Anthropic CEO Dario Amodei shared his thoughts on whether the AI industry is in a bubble at the New York Times Dealbook Summit on Wednesday. This was done in addition to overshadowing a specific, unnamed competitor, which is clearly OpenAI.
Amodei declined to give a simple yes or no answer, citing the complexities of bubbles, but instead elaborated on his thoughts on the economics of AI.
He said he was bullish about the technology’s potential, but warned that there could be players in the ecosystem who make “timing errors” and that “bad things” could happen in terms of financial returns.
“There is an inherent risk when the timing of economic value is uncertain,” Amodei explained. He said companies must take risks to compete with each other and with authoritarian adversaries, but added that some companies “are not managing that risk well and are taking unwise risks.”
The problem, he said, is uncertainty about how fast the economic value of AI will grow and how that can be properly mapped to delays in data center expansion.
“There is a real dilemma and we are trying to address it as responsibly as we can as a company,” Amodei said. “And I think some players are turning the risk dial up too high and ‘YOLO’,” he added, using the slang phrase “you only live once”, which is often used to justify taking risks. “I’m very concerned.”
Additionally, he answered questions regarding the deprecation schedule for AI chips. This is also a hot topic and a factor that could negatively impact the industry’s economics if GPUs become obsolete and lose value sooner than expected.
tech crunch event
san francisco
|
October 13-15, 2026
“The issue is not the longevity of the chips. Chips continue to work for a long time. The issue is that new chips come along that are faster and cheaper… so the value of the old chips may go down somewhat,” Amodei said.
He said Anthropic is making conservative assumptions on this and other fronts as it plans for an uncertain future.
The CEO said the company’s revenue has grown 10x annually over the past three years, going from zero to $100 million in 2023, $100 million to $1 billion in 2024, and between $8 billion and $10 billion by the end of this year.
But Amodei said it would be “really foolish” to simply assume this pattern would continue. “I don’t know if it’s going to be 20 billion or 5 billion a year from now…It’s very uncertain. I try to plan conservatively. So I’m planning for the lower end of that, which is very disconcerting,” he said.
AI companies like his need to plan how much computing they will need over the next few years and how much they should invest in data centers. If you do not purchase sufficient quantities, you may not be able to serve your customers. And if you buy too much, you may struggle to keep up with costs or, in the worst case scenario, go bankrupt.
Last month, OpenAI suffered a PR crisis when its CFO said it wanted the U.S. government to “backstop” its infrastructure financing, meaning it would insure taxpayers with an option if OpenAI couldn’t. After the uproar, she responded with a comment.
In a veiled reference to OpenAI CEO Sam Altman, Amodei warned that people who take more risks can overextend themselves, especially if they’re “biologically people who just want to ‘YOLO’ things, or just like big numbers.”
“We basically think we’re OK in most of the world…I can’t speak for other companies,” he said.
