Some of the world’s most powerful and well-connected leaders converged in Manhattan on Wednesday, and one subject dominated the room: artificial intelligence.
Three years after ChatGPT launched, revolutionizing the way the world thinks about and uses AI, the arms race is bigger than ever — perhaps too big.
At the annual New York Times DealBook Summit, leaders from BlackRock’s Larry Fink to Lai Ching-te, the President of Taiwan, were asked about the state of the AI industry and the potential for an AI bubble, a hot topic in Silicon Valley and on Wall Street over the past few months.
Tech giants like Meta, Alphabet, and Microsoft are spending tens of billions of dollars in capital expenditures — largely AI infrastructure — this year. So far, 49 US AI startups have raised at least $100 million this year, TechCrunch reported in November.
While the big names at DealBook agreed that AI is here and growing, they also said the boom would likely leave casualties in its wake.
“There are going to be some huge winners and huge failures,” Fink said to DealBook founder Andrew Ross Sorkin. “I’m not here to suggest there’s not going to be some headline blow-ups.”
That said, the financial bigwig said he’s confident that the demand will be there and that the “hyperscalers” — companies like Amazon, Google, and Microsoft that provide the resources that keep AI running — who he has spoken to are short on compute.
It was an opinion parroted by other leaders speaking at the event, those with and without a stake in the industry.
Dario Amodei, the CEO of AI company Anthropic, said the industry is inherently risky given the large amount of capital needed to build the data centers that power AI,
“Even if the technology fulfills all its promises, I think there are players in the ecosystem who, if they just make a timing error, they just get it off by a little bit, bad things could happen,” he said.
Amodei said Anthropic is managing its risk by working with large enterprise clients and investing in compute in a conservative way. He suggested some of his competitors haven’t been as cautious.
“There are some players who are YOLOing,” Amodei said.
He wouldn’t name names, but took a veiled jab at his former employer, OpenAI, and boss, Sam Altman, later in the conversation: “We don’t have to do any code reds,” he said, referring to OpenAI’s recent memo about the threat of Google’s AI model.
Those that do fail will be a necessary part of creating the best technology, which ideally would not only make money but also surpass that of other countries, many of the leaders said Wednesday.
“If we don’t spend enough faster in AI, in digitization, and in tokenization, other countries are going to beat us,” Fink said, endorsing the investment in the industry.
Amodei said the government would be pivotal to that. He argued for a more proactive approach to regulation and that Nvidia chips should not be sold in China.
If the best model is “plopped down in an authoritarian country, I feel like they can outsmart us in every way: intelligence, defense, economic value, R&D,” he said
Lai, for his part, agrees that work must be done to prevent any sort of AI calamity.
“Leaders around the world, especially those from countries with AI-related industries, should work together and take necessary measures to ensure AI develops sustainably and has a soft landing so that it can drive long-term global growth,” he said.
Of course, like the rest of those who spoke about an AI bubble on Wednesday, Lai has skin in the game. Taiwan leads the world in making the chips necessary for AI — and fueling the very potential bubble of which he spoke.
