On December 4, according to Bloomberg, Dario Amodei, CEO of American AI startup Anthropic, pointed out on Wednesday that some AI companies have taken excessive risks by committing hundreds of billions of dollars or more to develop and support AI systems.
Amodai said in an interview at the New York Times DealBook Summit on Wednesday,The industry is facing a "real dilemma":On the one hand, it requires huge long-term investment in data centers, and the construction period may last for several years; on the other hand, there is still "uncertainty" in the growth rate of AI's economic value.
"Our company is trying to manage risk as responsibly as possible," he said, "but I think some players are taking risks in a carpe diem way."
Amodai did not name specific companies, but he said some companies "took the risk dial too far."
In recent months, leading AI companies such as OpenAI, Meta and Google have significantly increased their investment in data centers and chips to build more advanced AI systems and pave the way for the widespread application of this technology. OpenAI alone has committed $1.4 trillion to AI infrastructure projects, fueling fears of an AI bubble.
Anthropic has also increased investment, but not to the same extent as OpenAI.For example, Anthropic recently announced that it will invest $50 billion to build its first custom data centers in multiple locations in the United States.
Anthropic was founded in 2021 by former OpenAI employees and is committed to becoming a more responsible AI steward than its competitors. The company is primarily focused on growing its enterprise business rather than the consumer market. In September this year, Anthropic completed a $13 billion financing round at a valuation of $183 billion.