Elon Musk provided erratic testimony in court this week, facing intense cross-examination regarding his departure from the OpenAI board and his efforts to influence the non-profit's direction, according to reporting by theverge.com.
During the proceedings, Musk's demeanor shifted from a self-described heroic figure to a combative witness. While his direct testimony initially showed improvement, theverge.com reported that his cross-examination was "absolutely miserable."
Defense lawyer William Savitt challenged Musk on inconsistencies between his current testimony and previous depositions. The report noted that Musk frequently refused to provide simple yes or no answers and occasionally scolded Savitt during the exchange.
Allegations of a Tesla-OpenAI merger
The cross-examination suggested that Musk ceased his quarterly payments to OpenAI because he failed to secure full control of the organization. According to theverost.com, Musk initially sought four board seats and 51 percent of the company's shares.
Documents and testimony indicated that Musk's plan involved a board of seven members, where he would have maintained primary control. When this arrangement failed, Musk reportedly pulled his funding commitments and hired Andrej Karpathy, a key OpenAI engineer, to Tesla in 2017.
Musk testified that he believed Tesla was the only entity capable of competing with Google. "In my and Andrej’s opinion, Tesla is the only path that could even hope to hold a candle to Google," Musk said on the stand.
By 2018, Musk had expressed concerns via email that OpenAI was on a "path of certain failure." He proposed merging Tesla with the non-profit, a plan that never materialized before his resignation from the board.
Musk maintained that his primary motivation was protecting the non-profit mission. He accused OpenAI leadership of "stealing a charity" and "looting a non-profit," claiming he was a "fool" who provided $38 million in funding that eventually helped create an $800 billion for-profit entity.
However, theverge.com noted that Musk's testimony also revealed long-standing doubts about the non-profit structure. In a 2016 email to a Neuralink colleague, Musk wrote that setting OpenAI up as a non-profit might have been a mistake due to a lack of urgency.
When confronted with these statements, Musk characterized the inquiry as hypothetical. The cross-examination concluded with Musk accusing the defense counsel of using "unfair questions" designed to trick him.
