On the stand, Elon Musk can’t escape his own tweets

Elon Musk appeared in a federal court in California on Wednesday, presenting arguments in a case against Sam Altman and his co-founders, alleging that they misappropriated a charity initiative. However, the day took a surprising turn when Musk, under oath, stated that Tesla is not currently engaged in the pursuit of artificial general intelligence (AGI), directly contradicting a tweet he had made just weeks prior. The lawsuit Musk filed challenges the framework of OpenAI, claiming that Altman and others deceived him into supporting a non-profit organization, which subsequently transitioned into a for-profit entity that now dominates the AI landscape. Throughout a lengthy and at times contentious testimony, Musk's case hinges on how jurors and Judge Yvonne Gonzalez Rogers perceive the distinction between investors with capped profits versus those with unlimited returns. Musk recounted his initial trust in the founders of OpenAI, which included Altman, Ilya Sutskever, and Greg Brockman. Over time, however, he grew wary of their intentions, ultimately concluding that they were misusing the nonprofit's resources. During cross-examination, OpenAI's attorney William Savitt attempted to complicate Musk's narrative, illustrating that Musk had previously supported the organization’s shift towards a for-profit model to secure funding against competitors like Google. Musk acknowledged that he had contemplated converting OpenAI into a for-profit entity as early as 2016, expressing intentions to maintain majority control. When those plans fell through, he ceased regular donations, although he continued to fund office space until 2020. He emphasized the critical difference between profit-capped investors like Microsoft and those with no limits on earnings, asserting that this disparity contributed to his decision to file the lawsuit. The cross-examination also delved into Tesla's AI capabilities, with Musk repeatedly finding himself at odds with his own statements on social media. When pressed about Tesla's focus, he confirmed that the company is not working on AGI, despite previous claims that it would be among the leaders in this field. Additionally, he was challenged on the extent of his financial contributions to OpenAI, asserting that his reputation compensated for the difference between his claimed investment of $100 million and the actual $38 million. The court also reviewed Musk's attempts to recruit OpenAI staff for Tesla and Neuralink, revealing a complex web of interactions while he was still on OpenAI's board. A significant aspect of Musk's argument is the assertion that OpenAI's shift to a corporate model poses societal risks by compromising safety priorities. However, Savitt prompted Musk to admit that all AI companies, including his own, face similar safety challenges. As Judge Gonzalez Rogers indicated that this line of questioning would continue, the focus shifted to the implications of AI development on public safety. Musk's legal team hinted at discussing the controversial role of AI in recent tragic incidents, although the judge expressed her desire to keep the conversation centered on safety measures within AI companies. Musk is set to return to court on Thursday for further questioning, with additional testimonies expected from key figures including his family office manager and experts in AI safety.