US voters abroad, have you got your ballot yet? All states are now sending ballots to overseas voters. So if you want to vote this year, you should request your ballot right away. Go to votefromabroad.org and click request my ballot as to receive a by email.
Questions? Ask our support team directly. Don't miss this Gens to shape the future. Don't delay.
Votefromabroad.org. But I think you'll know this case was really about Elon Musk being mad at same woman, or at opening eye, or being successful without him, and wanting Sam and the company punished in some way. So in a room full of very emotional, untrustworthy, unreliable narrators all fighting with each with each other, did anyone even have a reputation enough to lose? Is there a floor here for the AI industry?
Liz has a lot of thoughts here. I always enjoy having her on the show. Here's Lizopado, a must-be-alman. Here we go.
Lizopado, if you are a senior chaos reporter here at TheVerge.com, you just covered the same-alman-ill on Musk Trial. Welcome to the coder. 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. Every, a few times a year we drive you absolutely bad 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 the must-be-alman. You got to see a bunch of the testimony live as these guys took the stand, as Mira and others took the stand.
We'll start the highlight. I think the audience probably knows 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. Those are two entirely separate things. So, ostensibly, the case was about the violation of a charitable trust, which was that Elon Musk had donated a bunch of money to OpenAI Foundation. Then they created a for-profit.
He thinks that's a violation of his charitable trust. Also, the timing of that was right around what is known as the blip when Sam Alton was briefly removed and brought back. Put a pin on that. It's going to be a worker.
That's what we're ostensibly there for. Because it was around the blip, Microsoft was accused of aiding and abetting. Microsoft very quickly became my favorite part of the case. That's what we were theoretically doing.
But in reality, there had been so many changing legal strategies around this. This case was filed, I think, two years ago and stay 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. To me, I think that the main point of this was punishing Sam Alton and maybe trying to kneecap OpenAI.
This is a case where it's the two worst people you know are fighting. It's 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, can they both go to jail? 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 shoot 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. Then parading behind them would be a guy in a cyber truck holding an Elon sucks side. That was what that was.
I want to come to the legal issues and particularly the ruling from the jury. There's a lot of mechanics there. I just want to say on the point that the goal here was for Elon Musk to punish Sam Altman and connect that to the protests and the comments are getting on social media. And certainly the comments we had every time we published on the AI.
Is there any reputation left to damage for Sam Altman or the AI industry as a whole? Because it seems like both 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?
I kind of think there's no floor about these things. One of the things that I will say is that from my perspective, I also view Sam Altman as untrustworthy, which is one of the things that this trial was really driving home. It's one of the points that Elon Musk's lawyers were making. 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 I think the person who really got damaged the most was me or Marathi, 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 Open AI board members, we found out that she was one of the reasons that Sam Altman got fired. And there was immediately texting Sam Altman. Oh no, Sam, it's very bad. It's very bad, Sam.
You remember during the split that Altman was fired for what was 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 gossiped with a bunch of journalists and we are ferocious gasses, all of us were like, oh, he did something illegal.
Let's find out what he legal think he did. As far as I can tell, 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.
And she was like the CEO, the interim CEO, and then publicly supported him and then publicly it was involved in bringing him back. I think 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 and the way that they 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.
And I can certainly imagine after having been really publicly and embarrassingly fired and having gone through all of the annoying things that went through as a manager and especially as a CEO, just being like, you know what, I just want to 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, that actually, that, I believe that, that would, that sounds real.
And then he obviously changed his mind. But one of the things I thought was really interesting about that is that also Helen Toner, who we saw in deposition testimony, one of the things that we found out was that she was involved in potentially trying to sell OpenAI to anthropic company that she has some ties to through the EA movement. And so again, no one here comes off looking good. Like, I thought for a while that Helen Toner was maybe the most reliable witness we had heard from in the cross on the deposition.
It was like, so tell us about your relationship with the graphic. And I was like, aww. That's actually the thing that struck me about this entire trial. Helen Toner being wrapped up in the project is one thing, but the entire AI industry at the top is like 10 people who are wrapped up in each other emotionally, professionally, they're writing each other up sequence.
You know, particularly the 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. Like these are the leaders of a new religion and like a real way, you can see it. And then they all lack any management instincts or emotional maturity to deal with the kinds of tasks they're putting in front of them or the stakes or the money.
And you can just see it. Like 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 Sashanadella, who comes off as the coolest cucumber around because he's just like, I don't know, just gonna make money. Don't call me. Like that's basically his whole vibe. Again, I loved Microsoft in this case.
Like, I'm not a Microsoft user really, like, you know, I'm familiar with their products, which by the way, their opening statement was so good. It was just a list of things of Microsoft products you might have used at some length. It was fantastic. They were just like, I wear not sure why we're here, but you know, we're Microsoft.
You've used Windows shortly. Do you like Xbox? That's us. So that was great.
There was really a sense in which the only adult in the room at any given time was somebody from Microsoft. We saw that over and over again, where Satya Nadella was like, don't text me. Don't leave a paper trail. His emails are not especially spicy.
Like 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. He doesn't want to be the commodity provider of data center hardware and their software be the important thing, which is what happened to Microsoft. That's right. Which by the way, totally understandable sentiment, I feel.
Yeah, it's not just from Microsoft. He's like, I know, I know what's happening here. That was like the spiciest thing we got out of Microsoft. Like that was it.
So these are people who, in addition to sort of having the management chops and like having the sense of like what you do and don't do, we're also just like a little bit less dramatic. And like over and over again, we'd have a witness and there would be some like 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. Great. It was like a beautiful punch line every single time. That's very funny.
We're gonna pause here for a quick break. We'll be right back. So Microsoft obviously put a bunch of money into OpenAI. They all have that famous quote about being above them, below them, around them, referring to Azure and dependency in Azure and how they deploy OpenAI's models.
But eventually the trial comes down to did they legally convert this charity to a for-profit and along the way, take something from Elon Musk. What was the actual jury verdict on those kinds? The jury verdict was that Elon Musk filed the suit to late and the statute of limitations had run out. I'm gonna be real with you.
I think that had there not been a statute of limitations question, he still would have lost. This was a pretty weak case, but we're gonna start with the statute of limitation stuff because that is like 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 and I was left like month of my life. So one of the things that was part of Musk's case was that he claimed that he didn't think his press had been violated until the blip.
And so for this reason, he was still within the statute of limitations. And 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, 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, like him again. There would be these emails where it was like, hey, we're raising this round. Or he'd be like 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? Which by the way, 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. Yeah, how deep is today's cable?
Let's find out before we push. We were asked for money. There was evidence of Musk being read in every single step of the way, knowing about the Microsoft investments, knowing about like the fact that they were creating this for-profit. In fact, there was a bunch of email evidence that he thought that making open AI a nonprofit had been a mistake that it should have been for-profit from the jump.
So there's just like, there's a ton of evidence that separately from the timeline questions suggests that open AI would have won this case, not least because like the charitable trust thing, the definition of a charitable trust, I'm going to mangle this slightly because I'm not a lawyer. But you have to have a specific purpose on your donations. And so 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 Shivan Zillis was asked, were there strings attached to these donations? You're like, well, not that I recall.
And then in the closing statement, opening eyes lawyers 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 that money was gone very, very quickly, that they had been spending, because AI is expensive, they had like spent it. And they had spent it in the way that it was meant to be spent. 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 I think great about was that Musk had been paying the rent for open AI. They actually had to go back and ask him for money because Neuralink was in the building. And 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 be back. And so my suspicion is not that we went into this in any depth, but 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 is why then that money had to be paid back to open AI. There's a lot here. I mean, there's a lot of just Elon Musk.
There's infinitely complicated, like fractally expanding opening eye layers of companies within the nonprofit that have board control and people can fire some of them. And 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 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? Several things. Because I did get people asking me about this as well. Like, is it statute of limitations a legal issue?
Why didn't the judge rule on this? Yeah. 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 he'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 is that the jury was like, yeah, he definitely knew all of this was happening. If the goal was to trash the moment, of course you pick the blip, right? Because then you get to pull every document and email and text message from the blip into the trial into evidence.
You have to publish it. We published it. Was that the goal? Was that the goal?
Was it was Elon just saying I only knew about this when the when Sam Olin got fired in order to put all of that damaging evidence into the record? I think that was what was actually going on. I think it was also meant to distract opening eye 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 sub claims and all of the other sorts of things as well, which is why there is so much in our stories is that they were bringing forward as much as they could to defend every single part of every possible claim because they had to. I think that making Sam Olin look bad, distracting Sam Olin, maybe removing resources as ultimate approach to an IPO. I think those were probably the primary goals. Musk would have been happy with the win.
I think he certainly would have 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 of the number of things that I think he would have taken as icing on the cake. He said that he's going to continue this through the appeals process.
I think this is a dangerous precedent to set. If someone can take a non-profit 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, I mean, he said something very similar to that on stand, by the way.
He has some pet phrases he likes and dangerous precedent to set undermines all charitable giving in America. 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 on. That is a strategy you can use if you have unlimited resources, you can just bleed somebody out. But you feel like if you're Elon Musk and you're really worried about rich people using their charities to enrich themselves, they're handful of people in his direct orbit running the country.
You might want to take a closer look at it. 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 am unable to quite tell, I've been thinking about this for a while, is he personally pissed off at Sam Altman or is he just a frontage that OpenAI succeeded without him? We've got to take a picture of him. Hi, everyone. I'm Kara Swisher.
And I'm Scott Galloway. Twice a week, we get together on Pivot to sort through the biggest stories in tech business and politics and we're taking the show back on the road for a live tour in November. So we'll bring hot tags, I'll bring the inside and the dick jokes that Kara will try, he would have tried, keep me in check. You'll get to see her hit me in person live.
I'll do it. If someone has to do it, we'll cover AI, whatever chaos Washington has cooked up that week and every billionaire misbehaving. That's all of them. And you'll also get to participate in all his Q&A, so bring your best questions, Scott.
We'll probably answer ones you didn't ask. We also have secret special guests which you'll enjoy for every city. Get tickets at pivot tour.com, that's pivot tour.com. We are selling out, so don't wait and we look forward to seeing you.
This is my other question. Maybe you kill up an AI and it goes away and you've bought yourself some time. Elon has publicly said that they built Grog incorrectly and they need to start over. They are selling a huge amount of data center capacity at Colossus 1, two anthropic who Elon has hated in the past, but he says it's all fine now because I showed over the check to buy his data center capacity.
Even if you kill up an AI, it doesn't make X AI 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 up an AI?
It doesn't seem like I can identify the competitive advantage here. Killing a competitor is not necessarily not a competitive advantage, especially if you kill the open eyes in first or second or third or something, or running 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 in Google here. Think about this. Let's say Musk wins and OpenAI has to discourage 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 can patch it up, but let's say he can't.
OpenAI is at the center of a web of deals. They are huge deals with places like Goreweave and Oracle and Microsoft. All of these companies, every company in the AI space is one degree of Kevin Bacon away from opening AI. 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 the revenue.
I wish I could ascribe that level of 3D chess. There's a part of me that says this is just personal and vindictive, and we're going to see appeals in further campaigns about how Sam Oppmann stole a charity, and that will be distracting for OpenAI on one level. On another level, they're just going to continue selling codecs 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. I mean, here's the thing. We knew going into this trial, the Sam Oppmann did not have a reputation for being perfectly honest, right? 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 and all of his companies. People know who these guys are.
None of this is a surprise, which is why I think that the person who got hurt the most here is Mira Maradi, who did not have a reputation trash before this. There's going to be an appeal. These companies are going to carry on sending 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 aside for now. We had all the fun of going through their emails.
We had their ridiculous text messages. The biggest thing I think that was the takeaway from the trial that matters is discovering that rock sucks, even though Elon Musk had distilled everybody's model, which 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.
I think that that really is a take home point. One of the consistent things that we were seeing in this trial is that the nerdiest of the nerds, Brockman and Ilya Scottz, were both like, he's not really serious about AI. 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 with, 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 recording was a billion dollars a month. They're starting over from scratch.
There's nothing. This is even with cheating by the building. Everybody's models. Right.
This is something we didn't build the right way. They didn't actually do a proper training run. They distilled all the other models. They're not on the frontier, which by the way, has happened to other companies.
I met as out there saying that they were not on the frontier. 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 in a way that work.
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, in many ways, pitching a vision of 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 in the Googles of the world be in charge of deploying this technology? Because at least the amount of bureaucracy in place that this company 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 grown-ups in charge of this technology for sure.
The immaturity, I thought, was really interesting because there was a recurring theme, which is that 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. And I just want to say, first of all, that's stupid because that's fiction.
And if you can't tell the difference between fiction and reality, we have bigger problems. And it's like, okay, but 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.
At least they sent the Victorian industrialist to finishing school. That's basically what you're getting at. By the way, I just want to point this out. As we're speaking, there's breaking news.
Andres Carpathy has joined Anthropic, 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 on the trial was Musk poaching open AI engineers. And of course, André Caprathey was one of them.
He went from open AI to Tesla because open AI, when it was a foundation, was asked by Elon in a way that suggested it was not actually an ask if you follow me to come work on autopilot because they were having a hard time with autopilot to Tesla. And so several engineers, including Greg Broppen, went over and worked on autopilot while they were theoretically working for open AI. So if anybody was stealing resources for a maturity, I kind of think it was Elon Musk. One of the people who permanently state was Carpathy.
He shows up again and again, like this recruiting push that Musk made out of opening AI while it was still a nonprofit while he was still theoretically involved with it while he was still theoretically on the board and had like a fiduciary duty to the nonprofit. He was using it as a recruiting ground for Tesla. Well, I have a feeling we're going to keep you very busy with these characters in the year's cup. You know, my prediction is that open AI is not in the year looking the same as it does now.
There will be yet more change with that company. I think that's right. The other cherry that I've 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. And it was an evidence dispute about whether or not the jury could be shown a jackass trophy.
And it was like imagine a participation trophy that is just the back half of a donkey. And it said something like, you know, never stop being a jackass for safety, for AI safety. It was presented to an AI safety guy who when Musk was on the way out opening an AI and was doing a Q&A session was like, Hey, it sounds like you're really interested in speed over safety. Like, I don't think that's a good idea and Musk called him a jackass.
So would you like to take a guess at who was one of the people involved in presenting that trophy? Was it Carpathy? It was Dario Amadeh. Amazing.
Perfect. And that tracks with everything in anthropic history. Right. This is everyone's leaving to start a safer AI company.
And Dario was like, I'm on the first perfect. Do you take the trophy with him? The lawyers had it. So I assume he's gotten it back.
We published a photo, but it remained very entertained by the stroke things. So like 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 fun to be on coder. My pleasure. I like to thank Liz for taking the time to join coder and thank you for listening. I've enjoyed it.
It's like, let us know what you thought about this episode or anything else at all. Drop us a line. You can email us at coderat diverge.com. We really do read emails where you can maybe have directly on threads or blue sky.
We're also on YouTube. You can watch full episodes at to coder pod. You also have a TikTok and Instagram. They're also at to coder pod.
No, a lot of fun. You guys are going to share with your friends and subscribe over your podcasts. Decoder's production version, pro unboxing, podcast network shows produced by Kate Cox and Xtad and edited by Ursa Wright or editorial directors Kevin Shane. The Decoder music is by breakmaster so we'll see you next time.