Emergency Pod: Sam Altman is Out at Open AI
Hard Fork- 762 views
- 18 Nov 2023
Sam Altman, the chief executive of Open AI, was pushed out of the company by its board of directors on Friday. The news was a complete shock to much of the company’s employee base and to its largest corporate partner, Microsoft. Silicon Valley insiders are scrambling to get answers on exactly what happened and why the board’s decision seemed so abrupt. We rundown what we know and the many things we still don’t.
Kavin, it's very late here. Why are you calling me?
Well, I just wanted to see what you had planned for the weekend. No, we are doing an emergency podcast because, Casey, some very big news just happened.
That's right. My red hard fork phone that I keep next to my bed is ringing and it says there's big news.
Today, just a few hours ago in a surprise move, OpenAI CEO, Sam Altman, was pushed out of the company. He was effectively fired by the company's board. This move was a total shock to everyone in the tech industry and was especially shocking to us because on Wednesday of this week, you and I both sat down for a very long, very wide-ranging interview with none other than Sam Alman.
We did. And looking back, I wish I'd just ask him, Hey, do you think you're about to get fired?
Yeah, that was not one of the questions we asked him shockingly. And to be frank, it didn't seem like he had any idea this was coming. We were talking to him on what was essentially the one year anniversary of ChatGPT. We asked him a lot of questions about OpenAI and ChatGPT and the future he saw for artificial intelligence. And at no point during our interview did he indicate that anything was a myth or that anything like this was about to happen. So I have to wonder whether he knew or whether this was a total shock to him, too.
Yeah. Well, like you said, he definitely did not betray any notion that this was about to happen when we talked to him.
So today in this emergency podcast, we are going to talk about this news, what happened and what it means. We are not going to share our interview that we recorded with Sam Alton just yet. Candidly, we're still trying to figure out what to do about it. But today, I just thought we should hop on the mics and talk through what I think is probably the most shocking text story of the year and maybe in several years.
You know what I would say, Kevin? You could train an entire large language model just based on the text messages that have been sent about Sam being fired today. It's true.
After this news came out, my phone melted. And I knew it was one of those news days where my phone is going to run out of battery by noon. And it did. I have charged my phone twice today because it is just one of those days. So let's just get the timeline down because this all happened very quickly and without much warning. So on Friday, OpenAI announced this firing in a blog post. They said that Sam Altman has been pushed out effective immediately and that Miramaradi, the company's CTO, chief technology officer, will be taking over as CEO on an interim basis. They did not give the exact reasons for the firing, but they said that his departure, quote, follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities. The board no longer has confidence in his ability to continue leading OpenAI. They also said that Greg Brockman, who was the President and co-founder of OpenAI, as well as the chairman of its board, was being pushed off the board, and they left it at that. Sam Altman then shared, what I would describe as a vague post on X after the announcement.
He said, I loved my time at OpenAI. It was transformative for me personally, and hopefully the world a little bit. Most of all, I loved working with such talented people. We'll have more to say about what's next later. So very vague, no details about what happened or how he felt about it. And then just a couple of hours after that, Greg Brockman announced to the employees of OpenAI that he was quitting in solidarity or what many people interpreted as solidarity with Sam. And then on Friday night, Greg Brockman released a statement saying that he and Sam were both shocked and saddened by what the board did today. He gave a little rundown of his side of the story, basically said that both he and Sam were blindsided by these texts from Ilias Petskever, the chief scientist of OpenAI, who told them both minutes before this news went out that they were being removed from their roles. And he claimed that he was still trying to figure out exactly what happened, but that he had been taken by total surprise on this. So within the span of several hours, OpenAI's board has pushed out Sam Altman, the CEO, Greg Brockman, the COO, and just left everyone at the company reeling and wondering what happened.
That's right. Based on our reporting, Kevin, we know that none of the rank and file at OpenAI appeared to know this before it was announced to the public. So this came as a shock to them. Ena Frid at Axios reported that Microsoft apparently got one minutes notice before this news was shared.
Microsoft, of course, is one of their biggest partners.
That's right. And I can tell you based on my own reporting that last week, Sam and Sacha Nadella, the CEO of Microsoft, were spending a lot of time together, and Sacha did not get any indication again whatsoever that anything was up and that this might happen. Microsoft's stock dropped after this news was shared, as you might imagine, as people now worry about what this means for the future of their AI efforts, Nadella put out a statement later in the day essentially affirming his commitment to OpenAI, saying they have a long term deal in place and expressing his confidence in Mira to lead the company.
Yeah. So, Casey, what are you hearing from your reporting? I assume you've been calling around all day trying to figure out what happened here.
Well, I have to tell you, Kevin, I wish I done a better job reporting because despite sending what I would estimate to be 40,000 text messages, I still don't have the answers that I want. What I don't want to do is to speculate responsibly. But what I can tell you, based on having seen other CEOs get fired in other circumstances, is that Greg Brockman quitting, I think, tells us a lot about what is going.
On here. What do you mean?
When CEOs resign due to personal scandals, they do not typically have a bunch of people lining up with them to follow them out the door. And so when you have a co-founder of this company saying, based on the fact that Sam was fired I am quitting this company, that is a signal to me that the dispute that is at the center of whatever this is, is related to OpenAI. Now, we don't know if it's related to a business decision, a product decision, an AI safety decision. I can tell you all of those rumors are swirling inside the company and out, but we don't know. I can also say that the number of people who know the real reason behind this right now is extremely limited. I have been talking to a number of people very close to this situation who I would think would know, and they're coming back and they're saying, We are hearing Cricket. So this information was very, very tightly held. I do think it will come out eventually and maybe soon, but as of recording time, we just don't have it.
Yeah. I expect that every tech reporter in America is chasing this story right now. I've also been making calls and texting people all day trying to figure out what the heck happened here. And I think you're right. This came as a total surprise to almost everyone at OpenAI. They learned about it the same time the rest of the world did. I heard that there was an all hands meeting in which employees were shocked and dismayed many of them. Sam Altman was a very popular CEO among the rank and file at OpenAI. So this was not a situation where people had an easy explanation ready to go, Oh, this was clearly this thing that they could point to. They seemed totally baffled and totally shocked by this news. And I heard it described as a hostile takeover or a coup. Those are the kinds of words that people inside and close to OpenAI have been using to describe what happened today.
I think it is important to underscore what you said, Kevin, which is that Sam Altman was a very popular leader at OpenAI. He's a very popular leader in Silicon Valley in general. Of course, before OpenAI, he ran Y Combinator, the incubator that is responsible for companies including Airbnb, Stripe, Dropbox, a lot of other famous companies. So Sam Altman is one of the best connected people in Silicon Valley. And after the news broke, I just saw boldface name after boldface name, leaping onto X to express their support for him. Brian Chesky from Airbnb, Eric Schmidt, the former CEO of Google, and many, many more. All of them were essentially saying, we stand with Sam. So it shows you the amount of support that he has. And I have to say, I think it raises real questions about the future of OpenAI, because part of being a great leader at OpenAI for Sam meant being a great recruiter, which is one of the real things that he did. There's a limited amount of top tier talent in AI world, right? And Sam, because he is so well known from running Y Comedator, was able to go out and get a lot of the best people to come in and work at that company.
So there were a lot of people who went to work today because they wanted to work for Sam Alton. And on Monday, they're going to go into work working for someone else. And so I think the question is, are they so tied to that mission that they are going to stick with the company when, inevitably, it comes out whatever this dispute was?
Yeah. It's totally baffling to me. Something just isn't adding up. This was a very unusual circumstance. You do not see boards firing CEOs, very popular CEOs of very popular companies in this way, particularly. I mean, usually if a CEO is going to be pushed out, the board will do the nicety of saying this person is stepping aside to spend more time with their family is the cliché or to work on other projects. The board of OpenAI really did throw Sam under the bus in this statement, which makes me think that whatever happened, it must have been bigger than just a simple disagreement.
This, I actually think, is a good point to talk about the mess of OpenAI in general. And we have talked about it on the show before. There's basically no company that we talk about that has a more complicated structure than this company. There is a nonprofit board that operates a for-profit company. On that nonprofit board, which voted to fire Sam today, most of the members do not own shares of the company. Sam himself did not directly own stock in the company. And so you just have this weird situation where you have a for-profit company owned by a nonprofit. That is just weird. And if you were to say, does that scenario make it likelier that we would see something like we saw today? Then not like, Yes, it actually does. Okay. So I think that's part of what's going on here. The second thing I would say is that OpenAI has just been a messy company always. It was famously co-founded by Elon Musk. Elon Musk had a massive falling out with those folks and wound up walking away. A bunch of other co-founders at the company had a big falling out with Sam and the other folks there, and they went to start Anthropic, whose CEO, Dario Amaday, we had on the show, he's the CEO of that company.
He also was an OpenAI before then. So in the AI world, there are a lot of disputes, and they end with people slamming doors and often going to start their own AI companies.
Yeah, we should talk just for a second about the board for one more beat, because there's something that you said that I want to draw out, which is that the board members of this company do not, for the most part, own shares in the company. That is unusual for company boards. Usually, you'd have a venture capitalist on there, or maybe some early executives or just business people that the CEO knows and is friends with. This board, it's a small board, first of all. Before today, it consisted of Greg Brockman, Sam Altman, obviously, OpenAI's Chief Scientist, Ilia Setskiver. And then there are these independent non-employee directors, including Adam Diangelo, who's the CEO of Quora, Tasha McCauley, who is a technology entrepreneur, and Helen Toner, who is a scholar at the Georgetown Center for Security and Emerging Technology. And of those board members, I think it's spending a beat talking about the last two that I mentioned, because this is one hint that a lot of people I've been talking to today have been pointing me toward to try to read the tea leaves about what happened during this transition period, which is that Tasha McCauley and Helen Toner are aligned with the effective altruism movement.
Now we could talk a lot more, and we have talked on the show about the effective altruism movement. It is behind a lot of the AI safety research that is going on in the AI industry. And these are people who tend to be more worried about things like existential risk from AI. Now, we don't know who voted which way on the board, but some folks that I've been talking today said, Well, it's possible that this was some an EA-led coup against Sam Altman for some reason having to do with AI safety. But again, that is very vague and not confirmed, and so we probably shouldn't speculate about it too much. So, Casey, what do you think this does mean for the future of OpenAI? Can they survive this scandal as they have survived other scandals in the history of the organization? Or is there something different about this?
I mean, sure. I think the next 30 toand 90 days are critical for this company. Is it the case that a bunch of other top executives and engineers at this company follow Sam and Greg out the door and start another company with them? If so, then yes. I think that's a very serious challenge to OpenAI. I think it also might be one of the best days Anthropic has had in a while. It's almost certainly the best day Google Bard has had in a while, okay? Because these folks now have an opportunity maybe to get ahead while their top competitor is trying to pick up the pieces. The second thing I would say is they still have a lot of advantages. Gpt-4 is the best generally available large language model. Chatgpt is probably the most popular chatbot on the market. It's almost certainly the one that's making the money. And none of that is going to stop being true because as fascinating to us as it is who is running this company to most of its users, the question is just, Hey, can I sign up for ChatGPT Plus yet? It's interesting, Kevin, when you and I sat down with Sam, it seemed like the biggest problem in his life was that he had to pause signups for ChatGPT Plus because it had gotten too popular.
So they were in the middle of their own infrastructure and engineering crisis before the leadership crisis started. I don't think OpenAI is going to collapse tomorrow, but is this a very real risk to the company? I do think it is. What do you think?
Yeah. Something is just not heading up for me. It is truly inconceivable to me that the board of OpenAI would push out Sam and by extension, Greg Brockman, without a really good reason for doing so. This is a CEO who has generated enormous value for OpenAI. And for all of the flaws that his critics might find with him, he's been a very effective leader in terms of positioning OpenAI at the head of the pack when it comes to AI. I don't know, I guess we'll find out more in the coming days. But it just truly does not square to me that it could be something like a disagreement about AI safety or something without being something... This is also a board of people that Sam Althman, for the most part, picked. So it's not as if these are people who he had a history of open conflict with. True.
Something happened here. Look, something did happen, and while I cannot speculate and I don't have any reporting, I would just say that Semifour reported today that Sam has been in the middle of raising a new venture fund. He has other side projects. He co-founded Worldcoin. That is the org that scans your eyeballs and gives you crypto. And might be that none of that has anything to do with what happened today. But I would say that in general, if you are a CEO and you're also a CEO of other companies, or you're building other projects, or you're engaged in a lot of other complicated deals and investments, that stuff does get messy, actually. And there are disclosure rules. If you don't put the right thing down on a form, the next thing you know, you could find yourself in a lot of trouble. So we have actually seen a lot of CEOs go down over the year for what are essentially financial entanglements. So again, as I grasp for what could possibly have happened that would make the board do this, I have to say that's the number one place that I've been coming down.
Yeah. So aside from the main question here, which is what the heck happened, there's also this question of, well, what is next for OpenAI and for Sam Altonman? For OpenAI, I imagine under Miram Maradi, they will continue to refine and launch new products. I imagine Microsoft, which is heavily dependent on OpenAI for a lot of its AI work, will try to continue to work with the company, but I can't imagine that Satya Nadella, the CEO of Microsoft, was happy to see this news. He and Sam Altman were on stage at OpenAI's Dev Day just recently, and it didn't seem like there was anything but love between the two. So I think for a lot of employees of OpenAI, what I'm hearing is that they are trying to figure out what to do with themselves. As you mentioned, Sam was the reason that a lot of them came to work at OpenAI. And so I imagine that the recruiters of Silicon Valley are going to be very busy in the coming weeks.
Yeah. Also, no disrespect to the new leadership at OpenAI. They have very strong backgrounds. They're clearly incredibly talented technologists. But I wouldn't be surprised if there were people over at Microsoft feeling like maybe they actually just gained a lot of power over OpenAI and that maybe they can now steer the future a little bit more strongly than they could when they had Sam Altman at the helm.
Yeah. All right. Well, Casey, I'm sorry that our planned week off next week has been interrupted. This keeps happening. This summer, when we both planned to take a week off, we had the launch of Threads that was a big surprise, and we had to cut our vacation short to do that. I'm beginning to think the problem here is us. We plan to have an off week next week and drop this big Sam Alton interview, and now Sam Altman has been pushed out. So you could call it The Hard Fork Curse, or you could just call it Bad luck.
Yeah. Well, call it what you want. But we do like making the podcast. And thanks to everybody who sent us messages asking us to do an emergency pod. We did want to hop on and tell you what we know, which isn't much. But obviously, this story has our full attention. When we know more, you will be the first to know. And my guess is that Sam Alman is going to continue to play a role in the.
History of AI. Yes. This does not feel like someone who is ready to ride off into the sunset.
I, on the other hand, am absolutely about to ride off into the sunset. Actually, the sunset was a long time ago, and now I have to go eat.
Dinner, Ken. Okay, well, thanks for doing the emergency pod with me, and we will reconvene when we have more information.
Hardfork is produced by Davis Land and Rachel Cone. We're edited by Jen Poyak. Today's show was engineered by Corey Schreppel. Original music by Dan Powell. Our audience editor is Nelga Lockley. Video production by Ryan Manning and Dylan Burgesson. Special thanks to Paula Schumann, Piwik Tam, Kate Laprestee, and Jeffrey Miranda. You can email us at hardfork@nytimes.
Com. Especially if you're an OpenAI board member who knows what happened.
Yeah, that'd be great.