Sam Altman was winning on the stand, but maybe not enough

After two weeks of hearing from various witnesses that it was a fake snake, the jury heard from the fake snake: Sam Altman. At the end of the testimony, his lawyer William Savitt asked him how it felt to be accused of stealing from a charity.

“We have created, through hard work, this very large grant, and I agree that you will not steal it,” said Altman. “Mr. Musk tried to kill it, I think.

Altman was in “the good kid from St. Louis”, and he was fascinated by the man who was surprised by what was happening to him. When he came down from the scene carrying a lot of evidence, he looked like a schoolboy. He looked nervous at the start of his direct testimony, though he warmed up quickly. Overall, he seemed to give credible evidence – and at times, it seemed like the jury liked him.

Throughout this trial I have struggled to figure out what the jury is making of all this because I know the people who are testifying very well. I have heard some swearing lies, like time Elon Musk he told us everything that he is not angry. (Then he continued to get angry by asking him questions.) Or like when Shivon ZilisThe mother of his children, told us that she did not know that Musk was starting xAI – which seems to directly contradict his messages. Or when Greg “What will get me to $1B?” Brockman He told us that it was all about the mission. I believe that Altman is not trustworthy – I mean, The New Yorker published more than 17,000 words about the number of his lies. But unlike Musk, there is contemporary literature that supports Altman’s version of the story. At least, mostly.

After OpenAI Dota 2 victoryThe discussion of the profit making group started in earnest. “Mr. Musk felt very strongly that if we were going to make a profit he had to have full control over it first,” Altman said. “He just trusted himself to make decisions that he didn’t know would be right.”

Altman testified that he was not happy with Musk’s insistence on control, not because Musk was not involved like everyone else, but because OpenAI was there so no one could control AGI. And at Y Combinator, the startup where he was president, Altman saw many management battles; no one wanted to give up power when things were going well. With structures such as supervoting shares, founders can retain control indefinitely. Surprisingly, Altman’s example was not very popular (Mark Zuckerberg in Meta); it was Musk and SpaceX. When Altman asked Musk about OpenAI’s successor plans, he got a particularly “hair-raising” answer: After Musk’s death, Musk said, “I never thought about it, but maybe control should go to my children.”

I don’t know about that. But I know that I saw a 2017 email from Altman to Zilis in which he wrote, “I’m worried about control. I don’t think that one person should be in control of the first AGI in the world – in fact, the reason we started OpenAI was that it wouldn’t happen.” He went on to say that he didn’t care about control right away and was open to “architecture” – which I understood to mean that, in order to establish Musk, Altman was ready to give him control over the main events of the company.

“My belief is that he wanted to be in power for a long time and that we would have agreed to what he wanted,” Altman said on the stand. This sounds about right. In a later video testimony from Sam Teller, we learned that Musk no longer does anything he can’t control. This is also in line with Musk’s long-term plan to ensure that he cannot be fired from his company as he was fired from PayPal.

Musk also tried to recruit Altman to Tesla. We saw a text between Altman and Teller, where Teller told Altman that Musk was committed to promoting Tesla’s AI no matter what, and that he hoped that Altman, Brockman, and Ilya Sutskever would want to join him eventually. “I read vaguely, like, a faint threat in there, that they’re going to do this inside Tesla with or without you,” Altman said. But he saw that Tesla was essentially a car company – allowing it to acquire OpenAI would defeat the purpose of OpenAI.

Later, in Teller’s testimony, we saw the text that Teller sent to Zilis at 12:40 AM on February 4th, 2018: “I don’t like OpenAI to continue without Elon,” he wrote. “I would like to stop recruiting leaders.”

After Musk suspended his quarterly offering, OpenAI was working on “shoes” and “very short money.” OpenAI had other sponsors, none of whom built it or joined Musk’s suit. (One exhibitor not subpoenaed was Alameda Research, the company is Sam Bankman-Fried (who is now in prison for fraud and embezzlement.) Musk’s resignation from the agency meant that “people were wondering if he was going to try to get back at us, or get back at us. On the other hand, Altman said that Musk “let down our key researchers” and did “Severe long-term damage to organizational culture.” So it seems that some people felt relieved by removing him.

We saw a lot of evidence that all the while Altman was setting up the profitable arm of OpenAI, he kept Musk in the loop, either directly or through Zilis or Teller. At no time did Musk deny it, and whatever he said publicly about Microsoft’s finances, there was plenty of evidence that he was informed privately.

During the interview, we received more than 10 minutes in which Steven Molo told Altman that various people called him a liar: Sutskever, Mira Murati, Toner, McCauley, Daniela and Dario Amodei (former OpenAI employees and founders of Anthropic), employees of Atlman’s Loop for the first time, recently. The New Yorker story, book called The Optimistetc. Molo made other points by asking Altman about the evidence in the case, which Altman said he did not care about. Molo acted as if this was impossible. Of course someone did he inform Altman of what was said?

It was a little funny and a little boring. Altman remained silent, looking distraught and confused as if he was lying. It was also the most successful part of the cross, which lost focus later. I’ve seen a lot of bad defense from Musk’s side all along, and today was the worst. At one point, Molo is trying to take advantage of Altman being CEO and on the board of the company, Altman said – as a matter of fact – that CEOs are always on the boards of the companies they run.

(At this point in my notes, I had written, “Boy, Molo is not very good at this.”)

There was also an uncertain argument about fundraising in non-profits, specifically that if Stanford can raise $3 billion a year, OpenAI should remain non-profit. Well, let’s just think about that for a minute. Stanford has a network of thousands of undergraduate donors. It is a school, which has very different capital needs. It is not competing with well-known for-profit companies. But leave it at that and think that another financial expert took part in the OpenAI Foundation: $ 3 billion is the two original Microsoft products combined, and it is not enough for OpenAI to reach its current position. If compute is the biggest barrier to developing AI models, then Molo’s arguments show that OpenAI could not succeed as a non-profit on its own. He is making a defense case for them.

But the truth is, Molo should not succeed in this project, because the point of this experiment is not to win – although I am sure that Musk will not be happy to win. The goal is to punish Altman, Brockman, and OpenAI. Musk has done a great job – convincing people that Atlman is a liar and a snake. this morning, I read the exclusive in The Wall Street Journal which included Republican AGs and the House Oversight committee wanted to look into Sam Altman’s finances. References to the experiment are presented throughout this article.

So yes, Altman was a convincing stand-in. He might even win a suit. But it seems that Musk’s revenge has just begun.