OpenAI's Brockman Says He Feared Musk Hit in 2017
OpenAI president Greg Brockman testified in federal court that he feared Elon Musk might strike him during a tense 2017 meeting. Brockman, a co-founder of the AI company, shared the details while defending against Musk's lawsuit. The case seeks to reverse OpenAI's move to a for-profit structure.
The testimony occurred in Oakland federal court during the second week of a month-long trial. Musk, an early backer and co-founder, sued OpenAI leaders including chief executive Sam Altman. Brockman faces claims tied to the company's business changes.
Tense 2017 Meeting Over Control
Brockman described rejecting Musk's push for greater influence at OpenAI. Musk's demeanor shifted suddenly after the refusal. "I actually thought he was going to hit me," Brockman told the jury, referring directly to Musk.
The exchange ended quickly. Musk then declared he would stop providing funds to OpenAI. He had supported the organization since its founding in 2015. OpenAI started as a non-profit focused on safe artificial intelligence development. Co-founders included Musk, Altman, Brockman, and Ilya Sutskever.
OpenAI gained massive attention with ChatGPT's release in late 2022. That product propelled the company to one of the world's most valuable tech firms. Musk departed the board in 2018 amid disagreements over direction and funding needs.
Musk's Efforts to Gain Influence
Before leaving, Musk worked to increase his authority over OpenAI. He tried to win over Brockman and Sutskever with favors. OpenAI lawyers presented text messages from August 2017 between Sutskever and Brockman. One read: "Will a model 3 make you be willing to accept massively unfavourable terms?"
Stay updated
Get the day's AI and automation news in your inbox. No spam, unsubscribe anytime.
Brockman's account centers on Musk's knowledge of OpenAI's evolution. The company began as a non-profit. It later created a for-profit subsidiary to attract billions in investments. Last year, leaders made the for-profit entity the main focus to support rapid growth in AI research.
Musk has criticized this path publicly. He argues it strays from OpenAI's original open-source, safety-first mission. The lawsuit claims Musk received assurances the structure would stay capped-profit. OpenAI counters that Musk approved early plans and sought CEO role himself.
Connections to Shivon Zilis
Brockman's court time precedes testimony from Shivon Zilis, a former OpenAI board member. Zilis is mother to four of Musk's children. She served on the board years after Musk exited.
Brockman said Zilis told him about giving birth to twins. He learned Musk was the father through news reports later. In a follow-up talk, Zilis explained the pregnancy came via IVF. She described her relationship with Musk as entirely platonic.
Brockman noted trust in Zilis to handle conflicts with Musk. "We trusted her to keep the Elon conflict under control," he said. Zilis departed the board in March 2023. That timing matched Musk's launch of xAI, his AI venture. xAI builds a chatbot rivaling ChatGPT.
Zilis works at Neuralink, Musk's brain-interface company. Her OpenAI role overlapped with Musk's xAI start, highlighting ongoing ties in the AI field. The trial continues to explore these personal and professional links.
OpenAI's rise stems from breakthroughs like GPT models. ChatGPT drew millions of users fast, sparking AI investment boom. Musk's xAI aims to counter with Grok, emphasizing truth-seeking AI. The feud underscores tensions in AI governance and profits.