The president of OpenAI does ‘everything,’ except answering the question

The most powerful piece of evidence in Elon Musk’s case against OpenAI so far has been Greg Brockman’s magazine. Brockman himself is running as a close second.

Brockman was called to the stand in an unusual way – he was asked questions, then directly tested – and he had the opposite power of high school teams. There was a lot of “I wouldn’t say that,” “I wouldn’t say that,” and “It sounds like what I wrote.” After Musk’s attorney, Steven Molo, read some testimony aloud, Brockman gently corrected him if he dropped a word, even if the word was “a” or “the.” When asked if Microsoft’s $10 billion investment was the most financial thing for OpenAI, Brockman replied that it was only $10 billion. Come on. on.

I’ve said it before if you mean the word “epistemology,” you don’t have to testify in your own defense. So the lawyer jumped the gun – is it worth taking the time of the judge to tell us all this? Save being the smartest kid in the world to your parents.

That would be bad enough. But the journal articles – a series of written files from his computer – were worse, because they were clear about the greed and opportunity of Brockman around 2017. Here is one: “btw another insight from this is that it would be a mistake to steal something that is not profitable for him. Here is another one: “Maybe we should turn to profit. Making money for us makes sense.” There is also this: “we cannot say that we are dedicated to non-profits. do not want to say that we have given up.

“It would be wrong to steal something worthless from him” is very close to Musk’s “stealing charity” line, I realize.

We haven’t completed a direct test, so I’m sure we’ll hear something specific about the events that inspired this post. But between Brockman’s thoughts on the cross and the magazine articles, I don’t think I can trust him to search my purse and use the restroom.

Musk’s team is trying to paint Brockman as greedy, which I buy. The famous “What will it take me to $1B?” from Brockman’s magazine appeared. We discovered that Brockman’s share of OpenAI’s profits was worth $30 billion. Molo asked Brockman why he didn’t give $29 billion to the non-profit arm of OpenAI if $1 billion was enough for him.

Brockman could have said something like, “If I lost everything I’ve ever owned at once, OpenAI would get less than $39 billion, because that’s how supply and demand works.” He might say, “It’s an important signal to other investors that I have some skin in the game.” Or maybe, “That’s my only money on paper.”

He didn’t do any of this. Brockman responded bluntly about the amount of money that wasn’t used to make a profit. Molo said he did not answer his question and asked again. We went back and forth on this for quite some time; the judges’ heads bobbed back and forth as if they were watching a tennis match. Brockman did not answer the question.

Nothing was too small to argue with. Molo asked if the purple boxes were an OpenAI feature that was often used to attract attention, and Brockman said no. Then we all read in the document that OpenAI often uses them in the letters of employees and vendors to show important things. In my notes I wrote, “Why are we fighting the purple box?”

Molo brought up another big problem by bringing up the various contracts that OpenAI had with companies that Brockman was involved with: Cerebras, CoreWeaveStripe, and Helion Energy. Given the number of companies using Stripe, its OpenAI mission seems daunting — but OpenAI’s commitment is important to Cerebras and CoreWeave.

Brockman also has a direct relationship to Altman because of the compensation he was given when he started OpenAI. He owns a 1/1 share in the Altman family office, which Brockman acquired instead of Y Combinator stock because “we ran out of Y Combinator stock fulfilling other (employees).” In a 2017 email, Musk’s boss, Jared Birchall, wrote to Musk that Altman had explained this to him, and Musk forwarded Birchall’s email to Brockman with a “??” Apparently Musk was unaware of the deal, and Brockman had to explain.

I’m looking at the different ways Brockman made himself unreliable because it’s fun to see strong men in action. But it has also changed my mind about his direct testimony, which began later. Brockman began by telling the story of the launch of OpenAI that seems to have been polished from a thousand podcasts and keynotes. When asked what he did as president of OpenAI, he replied, “I do all the things.” If we weren’t in court, I would have screamed. The Millennial word is a big disaster.

That being said, OpenAI was the brainchild of Brockman and Sam Altman. Brockman told Altman about his interest in AI when he left Stripe (“I’m thinking about doing an AI thing,” to which Altman apparently said, “I’m also thinking about doing an AI thing.”). They were connected. The original idea was to have a research arm of Y Combinator, which Musk shot down because he did not want to be associated with Y Combinator.

Imagine fancy dinners, trips to Napa (“our van was parked for an hour and a half and no one noticed” because the conversation was so good), AI meetings. On this! It was so beautiful-oh! Everyone was very cooperative and creative! We were treated to a long recap of Ilya Sutskever looking at leaving Google, then a photo Brockman took on the first day of OpenAI, everyone working from their homes. (Pictured: Altman. Missing: Musk.) I think you’re on your way; I definitely have Brockman’s. This was the son of Altman and Brockman. Just after Musk finished the phone call with the group Altman and Brockman had gathered when Musk told them he wanted to participate.

Musk appeared in testimony as aloof and intimidating at times. At dinner, he asked if Denis Hassabis of Google was evil. In fact, Musk appeared to be “steady and focused” on Hassabis, and he did not even mention Larry Page, himself. in Musk’s words that’s why OpenAI came to be. In text messages from Sutskever to Brockman, Sutskever wrote, “Elon can stay with us half a day a week.

Sutskever was right to complain; Musk is very difficult. I think we’ll hear more about that tomorrow. But as it stands now, the court will have to decide which of the two unreliable men it trusts more. I don’t admire the work.