OPINION Back in December 2017, an obscure American soft drinks company changed its name from Long Island Iced Tea to Long Blockchain.
Never mind that the company had no background in technology. Forget about the strange borrowed name – Long Island iced tea was actually a generic name for a very strong cocktail that was popular in the 1980s (although growing up in Seattle, we always called it “Electric Iced Tea,” perhaps because we were so far away from New York that “Long Island” meant nothing to us). It didn’t matter! This was during the first crypto boomlet and the company’s shares more than tripled in value, giving the tiny company a market cap of over $90 million.
That turned out to be the peak of that particular bubble. Crypto prices headed off into their first crypto winter, with Bitcoin losing more than half its value by next September. That boom-bust cycle has repeated itself twice more at greater volume since then. Fortunes were won and lost, fraudsters went to jail and were pardoned, and there’s still no clear, obvious, technical use for the blockchain or cryptocurrency in general, apart from speculation and situations where you might otherwise be using a suitcase full of unmarked US dollar bills (illegal transactions, escaping countries with collapsing currencies, and so on).
History is repeating itself. On Wednesday, Allbirds, a venture-backed company that made woolen shoes for Silicon Valley hipsters and went public in 2021, announced that it was pivoting to AI infrastructure and changing its name to NewBird AI. This came weeks after the company, once valued at over $4 billion by stock traders, closed its US full-priced stores and sold its intellectual property and assets for $39 million.
Investors, seeing the word “AI,” bid the stock up more than 600 percent. Never mind that the company has zero experience in AI and is trying to compete with massively funded companies to “acquire high-performance GPU assets,” which it can then rent to customers. At least when the crypto miners pivoted to AI, they had some experience building and running datacenters filled with banks of computers.
History doesn’t repeat, but it rhymes. We’ve been seeing all kinds of signs that the financial underpinnings of the AI bubble are getting rickety. Reports abound that Anthropic is throttling throughput and compromising quality to save money. Investors are reportedly calling foul on OpenAI’s last announced valuation of eight hundred eleventy gazillion dollars, weeks after its most recent fundraising round opened investment to retail traders (translation: bag holders) through an exchange-traded fund. OpenAI itself is pulling back on its plan to build massive numbers of hypermegadatacenters to power the AI revolution, as partner Oracle seems to be developing chilly feet.
What difference does this make to the humble IT worker in the trenches? You’re not an investor or gambler or financial TV talking head. Financial bubbles come and go, but the server rooms and networks have to keep running, patches need to be deployed, and user tickets must be handled.
But financial bubbles can distort reality. When this many people have this much money riding on a particular technology, there will be a flood of hype around said technology. Keep your head on straight and ask yourself, really, is the set of software products that the industry is now calling “AI” helping you do new things, or is it simply shuffling work around to different people? Is it saving you time, or creating new messes that you (and other people) have to then clean up later? Is it actually delivering results? Is it worth what you’re paying for it?
Because if there is, in fact, an AI bubble, any of these products that have actual value may become a lot cheaper after it pops.
By the way, if you’re curious about whatever happened to Long Blockchain, the SEC delisted it in February 2021 and charged three people connected with insider trading. One of the defendants later settled with the SEC without admitting or denying the allegations. ®