By the time a dormant penny stock company known as Applied Sciences managed to wrangle a listing on the Nasdaq in April 2022, it had reinvented itself as a cloud hosting service for bitcoin miners and changed its name to Applied Blockchain. But with the crypto world crashing that spring, the stock never took off. Within months, Applied Blockchain pivoted again — renaming itself Applied Digital.
If its previous iteration had been too late to cash in on the bitcoin mining craze, the company wasn’t going to miss the next big one: artificial intelligence.
Applied Digital’s stock finally began to soar in May, when CEO Wes Cummins announced that the company had signed a cloud hosting deal potentially worth $180 million with an unnamed but prominent AI customer, and another one for up to $460 million with another big player in the booming AI space. By July, the stock had surged some 450 percent for the year, becoming — for a time, at least — one of the big winners in today’s AI-driven stock market.
Applied Digital is hardly alone in trying to capture some AI magic. Ever since ChatGPT burst onto the scene last November — with its “large language model” boasting a humanlike writing capability that at first blush seems to ensure productivity gains for everyone from publishers and movie studios to investment banks and hedge funds — so-called “generative AI” has turned the markets on their heads.
Coming off the worst year in recent history for venture capitalists, private market players like Andreessen Horowitz, Coatue Management, Tiger Global, Sequoia Capital, and Softbank quickly redirected their dollars to AI upstarts. Meanwhile, the stock prices of the big tech names suspected to be the major beneficiaries of this often-called “revolutionary” form of artificial intelligence have skyrocketed. The 2022 tech downturn became a faint memory as some five tech companies — Alphabet, Amazon, Apple, Microsoft, and Nvidia — accounted for the lion’s share of the stock market’s remarkable comeback this year, with the S&P 500 rising 20 percent through July. Nvidia, maker of the superfast chips that switched from powering bitcoin mining to making generative AI possible, has gained more than 250 percent so far in 2023, making it the S&P index’s top gainer.
In the midst of this bonanza, Applied Digital’s CEO — who is also the president of B. Riley Asset Management and was at one point the owner of about 25 percent of Applied Digital’s shares — posted on Twitter (now X) in June that the company had ordered 26,000 top-tier H100 GPUs, or video processing cards, from Nvidia for $40,000 apiece. To those casting a wary eye on the company, it seemed too good to be true. A purchase that big would allow Applied Digital to “jump to the top of the pile in high-performance computing, alongside Google, Meta, and [Amazon Web Services],” short-seller Dan David said in a Wolfpack Research report that called the company “an embarrassing and predictable stock promotion.” He noted that the cost to purchase such equipment would run more than $1 billion — more than Applied Digital’s market cap of nearly $600 million.
The explosion of interest in AI after the emergence of ChatGPT has predictably attracted the worst promoters and scumbags to peddle fake AI wares to credulous investors,” says David, who claims Applied Digital is one of them. (The company did not respond to a request for comment.)
Applied Digital is one of nearly a dozen companies that short-sellers have been eyeing this year as questionable beneficiaries of AI mania. The shorts say their antennae are on the alert for even more. But so far, the skeptics are fighting an uphill battle. Short interest is relatively high in several of these stocks, and it’s been costly to bet against them. As of August 25, Applied Digital short-sellers, for example, had placed bets on 19 percent of the outstanding shares of the company. The short-sellers in aggregate are down almost 30 percent this year on the name, having lost about $10 million, says S3 Partners — although those who shorted Applied Digital at or near its peak would have profited.
Orso Partners co-founder Nate Koppikar, who is also short Applied Digital, has a term for what he sees going on. He calls the phenomenon “the grift shift” — arguing that companies and venture capital funds have pivoted from their losing crypto and tech bets to cash in on the AI moment.
All told, generative AI and machine learning start-ups raised about $39.4 billion this year, with $19.4 billion of that in the second quarter, according to PitchBook’s second-quarter Artificial Intelligence & Machine Learning Report.
Though money has been pouring into everything said to be “AI,” a few critics are starting to wonder whether the latest technology is really transformational or merely evolutionary. Meanwhile, large language model problems like “hallucinations,” “drift,” and “degradation” are starting to populate the tech literature, in reference to the various types of errors cropping up in generative AI as studies indicate that the products’ output appears to have worsened over time.
Some investors are starting to grow antsy. AI has dominated the spotlight this year, but “investors are growing impatient with the lack of revenue growth from generative AI innovators,” says PitchBook. “No longer are big tech stocks going up after new-product or partnership announcements, putting pressure on start-ups to gain traction.”
Says Koppikar, “This looks like a great market for fraudsters to chase because of the ability to ‘Theranos’ your product” with semifunctional prototypes and claims of AI being used in products that are really just powered by humans. He argues that many of the new AI companies “are complete frauds or will never have scalable revenue models,” adding, “There is going to be a staggering amount of capex burned on AI that goes nowhere.”
You can read my entire story here:
Michelle Celarier 