Can Anthropic's record breaking IPO stop the AI bubble bursting?
Anthropic's hotly anticipated IPO is expected to put it in the top 10 most valuable US companies, but the future is uncertain.
Can Anthropic's record breaking IPO stop the AI bubble bursting?
It was during November 2020 that somewhere in San Francisco, a woman called Daniela Amodei phoned a friend, asking for advice.
Daniela oversaw safety and policy at OpenAI, one of the most talked-about AI companies in the world.
But Daniela and her brother Dario - who worked there too - had misgivings over safety levels at OpenAI and thought it could be time to walk away.
So she did what many do when we can't make a decision about something, she phoned a friend.
The friend's reply was direct and to the point: "You don't need to be on the phone with me.
You already know the answer."
The Amodei siblings and a handful of other colleagues quit.
Together, they set up a rival – Anthropic.
Six years on and Anthropic, the maker of leading AI model Claude, has predicted revenues of $52bn in 2026 and $135bn in 2027, according to analyst estimates.
It is about to sell shares to the public for the first time in what's known as an Initial Public Offering or IPO, aiming for a valuation of an estimated $1.4tn (£1tn), which would propel it straight into the top 10 of the most valuable US companies.
The sale comes at a crucial time when many are wondering whether, amid increasing concern about the potential of AI, the industry's bubble is showing signs of bursting.
Where once investors were "all in" on a technological revolution to rival the railroads, the car, electricity and the internet, some are now hedging their bets, amid arguments that AI prices will be driven down as the technology becomes more common and concerns that a hefty wedge of the AI investment is tech giants investing in each other.
Shares in companies that supply the raw computing power behind AI, like Coreweave and Nebius, have seen their value cut in half while even the mighty Nvidia has lost its crown as the world's most valuable company with its shares down 30% - a staggering $2tn fall in value.
So could the Anthropic IPO reinflate the bubble or will it be remembered as the last gasp of a boom that is heading for a bust?
At an event last week, Anthropic CEO Dario Amodei told JPMorgan Chase's CEO Jamie Dimon that his guiding principle was to "never let your company be worth a single dollar more than it's worth".
The company is selling roughly $65bn worth of shares to the general public and a further $26bn on to other big tech companies.
It is comfortably the single biggest money raising event in US corporate history.
It's an old adage that during a gold rush you should buy companies that sell the picks and shovels as opposed to those staking everything on striking gold.
In AI's case that's the chip makers, the data centre builders, the energy providers.
But Anthropic is arguably the gold itself and this IPO will see what value people put on it, according to Fiona Cincotta.
"I think the reason that this IPO is so important is basically it's the first ability for the public investors to access a pure play AI trade," she says, adding that previous offerings have been about "other parts of the ecosystem, which are, the semiconductors, the chips, the, data centres.
But this is the first time that we were actually able to access the AI provider."
Daniela Amodei helped set up a rival company to OpenAI
What a company is worth depends almost entirely on predicted future earnings.
It's that calculation that many think presents the sharpest needle to the AI bubble.
Will current and future users of AI be prepared to pay enough to justify the trillions being spent on the chips, data centres, energy and employees, some of whom can command annual salaries of over $100m?
"AI is certainly in a financial bubble.
While LLMs [large language models] offer genuine utility and are undoubtedly here to stay, the economics don't add up," said Gary Marcus, Professor Emeritus at New York University and author of Taming Silicon Valley, speaking earlier this year.
"[The] more companies build AI models, the [more the] technology becomes commoditised, eliminating meaningful competitive advantages and driving down prices."
So far, the growth rates in Anthropic revenue are extraordinary.
But these explosive growth rates hide a potential problem.
Around a quarter of that revenue came from just a handful of companies, among them Google and Amazon.
It's also concentrated in the US which accounts for two-thirds of its revenue.
There is also the staggering network of circular financing found in the AI world.
For example, Anthropic has agreements to buy or lease hundreds of billions of dollars worth of chips and computing power from a handful of companies including Broadcom, Google, Amazon.
Many of those firms are also investors in Anthropic and major customers of its products.
It's not just some Wall Street investors questioning whether the AI boom is all it's cracked up to be.
Out on Main Street, ordinary Americans have concerns too.
With the US driving so much of global AI growth, a growing public backlash against a technology that threatens to replace millions of jobs and hoover up vast amounts of energy and water became a factor in the recent midterm elections in the US.
Full detailed coverage is shown on the council website above. Source attribution: BBC Business.
Back to all council news