Musk’s great loyalty was his biggest liability

I sat down Musk v. Altman today, you painfully knew that no one would ask Shivon Zilis the question on everyone’s mind: Girl, what is this? fuck He is to do?

Zilis, who testified under oath that she is the mother of Musk’s four children, was… what’s the best way to put this? Musk’s mentor? He denies he was “chief of staff” but says he worked on “Musk’s entire AI portfolio: Tesla, Neuralink, and OpenAI” starting in 2017. The two met through OpenAI, and had what he calls a “close-up” before becoming “friends and colleagues.” “One,” he confirmed, “was in a romantic mood.”

His job under Musk was to “find problems and solve them,” and he says he worked 80 to 100 hours a week doing that. “It was just a banana,” she said. Musk’s first two children – twins – were born in 2021, and Zilis served on the board of OpenAI. He kept this a secret. He did not tell the board who the father was until Business Insider he reported in court papers that listed Musk as his father.

“I called my father first,” said Zilis, who revealed that even his own family does not know who the children are. “The call after that went to Sam Altman.” Greg Brockman, president of OpenAI, testified that he found out about Zillis’ children from news reports. When she spoke to him about it, she said that her relationship with Musk was “platonic” and that she had children through IVF. This was confirmation enough for Brockman, who has been his friend since 2013. He stayed on the board.

On the stand, Zilis spoke softly and quickly. He seemed worried. A key part of what makes his testimony so damning for Musk is that he appeared to be the only person taking notes on what Brockman, Altman, Ilya Sutsekever, and Musk were discussing when the computer inventors were considering their options for creating a for-profit arm of OpenAI. He was also “assisting and facilitating communication between the main parties.” Those notes are the most important evidence in the case – more important, even, than Brockman’s diary.

The purpose of the direct evidence seemed to be to remove a loophole from what Zilis and the plaintiff’s lawyers knew was coming. So he told the court that his role also meant telling Altman when Musk was in a “good place” in the conversation — perhaps corroborating Brockman’s testimony yesterday that at one point he feared Musk would physically attack him — while adamantly denying that he passed the information on to Musk.

See, she and Musk testified that they live together and have a loving relationship with four children. He was originally the plaintiff in the case. He kept the paternity of his children a secret to his father. All of those things would be reason enough to doubt his testimony by assuming that OpenAI disclosed its mission during the chaos when Altman was fired by the board. He said Microsoft CEO Satya Nadella said something about the effect that “we are above them, we are below them, we are next to them” in a time of chaos that is “terrifying.” (The quote was “Below them, above them, around them.”)

But the notes are actually Musk’s indictment. Try as he might, Zilis could not explain them.

There were many ideas that happened in 2017 and 2018. We saw many emails from Zilis since then. Notably for one, the option was “to transition to profitability in the next few weeks (wow fast!).” Another e-mail noted that the “total non-negotiable” of Altman, Brockman and Sutskever was “a firm agreement that Elon (or anyone else) should not be with him at all.” [sic] to control the AGI they are building.” In another he wrote to Musk’s money manager, Jared Birchall, “They said they would not move forward without a guarantee from him in control.” You and I can argue that what we want is stupid but they stick to it.”

Zilis also knew about Musk stopping donations before OpenAI did. On August 20, 2017, he wrote, “Funding suspension: OpenAI may have seen this week that their $5M in Q3 is on hold, however, it’s still on hold. Not sure how this will affect the negotiations but I wanted to flag it as it could have a huge psychological impact on them if they find out.” Musk told Brockman and Sutskever about a week later, on September 1, that he would withdraw the funding.

There were other plans:

We’ve already seen one of his messages in the booth – and that’s when Musk leaves the board and asks him if he should stay “close and friendly” to continue giving him information. In his direct testimony, he tried to put that into context: “They were going through this weird breakup,” he said. But at the cross, we found that he did not remember that at his deposition.

“Your old memories are coming back,” said Sarah Eddy, a lawyer for OpenAI, in one of the trial’s funniest moments. Sure, Musk’s team objected and the opposition was sustained, but we all heard it. In fact, it was one of the few times Zilis seemed to have regained memories he didn’t have at his deposition, memories that – luckily I’m sure – happened to be good for Musk’s case.

To be fair, Zilis has done pretty well under scrutiny from everyone we’ve seen so far, but he doesn’t seem to be honest. And there was even more reason to doubt him when we found out how he left the board, which – according to his input – happened “because I took a call from Sam and he said, ‘I heard that Elon is starting a competitive career’ and I said, ‘If that’s true, this is the time to resign.’

Surprisingly, he had forgotten about that call between being put down and today. But he appeared to know that Musk was going ahead with AI when he texted his friend, who was on his phone as “Shahini Rubicon Fluffer.” (Awesome name. Thomas Pynchon would be so jealous.) “You should clear the OpenAI board btw,” he wrote. “E’s effort is now well known.” His friend didn’t seem surprised by the revelation. Zilis continued: “When the father of your children starts to compete and hire without OpenAI there is nothing to do.

Zilis added that Musk “profusely apologized for cutting off my network of friends over this.”

Here’s what it boils down to, as far as I’m concerned: His primary loyalty was to Musk and to Musk. To believe that he didn’t know about xAI, I would have to believe that despite—at the time—their three children and the amount of time he spent with them every week, he never discussed it with her. I do not believe in that. Who were they? There is enough evidence in his meeting notes to suggest that he has always withheld information from OpenAI on Musk’s behalf – xAI will be no different. I also can’t believe he didn’t give Musk information about the Microsoft deals he approved while sitting on the OpenAI board.

Musk had no problem turning all of OpenAI into a profit or bringing charities to their knees by hiring its strongest researchers. He didn’t mind the idea of ​​putting it into Tesla by any means. It’s him he did the mind could not control it. That’s what I took from Zilis’s posts and emails.

Brockman and the OpenAI board were incredibly naive to allow Zilis to continue working there after learning of the paternity of his twins. However, perhaps no one expected someone so gentle to be so cunning. He was smart enough not to raise his voice or the obvious questions when he was asked questions, so his reading reads as more honest than anyone we’ve seen. It’s just that all that is taken from his written statement is that he puts Musk first in his life. Everyone else – including, obviously, his father – comes second. So in that case, you might think you’re saying what Musk wants to hear too.