As Anthropic and OpenAI are set for IPOs, Broadcom CEO has a message for AI investors: For now, it is still losing money
With frontier artificial intelligence (AI) labs Anthropic and OpenAI moving toward historic Initial Public Offerings (IPO), Broadcom CEO Hock Tan has issued a blunt warning to AI investors: despite the towering valuations, the generative AI sector remains an unprofitable endeavour. Speaking at an industry gathering in Singapore organised by DBS Bank, the leader of the $1.7 trillion semiconductor supplier framed AI as a volatile and uncertain financial bet at current times.
“For now, it’s still losing money,” Tan said of the AI sector, adding, “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.”
Anthropic is expected to list publicly around Thanksgiving, targeting a valuation in the $2 trillion territory, with OpenAI projected to follow closely behind. Both companies are depending on a vast international build-out of data centres, semiconductor wafers, advanced processor manufacturing and cloud computing capacity to fulfill market demand and eventually turn a profit.
Tan warned that a looming shortage of investment capital will likely form the industry's next major roadblock to continued expansion.
‘Like they are in a casino’The assessment comes from an executive whose company directly finances its growth by supplying essential hardware components for the AI boom, with Broadcom's equity rising nearly 10% this year. Tan argued that many market participants are pouring money into technologies they fundamentally fail to grasp, comparing enterprise spending behavior to children wasting coins in an arcade.
“A lot of investors have no clue what they are investing in,” Tan remarked, according to Bloomberg. They “set very extreme expectations only to find out that it’s almost like they are in a casino.”
“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. That’s what happening in enterprises today,” he added.
Sharing the stage alongside Tan, Intel CEO Lip-Bu Tan voiced similar caution regarding infrastructure limits. He noted that physical hardware supplies will remain constrained over the next few years, warning that memory chip shortages could become the first segment within the broader AI ecosystem to experience a slowdown.
“For now, it’s still losing money,” Tan said of the AI sector, adding, “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.”
Anthropic is expected to list publicly around Thanksgiving, targeting a valuation in the $2 trillion territory, with OpenAI projected to follow closely behind. Both companies are depending on a vast international build-out of data centres, semiconductor wafers, advanced processor manufacturing and cloud computing capacity to fulfill market demand and eventually turn a profit.
Tan warned that a looming shortage of investment capital will likely form the industry's next major roadblock to continued expansion.
‘Like they are in a casino’The assessment comes from an executive whose company directly finances its growth by supplying essential hardware components for the AI boom, with Broadcom's equity rising nearly 10% this year. Tan argued that many market participants are pouring money into technologies they fundamentally fail to grasp, comparing enterprise spending behavior to children wasting coins in an arcade.
“A lot of investors have no clue what they are investing in,” Tan remarked, according to Bloomberg. They “set very extreme expectations only to find out that it’s almost like they are in a casino.”
“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. That’s what happening in enterprises today,” he added.
Sharing the stage alongside Tan, Intel CEO Lip-Bu Tan voiced similar caution regarding infrastructure limits. He noted that physical hardware supplies will remain constrained over the next few years, warning that memory chip shortages could become the first segment within the broader AI ecosystem to experience a slowdown.
Next Story