Today I’m talking with Liz Lopatto, who spent the last month covering the Musk v. Altman trial in all its chaos. You’ll hear her describe the courthouse as a “zoo” and explain that there were protests of one kind or another happening outside every day.
Both Elon Musk and Sam Altman are big personalities, and people have a lot of feelings about both of them and the AI industry. And in the end… nothing happened! The jury found that Elon had filed his lawsuit after the statute of limitations had run out. You’ll hear Liz explain exactly what’s going on there.
Beyond that, the trial was nominally about OpenAI’s conversion to a for-profit entity from a nonprofit one and if the way OpenAI went about it cost Elon Musk money. But really, the suit seems mostly to have been about Elon Musk being mad at Sam Altman — or at OpenAI, for being successful without him — and wanting him punished in some way.
So in a room full of untrustworthy, unreliable people all fighting with each other, did anyone even have a reputation left to lose? Is there a floor?
Okay: Liz Lopatto on Musk v. Altman. Here we go.
This interview has been lightly edited for length and clarity.
Liz Lopatto, you are a senior chaos reporter here at The Verge. You just covered the Sam Altman v. Elon Musk trial. Welcome to Decoder.
Thank you. Always a pleasure to be here. I feel like it’s always some new, relatively insane thing that we’re talking about.
We have to stop meeting under these circumstances.
I think these are your favorite circumstances.
They are my favorite circumstances.
A few times a year, we drive you absolutely batty by sending you to cover something, and this trial was 100% one of those situations. The copy got increasingly unhinged. I think the audience liked it. But you were in the courtroom for the majority of Musk v. Altman. You got to see a bunch of the testimony live as these guys took the stand, as Mira Murati and others took the stand.
Today I’m talking with Liz Lopatto, who spent the last month covering the Musk v. Altman trial in all its chaos. You’ll hear her describe the courthouse as a “zoo” and explain that there were protests of one kind or another happening outside every day.
Both Elon Musk and Sam Altman are big personalities, and people have a lot of feelings about both of them and the AI industry. And in the end… nothing happened! The jury found that Elon had filed his lawsuit after the statute of limitations had run out. You’ll hear Liz explain exactly what’s going on there.
Beyond that, the trial was nominally about OpenAI’s conversion to a for-profit entity from a nonprofit one and if the way OpenAI went about it cost Elon Musk money. But really, the suit seems mostly to have been about Elon Musk being mad at Sam Altman — or at OpenAI, for being successful without him — and wanting him punished in some way.
We’ll start at the high level. I think the audience probably knows that Elon Musk lost, but what was this case about and what were the vibes in the courtroom?
There are two things that we should distinguish. There was what the case was ostensibly about, and then there was what the case was actually about, and those are two entirely separate things.
Ostensibly, the case was about the violation of a charitable trust.Elon Musk had donated a bunch of money to OpenAI Foundation, and then they created a for-profit, and he thinks that’s a violation of his charitable trust. He also thinks that the timing of that was right around what is known as “the blip,” when Sam Altman was briefly removed and brought back. Put a pin in that. It’s going to be important here. That’s what we’re ostensibly there for.
Verge subscribers, don’t forget you get exclusive access to ad-free Decoder wherever you get your podcasts. Head here. Not a subscriber? You can sign up here.
Because it was around the blip, Microsoft was accused of aiding and abetting, and Microsoft very quickly became my favorite part of the case.
In reality, there had been so many changing legal strategies around this. This case was filed I think two years ago in state court and then withdrawn and then put in federal court. There’s just been a myriad of things that have shuffled around since then, including a charge that got dropped right before we went to court.
So to me, the main point of this was punishing Sam Altman and maybe trying to kneecap OpenAI. And this is a case where the two worst people you know are fighting so it’s kind of hard to root for anyone. The most common response that I tended to get when I would talk about this to people or when I would post about it on social media was like, “Can they both go to jail?” So that’s kind of the vibe.
The courtroom was a little bit of a zoo during Musk’s testimony. We had one woman who got called down in front of the courtroom by the judge and chewed out because she had been taking photos in the courthouse. On the very last day, we had a guy who was ejected because he had been recording the proceedings in the courtroom. There were some shenanigans.
Every time we would leave the courthouse, there would be some kind of protest going on, usually behind the lawyers as they were trying to give their daily summary and spin what they had done in the courtroom, and then parading behind them would be a guy in a Cybertruck holding an “Elon Sucks” sign.
Perfect.
So that was what that was.
I want to come to the legal issues and particularly the ruling from the jury, as there’s a lot of mechanics there. I just want to stick on a point that the goal here was for Elon Musk to punish Sam Altman, and connect that to the protests and the comments you’re getting on social media, and certainly the comments we get every time we publish anything about AI. Is there any reputation left to damage for Sam Altman or the AI industry as a whole? Because it seems like both of these guys are at all-time lows. I’m thinking about jury selection when the judge had to just say, “It seems like no one likes Elon Musk, but we’re going to have to trust that the jury will be fair.” What’s even left to take away here?
There’s no floor about these things. I also view Sam Altman as untrustworthy, which is one of the things that this trial was really driving home as one of the points that Elon Musk’s lawyers were making, and I agree. I also think everybody else in the trial was totally untrustworthy. It was not just Sam Altman, it was all of them.
One of the things that I found myself thinking about was that the person who really got damaged the most was Mira Murati who, at least as far as I know, didn’t have a reputation as being somebody who was untrustworthy, or conniving, or whatever. And then in testimony from former OpenAI board members, we found out that she was one of the reasons that Sam Altman got fired and then was immediately texting Sam Altman like, “Oh, no, Sam, it’s very bad. It’s very bad, Sam.” You remember during this blip that Altman was fired for a pattern of being untrustworthy or something.
It was “he was not consistently candid with the board,” which could have meant anything.
Anything! And the thing that I remember, because I gossip with a bunch of journalists and we are ferocious gossips, is all of us were like, “Oh, he did something illegal. Let’s find out what illegal thing he did.”
As far as I can tell, no, he didn’t. It was just that he was engaging in what I would characterize as relatively normal executive shenanigans, where you are maintaining your control of the company by pitching your subordinates against each other — a strategy that is widely used in corporate America, by the way.
So she wouldn’t tell people that she was involved in his removal. She was the interim CEO, and then publicly supported him, and then publicly was involved in bringing him back.
Someone on the stand, I don’t remember who, said Mira was waiting to see which way the wind would blow and didn’t realize she was the wind.
That was Helen Toner, who was one of the board members who stepped down in this debacle. Because obviously as this proceeded, it became clear that by firing Sam in the way that they had fired him, they had jeopardized the entire company. One of the things that I thought was really interesting from Sam’s testimony — that I did believe, by the way — is that he thought about just taking a job at Microsoft and getting paid and not having to deal with any headaches anymore. I can certainly imagine after having been really publicly and embarrassingly fired, and having gone through all of the annoying things that one goes through as a manager and especially as a CEO, just being like, “You know what? I just want a paycheck.”
Who among us has not thought about retiring to a comfy job at Microsoft?
Right? And so when he was talking about that, I was like, “Yeah, actually, I believe that. That sounds real.” Then he obviously changed his mind.
But one of the things that I thought was really interesting about that is that we found out Helen Toner, who we saw in deposition testimony, was involved in potentially trying to sell OpenAI to Anthropic, a company that she has some ties to through the Effective Altruism movement. So again, no one here comes off looking good. I thought for a while that Helen Toner was maybe the most reliable witness we had heard from and then in the cross on the deposition it was like, “So tell us about your relationship with Anthropic.” And I was like, “Awww.”
That’s actually the thing that struck me about this entire trial. Helen Toner being wrapped up in Anthropic is one thing, but the entire AI industry at the top is 10 people who are wrapped up in each other emotionally, professionally. They’re writing each other obsequious emails, particularly to Elon, just full of flattery and praise about how great everyone is.
The idea that they’re going to make AGI is taken for granted in some way. These are the leaders of a new religion in a real way, you can see it, and they all lack any management instincts or emotional maturity to deal with the kinds of tasks that are put in front of them or the stakes or the money. You can just see it. It’s in the trial, it’s in the evidence, that they’re cracking under the pressure that they’re putting one another under, and there’s no outlet. In fact, the only outlet might have been Satya Nadella, who comes off as the coolest cucumber around because he’s just like, “I don’t know, is this going to make money? Don’t call me.” That’s basically his whole vibe.
Again, I loved Microsoft in this case. I’m not a Microsoft user. I am familiar with their products. Which by the way, their opening statement was so good. It was just a list of Microsoft products you might’ve used at some length.
“Remember us?”
It was fantastic. They were just like, “We’re not sure why we’re here, but you know us. We’re Microsoft. You’ve used Windows, surely. Do you like Xbox? That’s us.” So that was great.
There was really a sense that the only adult in the room at any given time was somebody from Microsoft. We saw that over and over again where Satya Nadella is like, “Don’t text me. Don’t leave a paper trail.” His emails are not especially spicy. I think the spiciest they got is something like him being like, “Well, we don’t want to be IBM and have them be Microsoft.”
This is OpenAI. He doesn’t want to be the commodity provider of data center hardware and have their software be the important thing, which is what happened to IBM and Microsoft.
That’s right. Which, by the way, totally understandable sentiment, I feel.
Especially from Microsoft. He’s like, “I know what’s happening here.”
That was the spiciest thing we got out of Microsoft. That was it.
So these are people who, in addition to having the management chops and having the sense of what you do and don’t do, were also just a little bit less dramatic. Over and over again, we’d have a witness, and there would be some really brutal and devastating cross from OpenAI. And then Microsoft would get up and be like, “Was Microsoft there? Was Satya Nadella there? Does anyone from Microsoft know anything about any of this? No further questions, your honor.”
It was a beautiful punchline every single time.
That’s very funny. So Microsoft obviously put a bunch of money into OpenAI. Nadella had that famous quote about being above them, below them, around them, referring to Azure and its dependency on Azure and how they would deploy OpenAI’s models. But eventually the trial comes down to, “Did they illegally convert this charity to a for-profit, and along the way, take something from Elon Musk?” What was the actual jury verdict on those counts?
The jury verdict was that Elon Musk filed the suit too late, and the statute of limitations had run out. And I’m going to be real with you, I think that had there not been a statute of limitations question, he still would’ve lost. This was a pretty weak case.
We’re going to start with the statute of limitation stuff because that is the most relevant. And then I will walk you through all the rest of it because we did do all of this in exhausting detail for the last month of my life.
One of the things that was part of Musk’s case was that he claimed that he didn’t think his trust had been violated until the blip. For this reason, he was still within the statute of limitations. The law, I believe, is that you need to file within three years. We saw a bunch of evidence that he had been read in repeatedly on the conversion to a for-profit and the various investment rounds.
I found myself unexpectedly sympathetic to Sam Altman during this trial. So congrats, Sam. He kept trying to get Elon to like him again. There would be these emails where it was like, “Hey, we’re raising this round.” Or he’d be emailing people to see what kind of mood Musk was in, if it was a good time to talk to him, because he just wanted to make sure that Elon knew what he was doing, and was it a good time for them to chat? Was Elon in a good mood? If you have a person whose job it is to tell people whether you’re in a good mood or not, I strongly feel that suggests that you maybe are difficult.
“How deep is today’s K-hole? Let’s find out before we ask for money.”
Over and over again, there was evidence of Musk being read in every single step of the way. Knowing about the Microsoft investments, knowing about the fact that they were creating this for-profit. In fact, there was a bunch of email evidence that he thought that making OpenAI a nonprofit had been a mistake, that it should have been for-profit from the jump.
There’s a ton of evidence that, separately from the timeline question, suggests that OpenAI would’ve won this case. The definition of a charitable trust, and I’m going to mangle this slightly because I am not a lawyer, is that you have to have a specific purpose for your donations. You have to have established that this is a trust, and then the next thing you have to establish is that that trust was violated.
Just looking at all of the donations, which we did in some depth, there were no strings attached that any of us saw. No one at all remembered there being any strings attached. One of the more devastating lines of testimony was that Shivon Zilis was asked, “Were there strings attached to these donations?” And she was like, “Well, not that I recall.” And then in the closing statement, OpenAI’s lawyer’s like, “Man, not even the mother of his children can corroborate his account.” Okay.
That’s brutal.
So there were no strings attached. And then we had a financial analysis that showed that money was gone very, very quickly. , tThey had spent it, because AI is expensive. And they had spent it in the way that it was meant to be spent, and all the other money that happened afterwards had nothing to do with Elon Musk. So there was that.
One of the things that I’m just going to put an asterisk on here, that I thought was interesting but didn’t write about, was that Musk had been paying the rent for OpenAI. They actually had to go back and ask him for money because Neuralink was in the building. When they got accountants to try to get their books in order so that they could proceed, the accountants were like, “Oh yeah, you can’t be supporting somebody else’s for-profit business in this building. You need to get rent money from Neuralink. They need to pay you back.”
Wow.
Not that we went into this in any depth, but my suspicion is that Musk had been taking a write-off on all of those donations on this building, and had been also taking that write-off on the space that Neuralink was using, which was why that money then had to be paid back to OpenAI.
There’s a lot here. I mean, there’s a lot of just Elon Musk, there’s infinitely complicated fractally expanding OpenAI layers of companies within the nonprofit that have board control, and people can fire Sam Altman. All of that seems enormously complex, and maybe worth some future litigation. But the jury just went with statute of limitations. And it seems like that’s maybe all they should have been talking about, if that’s what was going to end the case this quickly. Why do you think that we spent all the time in the substance and the complication when Elon had just filed too late?
I did get people asking me about this as well. “Isn’t statute of limitations a legal issue? Why didn’t the judge rule on this?”
And the answer is there was a question of fact, which was, “When should Elon have known what was going on?” And he’s saying, “I didn’t know until the blip. And so I’m within the statute of limitations.” And everybody else was saying, “e’s known the entire time. It’s over.” That was the thing that was being litigated. It wasn’t the only thing that was being litigated, but that was the one that ended up mattering: that the jury was like, “Yeah, he definitely knew all of this was happening. This is ridiculous.”
If the goal was to trash Sam Altman, of course you would pick the blip because then you get to pull every document and email and text message from the blip into the trial into evidence. You get to publish it. We published it. Was that the goal? Was Elon just saying, “I only knew about this when Sam Altman got fired,” in order to put all of that damaging evidence into the record?
I think that was the goal. I think that was what was actually going on. It was also meant to distract OpenAI, because they did have to pay this very expensive law firm to do some very expensive work to defend them. They didn’t just defend the statute of limitations. They defended all of the subclaims and all of the other things as well, which is why there is so much in our stories. They were bringing forward as much as they could to defend every single part of every possible claim because they had to.
And so, yeah, making Sam Altman look bad, distracting Sam Altman, maybe removing resources as Altman approached an IPO, those were probably the primary goals. I think Musk would’ve been happy with a win. He certainly would’ve been thrilled to force OpenAI to give up a bunch of money, even if it went back to the OpenAI Foundation, as he belatedly decided it should go. There are any number of things that I think he would’ve taken as icing on the cake, and he said that he’s going to continue this through the appeals process.
Let me just read you the quote. Elon appeared at a Forbes conference, and he said, “I think this is a dangerous precedent to set. If someone can take a nonprofit and convert it to a for-profit, that undermines all charitable giving in America.” I don’t think Elon understands how precedent works, but it seems regardless of that, he’s going to keep tying OpenAI up in litigation for as long as he can.
Oh yeah. He said something very similar to that on the stand, by the way. He has some pet phrases he likes, and “dangerous precedent to set” and “undermines all charitable giving in America” are on the list.
I think he does intend to tie OpenAI up in litigation for as long as he possibly can, bleeding them for cash, which is a strategy that we’ve seen other billionaires use. Most famously, Sheldon Adelson, who went after a Las Vegas paper, if I remember correctly. Not because they had done anything wrong — and they were in fact ruled not to have done anything wrong — but because defending the case was so financially expensive that they nearly went under. And that is a strategy you can use if you have unlimited resources: you can just bleed somebody out.
I do feel like if you’re Elon Musk and you’re really worried about rich people using their charities to enrich themselves, there are a handful of people in his direct orbit running the country that he might want to take a closer look at. This seems like he’s saying it because he just wants to keep screwing with OpenAI.
Oh, absolutely. There’s no doubt in my mind that this is personal for him. The thing that I have been thinking about for a while and am unable to quite tell is, “Is he personally pissed off at Sam Altman, or is he just affronted that OpenAI succeeded without him?”
Well, so this is my other question. Maybe you kill OpenAI and it goes away and you’ve bought yourself some time. Elon has publicly said that they built Grok incorrectly and they need to start over. They are selling a huge amount of data center capacity at Colossus 1 to Anthropic, who Elon has hated in the past, but he says, “It’s all fine now” because they showed up with a check to buy his data center capacity.
Even if you kill OpenAI, it doesn’t make xAI the winner. They’re basically starting over, as they publicly said. They’re giving up their compute capacity. What is the point of this, except to just vindictively kill OpenAI? It doesn’t seem like I can identify the competitive advantage here.
I mean, killing a competitor is not necessarily not a competitive advantage.
Let’s say OpenAI is in first or second or third or something, or just running in a different direction on the track at this point. Who knows what they’re doing. If you’re in last, it doesn’t matter. In some way, he’s helped Anthropic and Google here.
Let’s say Musk wins and OpenAI has to disgorge all this money and that potentially just blows a hole in the side of the company. I can’t rule out that Altman is enough of a deals guy that he could patch it up, but let’s say he can’t.
OpenAI is at the center of a web of deals, huge deals with places like CoreWeave and Oracle and Microsoft. Every company in the AI space is one degree of Kevin Bacon away from OpenAI. If you knock that company out, not only do you have a bunch of talent that comes free and needs a job now, which you can maybe hire, you also have created conditions where you can negotiate really favorable terms in these now suddenly open data centers with companies that now suddenly have huge holes in their revenue.
I wish I could ascribe that level of 3D chess, but there’s a part of me that says this is just personal and vindictive. And we’re going to see appeals and further campaigns about how Sam Altman stole a charity, and that will be distracting for OpenAI on one level. And on another level, they’re just going to continue selling Codex to people, because it is good at writing code, and a lot of software companies seem very taken by that. Do you think this has any meaningful effect on OpenAI in the future?
No. We knew going into this trial that Sam Altman did not have a reputation for being perfectly honest. I mean, that was the upshot of the blip. There was a 17,000-word article in The New Yorker about this. This is something that I effectively think is priced in, in the same way that Elon Musk’s, let’s say, scattershot relationship with the truth is also priced in in all of his companies. People know who these guys are, none of this is a surprise, which is why I think, again, that the person who got hurt the most here is Mira Murati, who did not have her reputation trashed before this.
So there’s going to be an appeal. These companies are going to carry on spending money. What do you think happens next? What should people be looking for? Or is this one safe to set aside for now?
I would set it aside for now. We had all the fun of going through their emails, we had their ridiculous text messages. But the biggest takeaway from the trial that matters is discovering that Grok sucks, even though Elon Musk had distilled everybody’s models. To me, that’s shocking.
Not that I am an expert in AI. It’s entirely possible that you can distill all these models and have your AI still suck. But I think that that really is a take-home point, that one of the consistent things that we were seeing in this trial was that the nerdiest of the nerds, [OpenAI co-founders Greg] Brockman and Ilya Sutskever were both like, “He’s not really serious about AI.” And I came away being like, “Yeah, he’s not serious about AI. He doesn’t know what he’s doing.”
We have all of the things that you talked about: They’re starting over from scratch, they’re leasing out their data center capacity, they’re doing all of these things that suggest that whatever Musk did with whatever billions of dollars, because I think xAI was spending… The reporting was a billion dollars a month. They’re starting over from scratch, there’s nothing, and this is even with cheating by distilling everybody’s models.
Right. This is him saying, “We didn’t build it the right way.” They didn’t actually do a proper training run, they distilled all the other models. And so they’re not on the frontier. Which, by the way, has happened to other companies. Meta is out there saying that they were not on the frontier and they started over in a meaningful way. This is a nascent industry. It’s not clear how to do these things or build these things or ship these things in a way that works.
I think my big question coming out of all of this is, boy, this handful of people that have been entrusted with spending all this money and asking for all these resources and in many ways pitching a vision in the future, they seem so immature. And even if that’s priced in, did this trial just reveal that fundamentally they’re immature and maybe you should let the Microsofts and the Googles of the world be in charge of deploying this technology, because at least the amount of bureaucracy in place at those companies will slow them down.
That could be one takeaway. Given the way that Google has destroyed its own search engine for its AI models, I’m not clear that we want to include Google in this conversation.
I’m just saying.
We’re maybe talking about Microsoft and maybe Apple. But yeah, you want grownups in charge of this technology, for sure. And the immaturity I thought was really interesting because there was a recurring theme, again that didn’t seem worth writing about separately, but that I will mention here. Over and over again, you’d get somebody on the stand and they’d be like, “Ever since I was a child, I’ve dreamed of AI. I’ve thought about the smart computer and how amazing it would be. And it kept me up at nights when I was nine years old.”
First of all, that’s stupid because that’s fiction. If you can’t tell the difference between fiction and reality, we have bigger problems. I had some childhood dreams too, and I want to be real with you, I just don’t think that owning a horse is going to be a thing that makes sense for me.
By the way, I just want to point this out. As we’re speaking, there is breaking news. Andrej Karpathy has joined Anthropic.
[Laughs] Sorry. [Laughs] Oh my God.
Which is a perfect capstone on this trial. He’s like a main character. He gets recruited to and from all these companies and now he’s at Anthropic, which seems like far and away the winner of this whole thing. Hands the cleanest, products the most successful. Why did you start laughing that hard?
A recurring theme in the trial was Musk poaching OpenAI engineers. And of course, Andrej Karpathy was one of them, because he went from OpenAI to Tesla. Because OpenAI, when it was a foundation, was asked by Elon in a way that’s suggested was not actually an ask, if you follow me, to come work on autopilot because they were having a hard time with autopilot at Tesla. And so several engineers, including Greg Brockman, went over and worked on autopilot while they were theoretically working for OpenAI. So if anybody was stealing resources from a charity, I kind of think it was Elon Musk.
One of the people who permanently stayed was Karpathy and he shows up again and again. This recruiting push that Musk made out of OpenAI while it was still a nonprofit, while he was still theoretically involved with it, while he was still theoretically on the board and had a fiduciary duty to the nonprofit, he was using it as a recruiting ground for Tesla.
That’s very good. Well, Liz, I have a feeling we’re going to keep you very busy with these characters in the year to come. My prediction is that OpenAI does not end the year looking the same as it does now, that there will be yet more change at that company.
I think that’s right.
The other little cherry that I’d like to put on top of all of this, speaking of Anthropic, is that one of my personal favorite parts of this trial occurred while the jury was out of the room. It was an evidence dispute about whether or not the jury could be shown a jackass trophy. Imagine a participation trophy that is just the back half of a donkey. And it said something like, “Never stop being a jackass for AI safety.”
It was presented to an AI safety guy who, when Musk was on the way out at OpenAI and was doing a Q&A session, was like, “Hey, it sounds like you’re really interested in speed over safety. I don’t think that’s a good idea,” and Musk called him a jackass. And so would you like to take a guess at one of the people involved in presenting that trophy?
Was it Karpathy?
It was [Anthropic CEO] Dario Amodei.
Oh, amazing. Amazing. Perfect. That tracks with everything Anthropic has stood for. Everyone’s leaving to start a safer AI company, and Dario was among the first. Perfect. Did he take the trophy with him?
He did. The lawyers had it, so I assume he’s gotten it back. We published a photo because as I was live-tweeting this, I saw people asking for a photo, so I got ahold of one, but I remain very entertained by this trophy. So hats off to the fine engineers who eventually did leave and make Anthropic, because it seems like they have a pretty good sense of humor.
Yeah, they figured it out. All right, Liz, we’ll have you back soon, hopefully under more rational circumstances, but it’s always a pleasure. Thanks for being on Decoder.
My pleasure.
Questions or comments? Hit us up at decoder@theverge.com. We really do read every email!
So in a room full of untrustworthy, unreliable people all fighting with each other, did anyone even have a reputation left to lose? Is there a floor?
Okay: Liz Lopatto on Musk v. Altman. Here we go.
This interview has been lightly edited for length and clarity.
Liz Lopatto, you are a senior chaos reporter here at The Verge. You just covered the Sam Altman v. Elon Musk trial. Welcome to Decoder.
Thank you. Always a pleasure to be here. I feel like it’s always some new, relatively insane thing that we’re talking about.
We have to stop meeting under these circumstances.
I think these are your favorite circumstances.
They are my favorite circumstances.
A few times a year, we drive you absolutely batty by sending you to cover something, and this trial was 100% one of those situations. The copy got increasingly unhinged. I think the audience liked it. But you were in the courtroom for the majority of Musk v. Altman. You got to see a bunch of the testimony live as these guys took the stand, as Mira Murati and others took the stand.
We’ll start at the high level. I think the audience probably knows that Elon Musk lost, but what was this case about and what were the vibes in the courtroom?
There are two things that we should distinguish. There was what the case was ostensibly about, and then there was what the case was actually about, and those are two entirely separate things.
Ostensibly, the case was about the violation of a charitable trust.Elon Musk had donated a bunch of money to OpenAI Foundation, and then they created a for-profit, and he thinks that’s a violation of his charitable trust. He also thinks that the timing of that was right around what is known as “the blip,” when Sam Altman was briefly removed and brought back. Put a pin in that. It’s going to be important here. That’s what we’re ostensibly there for.
Verge subscribers, don’t forget you get exclusive access to ad-free Decoder wherever you get your podcasts. Head here. Not a subscriber? You can sign up here.
Because it was around the blip, Microsoft was accused of aiding and abetting, and Microsoft very quickly became my favorite part of the case.
In reality, there had been so many changing legal strategies around this. This case was filed I think two years ago in state court and then withdrawn and then put in federal court. There’s just been a myriad of things that have shuffled around since then, including a charge that got dropped right before we went to court.
So to me, the main point of this was punishing Sam Altman and maybe trying to kneecap OpenAI. And this is a case where the two worst people you know are fighting so it’s kind of hard to root for anyone. The most common response that I tended to get when I would talk about this to people or when I would post about it on social media was like, “Can they both go to jail?” So that’s kind of the vibe.
The courtroom was a little bit of a zoo during Musk’s testimony. We had one woman who got called down in front of the courtroom by the judge and chewed out because she had been taking photos in the courthouse. On the very last day, we had a guy who was ejected because he had been recording the proceedings in the courtroom. There were some shenanigans.
Every time we would leave the courthouse, there would be some kind of protest going on, usually behind the lawyers as they were trying to give their daily summary and spin what they had done in the courtroom, and then parading behind them would be a guy in a Cybertruck holding an “Elon Sucks” sign.
Perfect.
So that was what that was.
I want to come to the legal issues and particularly the ruling from the jury, as there’s a lot of mechanics there. I just want to stick on a point that the goal here was for Elon Musk to punish Sam Altman, and connect that to the protests and the comments you’re getting on social media, and certainly the comments we get every time we publish anything about AI. Is there any reputation left to damage for Sam Altman or the AI industry as a whole? Because it seems like both of these guys are at all-time lows. I’m thinking about jury selection when the judge had to just say, “It seems like no one likes Elon Musk, but we’re going to have to trust that the jury will be fair.” What’s even left to take away here?
There’s no floor about these things. I also view Sam Altman as untrustworthy, which is one of the things that this trial was really driving home as one of the points that Elon Musk’s lawyers were making, and I agree. I also think everybody else in the trial was totally untrustworthy. It was not just Sam Altman, it was all of them.
One of the things that I found myself thinking about was that the person who really got damaged the most was Mira Murati who, at least as far as I know, didn’t have a reputation as being somebody who was untrustworthy, or conniving, or whatever. And then in testimony from former OpenAI board members, we found out that she was one of the reasons that Sam Altman got fired and then was immediately texting Sam Altman like, “Oh, no, Sam, it’s very bad. It’s very bad, Sam.” You remember during this blip that Altman was fired for a pattern of being untrustworthy or something.
It was “he was not consistently candid with the board,” which could have meant anything.
Anything! And the thing that I remember, because I gossip with a bunch of journalists and we are ferocious gossips, is all of us were like, “Oh, he did something illegal. Let’s find out what illegal thing he did.”
As far as I can tell, no, he didn’t. It was just that he was engaging in what I would characterize as relatively normal executive shenanigans, where you are maintaining your control of the company by pitching your subordinates against each other — a strategy that is widely used in corporate America, by the way.
So she wouldn’t tell people that she was involved in his removal. She was the interim CEO, and then publicly supported him, and then publicly was involved in bringing him back.
Someone on the stand, I don’t remember who, said Mira was waiting to see which way the wind would blow and didn’t realize she was the wind.
That was Helen Toner, who was one of the board members who stepped down in this debacle. Because obviously as this proceeded, it became clear that by firing Sam in the way that they had fired him, they had jeopardized the entire company. One of the things that I thought was really interesting from Sam’s testimony — that I did believe, by the way — is that he thought about just taking a job at Microsoft and getting paid and not having to deal with any headaches anymore. I can certainly imagine after having been really publicly and embarrassingly fired, and having gone through all of the annoying things that one goes through as a manager and especially as a CEO, just being like, “You know what? I just want a paycheck.”
Who among us has not thought about retiring to a comfy job at Microsoft?
Right? And so when he was talking about that, I was like, “Yeah, actually, I believe that. That sounds real.” Then he obviously changed his mind.
But one of the things that I thought was really interesting about that is that we found out Helen Toner, who we saw in deposition testimony, was involved in potentially trying to sell OpenAI to Anthropic, a company that she has some ties to through the Effective Altruism movement. So again, no one here comes off looking good. I thought for a while that Helen Toner was maybe the most reliable witness we had heard from and then in the cross on the deposition it was like, “So tell us about your relationship with Anthropic.” And I was like, “Awww.”
That’s actually the thing that struck me about this entire trial. Helen Toner being wrapped up in Anthropic is one thing, but the entire AI industry at the top is 10 people who are wrapped up in each other emotionally, professionally. They’re writing each other obsequious emails, particularly to Elon, just full of flattery and praise about how great everyone is.
The idea that they’re going to make AGI is taken for granted in some way. These are the leaders of a new religion in a real way, you can see it, and they all lack any management instincts or emotional maturity to deal with the kinds of tasks that are put in front of them or the stakes or the money. You can just see it. It’s in the trial, it’s in the evidence, that they’re cracking under the pressure that they’re putting one another under, and there’s no outlet. In fact, the only outlet might have been Satya Nadella, who comes off as the coolest cucumber around because he’s just like, “I don’t know, is this going to make money? Don’t call me.” That’s basically his whole vibe.
Again, I loved Microsoft in this case. I’m not a Microsoft user. I am familiar with their products. Which by the way, their opening statement was so good. It was just a list of Microsoft products you might’ve used at some length.
“Remember us?”
It was fantastic. They were just like, “We’re not sure why we’re here, but you know us. We’re Microsoft. You’ve used Windows, surely. Do you like Xbox? That’s us.” So that was great.
There was really a sense that the only adult in the room at any given time was somebody from Microsoft. We saw that over and over again where Satya Nadella is like, “Don’t text me. Don’t leave a paper trail.” His emails are not especially spicy. I think the spiciest they got is something like him being like, “Well, we don’t want to be IBM and have them be Microsoft.”
This is OpenAI. He doesn’t want to be the commodity provider of data center hardware and have their software be the important thing, which is what happened to IBM and Microsoft.
That’s right. Which, by the way, totally understandable sentiment, I feel.
Especially from Microsoft. He’s like, “I know what’s happening here.”
That was the spiciest thing we got out of Microsoft. That was it.
So these are people who, in addition to having the management chops and having the sense of what you do and don’t do, were also just a little bit less dramatic. Over and over again, we’d have a witness, and there would be some really brutal and devastating cross from OpenAI. And then Microsoft would get up and be like, “Was Microsoft there? Was Satya Nadella there? Does anyone from Microsoft know anything about any of this? No further questions, your honor.”
It was a beautiful punchline every single time.
That’s very funny. So Microsoft obviously put a bunch of money into OpenAI. Nadella had that famous quote about being above them, below them, around them, referring to Azure and its dependency on Azure and how they would deploy OpenAI’s models. But eventually the trial comes down to, “Did they illegally convert this charity to a for-profit, and along the way, take something from Elon Musk?” What was the actual jury verdict on those counts?
The jury verdict was that Elon Musk filed the suit too late, and the statute of limitations had run out. And I’m going to be real with you, I think that had there not been a statute of limitations question, he still would’ve lost. This was a pretty weak case.
We’re going to start with the statute of limitation stuff because that is the most relevant. And then I will walk you through all the rest of it because we did do all of this in exhausting detail for the last month of my life.
One of the things that was part of Musk’s case was that he claimed that he didn’t think his trust had been violated until the blip. For this reason, he was still within the statute of limitations. The law, I believe, is that you need to file within three years. We saw a bunch of evidence that he had been read in repeatedly on the conversion to a for-profit and the various investment rounds.
I found myself unexpectedly sympathetic to Sam Altman during this trial. So congrats, Sam. He kept trying to get Elon to like him again. There would be these emails where it was like, “Hey, we’re raising this round.” Or he’d be emailing people to see what kind of mood Musk was in, if it was a good time to talk to him, because he just wanted to make sure that Elon knew what he was doing, and was it a good time for them to chat? Was Elon in a good mood? If you have a person whose job it is to tell people whether you’re in a good mood or not, I strongly feel that suggests that you maybe are difficult.
“How deep is today’s K-hole? Let’s find out before we ask for money.”
Over and over again, there was evidence of Musk being read in every single step of the way. Knowing about the Microsoft investments, knowing about the fact that they were creating this for-profit. In fact, there was a bunch of email evidence that he thought that making OpenAI a nonprofit had been a mistake, that it should have been for-profit from the jump.
There’s a ton of evidence that, separately from the timeline question, suggests that OpenAI would’ve won this case. The definition of a charitable trust, and I’m going to mangle this slightly because I am not a lawyer, is that you have to have a specific purpose for your donations. You have to have established that this is a trust, and then the next thing you have to establish is that that trust was violated.
Just looking at all of the donations, which we did in some depth, there were no strings attached that any of us saw. No one at all remembered there being any strings attached. One of the more devastating lines of testimony was that Shivon Zilis was asked, “Were there strings attached to these donations?” And she was like, “Well, not that I recall.” And then in the closing statement, OpenAI’s lawyer’s like, “Man, not even the mother of his children can corroborate his account.” Okay.
That’s brutal.
So there were no strings attached. And then we had a financial analysis that showed that money was gone very, very quickly. , tThey had spent it, because AI is expensive. And they had spent it in the way that it was meant to be spent, and all the other money that happened afterwards had nothing to do with Elon Musk. So there was that.
One of the things that I’m just going to put an asterisk on here, that I thought was interesting but didn’t write about, was that Musk had been paying the rent for OpenAI. They actually had to go back and ask him for money because Neuralink was in the building. When they got accountants to try to get their books in order so that they could proceed, the accountants were like, “Oh yeah, you can’t be supporting somebody else’s for-profit business in this building. You need to get rent money from Neuralink. They need to pay you back.”
Wow.
Not that we went into this in any depth, but my suspicion is that Musk had been taking a write-off on all of those donations on this building, and had been also taking that write-off on the space that Neuralink was using, which was why that money then had to be paid back to OpenAI.
There’s a lot here. I mean, there’s a lot of just Elon Musk, there’s infinitely complicated fractally expanding OpenAI layers of companies within the nonprofit that have board control, and people can fire Sam Altman. All of that seems enormously complex, and maybe worth some future litigation. But the jury just went with statute of limitations. And it seems like that’s maybe all they should have been talking about, if that’s what was going to end the case this quickly. Why do you think that we spent all the time in the substance and the complication when Elon had just filed too late?
I did get people asking me about this as well. “Isn’t statute of limitations a legal issue? Why didn’t the judge rule on this?”
And the answer is there was a question of fact, which was, “When should Elon have known what was going on?” And he’s saying, “I didn’t know until the blip. And so I’m within the statute of limitations.” And everybody else was saying, “e’s known the entire time. It’s over.” That was the thing that was being litigated. It wasn’t the only thing that was being litigated, but that was the one that ended up mattering: that the jury was like, “Yeah, he definitely knew all of this was happening. This is ridiculous.”
If the goal was to trash Sam Altman, of course you would pick the blip because then you get to pull every document and email and text message from the blip into the trial into evidence. You get to publish it. We published it. Was that the goal? Was Elon just saying, “I only knew about this when Sam Altman got fired,” in order to put all of that damaging evidence into the record?
I think that was the goal. I think that was what was actually going on. It was also meant to distract OpenAI, because they did have to pay this very expensive law firm to do some very expensive work to defend them. They didn’t just defend the statute of limitations. They defended all of the subclaims and all of the other things as well, which is why there is so much in our stories. They were bringing forward as much as they could to defend every single part of every possible claim because they had to.
And so, yeah, making Sam Altman look bad, distracting Sam Altman, maybe removing resources as Altman approached an IPO, those were probably the primary goals. I think Musk would’ve been happy with a win. He certainly would’ve been thrilled to force OpenAI to give up a bunch of money, even if it went back to the OpenAI Foundation, as he belatedly decided it should go. There are any number of things that I think he would’ve taken as icing on the cake, and he said that he’s going to continue this through the appeals process.
Let me just read you the quote. Elon appeared at a Forbes conference, and he said, “I think this is a dangerous precedent to set. If someone can take a nonprofit and convert it to a for-profit, that undermines all charitable giving in America.” I don’t think Elon understands how precedent works, but it seems regardless of that, he’s going to keep tying OpenAI up in litigation for as long as he can.
Oh yeah. He said something very similar to that on the stand, by the way. He has some pet phrases he likes, and “dangerous precedent to set” and “undermines all charitable giving in America” are on the list.
I think he does intend to tie OpenAI up in litigation for as long as he possibly can, bleeding them for cash, which is a strategy that we’ve seen other billionaires use. Most famously, Sheldon Adelson, who went after a Las Vegas paper, if I remember correctly. Not because they had done anything wrong — and they were in fact ruled not to have done anything wrong — but because defending the case was so financially expensive that they nearly went under. And that is a strategy you can use if you have unlimited resources: you can just bleed somebody out.
I do feel like if you’re Elon Musk and you’re really worried about rich people using their charities to enrich themselves, there are a handful of people in his direct orbit running the country that he might want to take a closer look at. This seems like he’s saying it because he just wants to keep screwing with OpenAI.
Oh, absolutely. There’s no doubt in my mind that this is personal for him. The thing that I have been thinking about for a while and am unable to quite tell is, “Is he personally pissed off at Sam Altman, or is he just affronted that OpenAI succeeded without him?”
Well, so this is my other question. Maybe you kill OpenAI and it goes away and you’ve bought yourself some time. Elon has publicly said that they built Grok incorrectly and they need to start over. They are selling a huge amount of data center capacity at Colossus 1 to Anthropic, who Elon has hated in the past, but he says, “It’s all fine now” because they showed up with a check to buy his data center capacity.
Even if you kill OpenAI, it doesn’t make xAI the winner. They’re basically starting over, as they publicly said. They’re giving up their compute capacity. What is the point of this, except to just vindictively kill OpenAI? It doesn’t seem like I can identify the competitive advantage here.
I mean, killing a competitor is not necessarily not a competitive advantage.
Let’s say OpenAI is in first or second or third or something, or just running in a different direction on the track at this point. Who knows what they’re doing. If you’re in last, it doesn’t matter. In some way, he’s helped Anthropic and Google here.
Let’s say Musk wins and OpenAI has to disgorge all this money and that potentially just blows a hole in the side of the company. I can’t rule out that Altman is enough of a deals guy that he could patch it up, but let’s say he can’t.
OpenAI is at the center of a web of deals, huge deals with places like CoreWeave and Oracle and Microsoft. Every company in the AI space is one degree of Kevin Bacon away from OpenAI. If you knock that company out, not only do you have a bunch of talent that comes free and needs a job now, which you can maybe hire, you also have created conditions where you can negotiate really favorable terms in these now suddenly open data centers with companies that now suddenly have huge holes in their revenue.
I wish I could ascribe that level of 3D chess, but there’s a part of me that says this is just personal and vindictive. And we’re going to see appeals and further campaigns about how Sam Altman stole a charity, and that will be distracting for OpenAI on one level. And on another level, they’re just going to continue selling Codex to people, because it is good at writing code, and a lot of software companies seem very taken by that. Do you think this has any meaningful effect on OpenAI in the future?
No. We knew going into this trial that Sam Altman did not have a reputation for being perfectly honest. I mean, that was the upshot of the blip. There was a 17,000-word article in The New Yorker about this. This is something that I effectively think is priced in, in the same way that Elon Musk’s, let’s say, scattershot relationship with the truth is also priced in in all of his companies. People know who these guys are, none of this is a surprise, which is why I think, again, that the person who got hurt the most here is Mira Murati, who did not have her reputation trashed before this.
So there’s going to be an appeal. These companies are going to carry on spending money. What do you think happens next? What should people be looking for? Or is this one safe to set aside for now?
I would set it aside for now. We had all the fun of going through their emails, we had their ridiculous text messages. But the biggest takeaway from the trial that matters is discovering that Grok sucks, even though Elon Musk had distilled everybody’s models. To me, that’s shocking.
Not that I am an expert in AI. It’s entirely possible that you can distill all these models and have your AI still suck. But I think that that really is a take-home point, that one of the consistent things that we were seeing in this trial was that the nerdiest of the nerds, [OpenAI co-founders Greg] Brockman and Ilya Sutskever were both like, “He’s not really serious about AI.” And I came away being like, “Yeah, he’s not serious about AI. He doesn’t know what he’s doing.”
We have all of the things that you talked about: They’re starting over from scratch, they’re leasing out their data center capacity, they’re doing all of these things that suggest that whatever Musk did with whatever billions of dollars, because I think xAI was spending… The reporting was a billion dollars a month. They’re starting over from scratch, there’s nothing, and this is even with cheating by distilling everybody’s models.
Right. This is him saying, “We didn’t build it the right way.” They didn’t actually do a proper training run, they distilled all the other models. And so they’re not on the frontier. Which, by the way, has happened to other companies. Meta is out there saying that they were not on the frontier and they started over in a meaningful way. This is a nascent industry. It’s not clear how to do these things or build these things or ship these things in a way that works.
I think my big question coming out of all of this is, boy, this handful of people that have been entrusted with spending all this money and asking for all these resources and in many ways pitching a vision in the future, they seem so immature. And even if that’s priced in, did this trial just reveal that fundamentally they’re immature and maybe you should let the Microsofts and the Googles of the world be in charge of deploying this technology, because at least the amount of bureaucracy in place at those companies will slow them down.
That could be one takeaway. Given the way that Google has destroyed its own search engine for its AI models, I’m not clear that we want to include Google in this conversation.
I’m just saying.
We’re maybe talking about Microsoft and maybe Apple. But yeah, you want grownups in charge of this technology, for sure. And the immaturity I thought was really interesting because there was a recurring theme, again that didn’t seem worth writing about separately, but that I will mention here. Over and over again, you’d get somebody on the stand and they’d be like, “Ever since I was a child, I’ve dreamed of AI. I’ve thought about the smart computer and how amazing it would be. And it kept me up at nights when I was nine years old.”
First of all, that’s stupid because that’s fiction. If you can’t tell the difference between fiction and reality, we have bigger problems. I had some childhood dreams too, and I want to be real with you, I just don’t think that owning a horse is going to be a thing that makes sense for me.
By the way, I just want to point this out. As we’re speaking, there is breaking news. Andrej Karpathy has joined Anthropic.
[Laughs] Sorry. [Laughs] Oh my God.
Which is a perfect capstone on this trial. He’s like a main character. He gets recruited to and from all these companies and now he’s at Anthropic, which seems like far and away the winner of this whole thing. Hands the cleanest, products the most successful. Why did you start laughing that hard?
A recurring theme in the trial was Musk poaching OpenAI engineers. And of course, Andrej Karpathy was one of them, because he went from OpenAI to Tesla. Because OpenAI, when it was a foundation, was asked by Elon in a way that’s suggested was not actually an ask, if you follow me, to come work on autopilot because they were having a hard time with autopilot at Tesla. And so several engineers, including Greg Brockman, went over and worked on autopilot while they were theoretically working for OpenAI. So if anybody was stealing resources from a charity, I kind of think it was Elon Musk.
One of the people who permanently stayed was Karpathy and he shows up again and again. This recruiting push that Musk made out of OpenAI while it was still a nonprofit, while he was still theoretically involved with it, while he was still theoretically on the board and had a fiduciary duty to the nonprofit, he was using it as a recruiting ground for Tesla.
That’s very good. Well, Liz, I have a feeling we’re going to keep you very busy with these characters in the year to come. My prediction is that OpenAI does not end the year looking the same as it does now, that there will be yet more change at that company.
I think that’s right.
The other little cherry that I’d like to put on top of all of this, speaking of Anthropic, is that one of my personal favorite parts of this trial occurred while the jury was out of the room. It was an evidence dispute about whether or not the jury could be shown a jackass trophy. Imagine a participation trophy that is just the back half of a donkey. And it said something like, “Never stop being a jackass for AI safety.”
It was presented to an AI safety guy who, when Musk was on the way out at OpenAI and was doing a Q&A session, was like, “Hey, it sounds like you’re really interested in speed over safety. I don’t think that’s a good idea,” and Musk called him a jackass. And so would you like to take a guess at one of the people involved in presenting that trophy?
Was it Karpathy?
It was [Anthropic CEO] Dario Amodei.
Oh, amazing. Amazing. Perfect. That tracks with everything Anthropic has stood for. Everyone’s leaving to start a safer AI company, and Dario was among the first. Perfect. Did he take the trophy with him?
He did. The lawyers had it, so I assume he’s gotten it back. We published a photo because as I was live-tweeting this, I saw people asking for a photo, so I got ahold of one, but I remain very entertained by this trophy. So hats off to the fine engineers who eventually did leave and make Anthropic, because it seems like they have a pretty good sense of humor.
Yeah, they figured it out. All right, Liz, we’ll have you back soon, hopefully under more rational circumstances, but it’s always a pleasure. Thanks for being on Decoder.
My pleasure.
Questions or comments? Hit us up at decoder@theverge.com. We really do read every email!