OpenAI’s Sam Altman Expects To Spend ‘Trillions’ On Infrastructure

OpenAI Chief Executive Officer Sam Altman wants to spend trillions of dollars over time on the infrastructure required to develop and run artificial intelligence services. He just needs to figure out a way to raise that kind of money.

“You should expect OpenAI to spend trillions of dollars” on data center construction in the “not very distant future,” Altman told a group of reporters on Thursday. “And you should expect a bunch of economists to wring their hands and say, ‘This is so crazy, it’s so reckless, and whatever. And we’ll just be like, ‘You know what? Let us do our thing.’”

Altman said the startup is devising a novel way to bankroll that outlay. “I suspect we can design a very interesting new kind of financial instrument for finance and compute that the world has not yet figured it out,” he said. “We’re working on it.”

In January, Altman joined SoftBank Group Corp.’s Masayoshi Son and Oracle Corp.’s Larry Ellison at the White House to announce a $500 billion, four-year infrastructure venture called Stargate. Still, Altman envisions spending far more than Stargate now entails.

Altman’s remarks came during a rare and wide-ranging discussion with reporters over dinner. He spoke about ChatGPT’s rapid growth, the possibility of encrypting chats for better privacy and the new browser wars. 

When asked about AI startup Perplexity’s $34.5 billion bid to buy Google’s Chrome browser, Altman said OpenAI should also “take a look” at it. An OpenAI executive previously said the company would be interested in bidding for Chrome if a federal court ordered Alphabet Inc.’s Google to spin it out. 

Altman also said he sees parallels between the current investment frenzy in artificial intelligence and the dot-com bubble in the late 1990s. In both cases, Altman said, “smart people” became “overexcited” by a new technology. But in each instance, he said, that technology was “real” and poised to eventually have lasting impacts on the business world and society. 

“Are we in a phase where investors as a whole are overexcited by AI? In my opinion, yes,” Altman said in response to a question from Bloomberg News during the conversation. “Is AI the most important thing to happen in a very long time? My opinion is also yes.” 

Altman said “society as a whole” is unlikely to regret the massive investment in AI, but also admitted he thinks some current startup valuations are “insane” and “irrational behavior.” He added: “Someone’s gonna get burned there.”

Altman, more than almost anyone, has helped fuel that investor enthusiasm. OpenAI kicked off the current AI boom with the release of ChatGPT in late 2022, prompting other tech firms and startups to spend billions of dollars on talent, chips and data centers to build similar AI software. While Altman expects OpenAI to “spend a lot of money,” he also thinks the company will see “huge profits” eventually. 

“It’s very rational for us to keep investing right now,” he said.

At least part of OpenAI’s path to funding the significant cost of AI development will likely include a public offering in the future, though Altman declined to give a specific timeline. OpenAI is currently working to complete a complex corporate restructuring effort that has been underway for months. 

“I do think we have to go public someday, probably,” Altman said. But Altman also noted he is not as “well-suited” to be CEO of a public company.

Altman said he now sees OpenAI as being more like four companies: a consumer technology business, a “mega scale” infrastructure operation, a research lab and “all of the new stuff,” including planned hardware devices. To that end, his company recently brought on former Instacart boss Fidji Simo to be the CEO of OpenAI’s applications business.

“I can’t run four companies,” Altman said. “It’s an open question if I can run one.”

Much of his focus in the discussion was on the rocky rollout of GPT-5 over the past week. The new AI model, which Altman had hyped up for months, was met with mixed reviews and confusion from many users. 

Some people were frustrated by GPT-5’s writing style and personality; others complained that their favorite models had effectively been phased out after the launch. In response, OpenAI has issued a series of fixes and mea culpas.

“I think we totally screwed up some things on the rollout,” Altman said. “We’ve learned a lesson about what it means to upgrade a product for hundreds of millions of people in one day, and the differences in the kinds of attachment people have with this product versus previous products.”