Musk Demands OpenAI Return To Being Weird Little Nonprofit Nobody Understood
OAKLAND, Calif. – The federal courtroom overseeing the increasingly hostile legal dispute between Elon Musk and OpenAI briefly paused Wednesday after both parties attempted to submit the exact same 2017 Slack messages as evidence that the other man had personally destroyed humanity.
The lawsuit, which centers on Musk’s claim that OpenAI betrayed its founding nonprofit mission by becoming a multi hundred billion dollar AI company, entered its third week with testimony from executives, former board members, venture capitalists, and one exhausted court stenographer who reportedly requested witness protection after hearing the phrase “benefit humanity” 6,400 times in a single afternoon.
Musk, who helped found OpenAI in 2015 before leaving the organization, told the court he had been “emotionally and spiritually deceived” when the company evolved from a niche research lab into what he described as “basically Adobe Creative Cloud with nuclear capabilities.”
According to court documents, Musk is seeking over $150 billion in damages, restoration of OpenAI’s nonprofit structure, removal of CEO Sam Altman, and “full custody of the original vibes.”
“This was supposed to be a humble nonprofit devoted to safe AI research,” Musk testified while holding up a printed screenshot of an old group email titled AGI Brainstorm FINAL FINAL v12. “Now they’re selling enterprise subscriptions to governments and letting middle managers generate emails about yogurt.”
OpenAI CEO Sam Altman rejected the allegations, arguing that Musk had personally pushed for aggressive commercialization during the company’s early years and once proposed merging OpenAI with Tesla “so the robots would already have somewhere to work.”
“He wanted ninety percent ownership and possibly a flamethrower division,” Altman said during testimony. “There are emails where he asks if AGI can be launched during halftime at the Super Bowl.”
The trial has increasingly devolved into a forensic reconstruction of Silicon Valley group chats from 2016 through 2018, forcing jurors to spend several days reviewing old Signal messages containing phrases like “nonprofit but epic,” “ethical capitalism,” and “we should maybe own New Zealand.”
Court reporters confirmed that at least two jurors visibly aged during testimony from former OpenAI executives regarding governance structures, capped-profit models, and a 47-minute argument over whether the phrase “open” in OpenAI was legally binding or “more of a spiritual adjective.”
One pivotal moment came when Microsoft CEO Satya Nadella testified that Microsoft’s billions invested into OpenAI were “not charitable donations,” but instead “normal business investments intended to gently reshape civilization.”
The courtroom also heard testimony from AI safety researchers who warned that both parties appeared less interested in protecting humanity than in controlling whichever machine eventually becomes powerful enough to replace it.
“This case has really highlighted the core issue in artificial intelligence,” said Stanford researcher Daniel Kravitz outside the courthouse. “Nobody involved seems emotionally capable of losing an argument.”
The judge repeatedly struggled to keep proceedings on track after Musk attempted to explain the plot of Terminator 2 for the fourth time this month.
At one point, attorneys for OpenAI presented evidence showing Musk had texted Greg Brockman just days before trial seeking a settlement involving “shared AGI governance” and “possibly an X-branded moon monastery.”
Legal analysts say the trial could fundamentally reshape the future of artificial intelligence governance, corporate law, and whether any human being should legally be allowed to use the phrase “save humanity” while wearing a $3,000 leather jacket.
By Wednesday afternoon, the court had reportedly approved emergency funding for an additional twelve terabytes of evidence, most of which consisted of old podcasts where both men accurately predicted that the other would eventually become intolerable.