But on Monday, Broadcom CEO Hock Tan likened many of the investors drawn to the AI boom to gamblers and questioned the economics of AI labs as Wall Street debates whether the AI bubble will burst.

“A lot of investors have no clue what they are investing in,” he said at the DBS Asia Future Forward Conference in Singapore, according to Bloomberg. He said they “set very extreme expectations only to find out that it’s almost like they are in a casino.”

Broadcom supplies the hardware that frontier AI labs need to build their systems. But while Broadcom can profit from selling chips, the AI labs buying computing power must still find enough paying customers to cover the cost of developing and running their models.

The challenge is to turn demand into sustainable revenue. Tan acknowledged that OpenAI’s reported $50 billion in annualized revenue at the end of September points to a viable business model for AI labs, but said it doesn’t guarantee that they will be profitable. Annualized revenue extrapolates a company’s current sales pace across a year. It does not measure revenue collected over 12 months, and labs have expenses to cover. OpenAI expects to spend over $25 billion on training alone this year. 

“For now it’s still losing money because they’re spending a lot on training,” Tan said. “So frankly, it doesn’t tell anything at this point except that there is a lot of potential revenue growth out of providing frontier generative AI models.”

AI’s next bottleneck could be money

The cost of developing AI also raises questions about whether investors will keep footing the bill, with some analysts predicting that AI capital expenditure growth will slow next year. Tan said a lack of capital “might be the biggest constraint” on the AI industry, more than a memory, semiconductor, or electricity shortage. 

There are already signs investors are questioning how much money AI infrastructure companies are asking for. Firmus, an Australian data center company backed by Nvidia, withdrew its planned IPO on Oct. 9 after weak investor demand for its marketed share prices, according to Bloomberg. 

That doesn’t necessarily mean demand for AI computing is weakening. Intel CEO Lip-Bu Tan, who shared the panel with Tan at the conference, said hardware could remain in short supply for another two to three years, per Bloomberg. However, demand for chips doesn’t guarantee the companies buying them will earn enough to justify their investments.

Companies ‘wasting money’ on AI experimentation

Tan’s concerns also extend to the companies paying to use AI. 

He criticized them for “wasting money” on tokens to experiment with AI instead of figuring out where using it generates returns, comparing that to children feeding arcade tokens into games with nothing to show for it in the end. 

“When you are a kid, you go to a video arcade, you play. Your parents give you $10 of tokens; you finish it in 10 minutes, and you haven’t learned anything,” he said. “That’s what’s happening in enterprises today.”

His concerns echo the findings of a McKinsey survey on AI’s ROI published in August. Although 80% of respondents said AI had improved their individual productivity, only 37% reported that it had contributed to their organization’s earnings before interest and taxes. About one in five respondents said AI-related operating expenses, including token costs, had constrained their organization’s use of the technology.