AI bubble inflated by SoftBank CEO who once threatened to set himself on fire for…

(Bloomberg Opinion) – Masayoshi Son is man of extremes. His $20 million bet on Alibaba Group Holding Ltd. became the most successful of all time when it swelled to more than $70 billion after a decade.

He also lost $70 billion of his net worth in the dotcom crash, but then managed to raise one of the biggest investment funds of all time – the $100 billion Vision Fund – in 2017. The 67-year-old has been on a roller coaster that would put most mortals in therapy for life.

He is unapologetically dramatic, once threatening to set himself on fire if he wasn’t granted a telecom license in Japan, and shoots from the hip in business deals. He says he backed Alibaba because of founder Jack Ma’s “strong eyes, shining eyes,” and he convinced the crown prince of Saudi Arabia to put $45 billion into his Vision Fund in a single, 45-minute conversation, by offering him “a Masa gift, a trillion-dollar gift,” according to Son’s 2017 interview with Bloomberg’s David Rubinstein.

So unusual is Son — known as “Masa” — that he is now the subject of two books, one by former Financial Times editor Lionel Barber and another by Alok Sama, a former president of Son’s telecommunications and technology giant Softbank Group International. Both paint a portrait of a global maverick who never seems to sleep, but it’s Sama’s book The Money Trap that looks closely at Son’s potential impact on the booming AI market. “Masa Son’s ambition is to be [AI’s] high priest,” writes Sama.

“He’s not a gambling man,” Sama tells me, adding that Son has made 10 times what he lost on WeWork through his investment in Arm Holdings Plc, the chip-design firm that’s now at the heart of his AI ambitions. “He’s living in the future.”

Son founded the Vision Fund to capitalize on the so-called singularity, a hypothetical threshold sometimes defined as being the moment when AI surpasses the human brain. When he spoke at Softbank’s annual general meeting this past summer, Son said he’d become obsessed with realizing “artificial super intelligence” and was born to make it happen. His past investments were just a warm up.

No surprise then that Son has been on an AI tear. Last month, he invested $500 million in OpenAI through his Vision Fund, having missed out on the AI company’s previous funding rounds. Earlier this year, he led a $1 billion investment into British self-driving car maker Wayve and put as much as $20 million into Perplexity AI, which is going head-to-head with Alphabet Inc.’s Google Search. Son recently told investors he’ll continue fishing in areas like autonomous driving, data centers and AI robotics. He’ll have rich pickings, with dozens more generative AI startups struggling to cover the costs of computing power and eager for funding in a business dominated by the tech giants.

But Son poses some risk for everyone else in the pond. The clue is in the title of Barber’s book: Gambling Man. Though he’s worth about $16 billion, according to the Bloomberg Billionaires Index, Son’s career has been nothing short of volatile. His bet on WeWork – another shoot-from-the-hip deal that came from a 12-minute meeting and car ride with founder Adam Neumann – led to a $32 billion loss for the Vision Fund as the startup went bankrupt. Son later called the bet “foolish,” but his approach to investing doesn’t appear to have been tempered.

That is hardly what today’s AI market needs. It has already shown bubbly dynamics, with soaring valuations (hello, Nvidia Corp.) and intense hype. Son threatens to fuel those dynamics further, pushing the market toward unstainable growth. In the two years since the launch of ChatGPT, investors have added $8.2 trillion to the market valuations of tech’s six biggest firms, but the generative AI market is still in its early stages and doesn’t need the kind of volatility that an eccentric billionaire’s big gambles could introduce.

The obvious consequence of injecting enormous capital into new markets is a pressure-cooker environment, where companies burn through cash to grow exponentially. When many of those companies fail, they can implode spectacularly, just as WeWork did, with painful repercussions.

Today’s AI boom has been defined by people with noble objectives. Its leading visionaries, OpenAI founder Sam Altman and Google DeepMind founder Demis Hassabis, both set out to build their own versions of “super AI” in the hope it would elevate global living standards, cure cancer and solve climate change. Instead, they’ve acted as de facto product arms for Microsoft Corp. and Google, extending the dominance of those companies.

The road to hell is paved with good intentions. At a minimum, expect Son’s intervention to make that road much bumpier.