Elon Musk calls himself a fool for giving OpenAI $38 million that became an $800 billion company

Elon Musk and Sam Altman squared off in federal court in Oakland this week in what may be the AI industry's most consequential trial yet.

Musk is suing OpenAI to unwind its shift from a nonprofit research lab into a for-profit company. He's also pushing for the removal of Altman and OpenAI president Greg Brockman. Musk testified for more than seven hours over three days, according to Reuters.

"I was a fool who provided them free funding to create a startup. I gave them $38 million of essentially free funding, which they then used to create what would become an $800 billion company," Musk said, according to MIT Technology Review. He added that OpenAI was "specifically meant to ​be for a charity that does not benefit any individual person. I could've started it as a for-profit and I specifically chose not to," according to Reuters.

Musk argued that OpenAI wouldn't exist without him: "I came up with the idea, the name, recruited the key people, taught them everything I know, provided all of the initial funding," he said, according to Reuters. He also said he poached researcher Ilya Sutskever from Google, adding, "After I recruited Ilya to OpenAI, Larry Page refused to speak to me ever again."

xAI appears to have distilled OpenAI's models

Musk's cross-examination by OpenAI lawyer William Savitt got tense, according to Reuters. Savitt argued that Musk only took action once ChatGPT took off. When Savitt confronted him with his own emails showing that Musk had backed a for-profit structure and a Tesla takeover, Musk snapped: "Your questions are not simple. They're designed to trick me," adding that "few answers are going to be complete especially when you cut me off all the time."

Musk also warned of a "Terminator" scenario. "The worst-case scenario is a Terminator situation where AI kills us all," he said, according to MIT Technology Review. In a separate remark, he said, "If we build the robots, I can make sure that they're safe, and we don't have a Terminator future situation," according to The Ringer. The judge later barred further discussion of existential AI risk.

Musk also admitted that xAI "partly" distills OpenAI's models, according to MIT Technology Review. That's less a bombshell than a confirmation of what was, at least early on, fairly obvious. Musk defended the practice, saying, "It is standard practice to use other AIs to validate your AI," according to Reuters. It's still unclear whether, and to what extent, xAI used OpenAI models to train its own systems.

The trial's outcome could put OpenAI's planned IPO at risk. OpenAI co-founders Sam Altman and Greg Brockman, along with Microsoft's Satya Nadella, are among those expected to testify next week.