OpenAI CEO Sam Altman told reporters on Thursday that the company plans to spend trillions of dollars to build the infrastructure needed to run advanced artificial intelligence.
The conversation, which took place with Bloomberg over dinner, covered OpenAI’s funding plans, the messy rollout of GPT-5, ChatGPT’s growth, privacy concerns, and a potential browser acquisition.
Sam said people should “expect OpenAI to spend trillions of dollars on infrastructure in the not very distant future.” He didn’t hide the fact that economists and analysts will probably call it reckless. He said, “We’ll just be like, ‘You know what? Let us do our thing.’”
OpenAI wants to invent new financial instruments to fund AI buildout
OpenAI doesn’t yet have those trillions in its bank account. Instead, Sam said they’re working on creating a “very interesting new kind of financial instrument” that doesn’t currently exist. The goal is to finance computing and infrastructure costs without relying on traditional fundraising routes. “We’re working on it,” he said.
Earlier this year, in January, Sam stood beside Masayoshi Son of SoftBank and Larry Ellison of Oracle at the White House to announce Stargate, a $500 billion plan to build AI infrastructure over four years.
That figure already dwarfs anything happening in the crypto industry, but Sam said OpenAI now expects to spend far more than that. Stargate, in his view, is just the beginning.
Sam also discussed how OpenAI could eventually go public to fund part of this plan. He admitted a public offering is likely at some point. “I do think we have to go public someday, probably,” he said. But he also said he’s not sure he’s the right guy to run a public company, saying he’s not “well-suited” for that kind of role. At the moment, OpenAI is still deep in a corporate restructuring process that’s been dragging on for months.
The conversation also touched on what Sam called an irrational hype cycle. He compared the AI craze to the dot-com bubble of the late 1990s, saying both saw smart people getting way too excited about a new technology.
But he added that AI, like the internet, is real and will have a lasting impact. “Are we in a phase where investors as a whole are overexcited by AI? In my opinion, yes,” Sam said. “Is AI the most important thing to happen in a very long time? My opinion is also yes.”
Sam Altman admits GPT-5 rollout failed, says OpenAI is learning
Another major topic at the dinner was the GPT-5 rollout, which hasn’t gone well. Sam said OpenAI made serious mistakes when it launched the model last week. He admitted the company “totally screwed up some things on the rollout.” GPT-5 had been hyped up for months, but users complained about everything from the chatbot’s new tone to the removal of older models they preferred.
Some users were frustrated by the writing style and personality changes in GPT-5. Others were angry that models they had grown attached to were quietly phased out. Sam said OpenAI didn’t fully understand how emotionally attached people had become to the product. He 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.”
He also talked about possible new features for ChatGPT, like encrypted chats to improve privacy. There’s no official launch date for that, but it’s something the company is considering.
Another topic that came up was OpenAI’s interest in buying Google Chrome if a federal court forces Alphabet to spin it off. Sam confirmed OpenAI still wants it. Owning the world’s most-used browser would give OpenAI a direct line to billions of users.
Still, Sam acknowledged that the hype has created unrealistic startup valuations. Some companies are pulling in absurd amounts of money without delivering results. “Someone’s gonna get burned there,” Sam said.
If you're reading this, you’re already ahead. Stay there with our newsletter.
