “trillions of dollars of economic value.” He said the technology is also breaking new ground in the fields of drug development and climate change modeling.
The release of ChatGPT, coupled with Microsoft's major investment in OpenAI, triggered a “sudden awakening for generalist investors” to the extraordinary potential of AI. It was also the moment when everyone finally had access to technology.
It was “kind of magical,” he says.
