Elon Musk Testifies That He Started OpenAI to Prevent a ‘Terminator Outcome’

Elon Musk and Sam Altman appeared successful a national courtroom unneurotic for the archetypal clip connected Tuesday arsenic they combat implicit OpenAI’s decade-long improvement and what it means for the company’s future.

The proceedings successful Musk’s suit against Altman could effect successful fiscal damages and, much significantly, governance changes astatine OpenAI that whitethorn complicate its plans for an archetypal nationalist offering arsenic soon arsenic this year.

As the archetypal witnesser connected the stand, Musk instantly sought to framework his lawsuit arsenic much than conscionable astir OpenAI. Siding with Altman “will springiness licence to looting each foundation successful America” and shingle the “entire instauration of charitable giving,” Musk told a sheet of 9 jurors advising US District Judge Yvonne Gonzalez Rogers connected however to rule.

Musk has been acrophobic astir computers becoming smarter than radical “since helium was a young antheral successful college,” his lawyer Steven Molo told jurors. Molo explained that Musk lobbied governments to walk regulations addressing the imaginable of alleged artificial wide intelligence, including gathering with then-President Barack Obama successful 2015. “But the authorities was not stepping up,” Molo said. “Elon felt helium had to bash something.”

Around the aforesaid time, Musk met with Altman, a then-30-year-old capitalist “whom helium didn’t cognize precise well,” Molo said. They soon launched OpenAI unneurotic arsenic a nonprofit. Google’s unchecked advancement connected AI improvement had sparked concerns for some OpenAI cofounders, and they wanted to make a competing laboratory with a greater absorption connected safety. “My position is [OpenAI] exists due to the fact that Larry Page called maine a speciesist for being pro-humanity,” Musk said, referring to the Google cofounder. “What would beryllium the other of Google? An open-source nonprofit.”

While Musk believes AI could cure diseases and make prosperity for humanity, helium besides told the tribunal that helium thinks the exertion could veer disconnected into catastrophic scenarios consecutive retired of subject fiction. “It could besides termination each of america … the Terminator outcome. I deliberation we privation to beryllium successful a movie … similar Star Trek, not a James Cameron movie,” Musk said. (While Musk has agelong raised alarms astir AI safety, his existent firm, xAI, has been criticized by researchers astatine different AI labs for its "reckless" information culture.)

As OpenAI began notching immoderate of its ain successes, Musk and Altman agreed that a for-profit limb with fixed returns for investors was indispensable to rise bonzer sums of wealth needed to money hiring and computing, according to Molo. He compared it to a nonprofit depository that receives immoderate proceeds from a for-profit store. “I was not opposed to determination being a tiny for-profit arsenic agelong arsenic the process didn’t wag the dog,” Musk said connected the stand.

Musk felt that the attack had gone excessively acold erstwhile Microsoft, different suspect successful the trial, agreed to put $10 cardinal successful 2023, and OpenAI progressively moved intelligence spot and unit to the for-profit company. “The depository store sold the Picassos truthful they were locked up wherever nary 1 could spot them,” Molo said.

OpenAI’s Rebuttal

William Savitt, an lawyer for OpenAI, told jurors that OpenAI ne'er promised Musk that it would stay a nonprofit and people each its code. “The grounds present volition amusement what Musk says happened did not happen,” Savitt said.

He added that Musk knew astir plans to rise firm concern exceeding $10 cardinal arsenic acold backmost arsenic 2018. Musk adjacent raised concerns astir Microsoft’s engagement successful a 2020 tweet. But helium didn’t record a suit until helium founded a competitor, xAI, successful 2023.