Sam Altman & the Battle for OpenAI | Is OpenAI Out of the Woods? | 4

0:05I'm David Brown and this is Business Wars.

0:30The tech startup OpenAI quickly became an industry leader after its founding in late 2015. But as its technology became more and more powerful, there was drama at the top. And we covered the attempt to oust its CEO in our latest series, Sam Altman and the Battle for OpenAI. But this soap opera is still ongoing. Since Altman was reinstated as CEO after the unsuccessful coup, The new board launched an independent investigation to determine whether his firing was justified. They declared it wasn't, and Altman was allowed to resume his membership on the board. The report was thin on details, however, and disappointing to those looking to understand the board's rationale in trying to remove Altman in the first place.

1:15Our guest today is Julia Black, reporter at the subscriber-based independent tech publication The Information. And she joins us to help us understand OpenAI's transformation from a mission-driven AI company looking to save the world into one that critics say is moving too quickly. We'll talk about Sam Altman's leadership, the latest developments at OpenAI, and whether the future of AI is in hands we can trust. Stick around.

1:46When planning for your future, you want someone with a history of keeping their word year after year. For nearly 160 years, Pacific Life has been a trusted name in the industry. But that isn't just a number. It's experience that matters. It's 160 years of promises held, helping generations retire with confidence, protect their loved ones, and plan for whatever comes next. Whether you're looking for life insurance, employee benefits, or retirement income solutions, when your future is on the line, you want history on your side. And believe me, Pacific Life has been there, always there through changing times, always focused on your needs.

2:24Ready to secure your tomorrow? Ask a financial professional how Pacific Life can help you feel prepared for what's next. Pacific Life Insurance Company, Omaha, Nebraska, and in New York, Pacific Life and Annuity, Phoenix, Arizona. because with Pacific Life, you're not just planning for the future, you're partnering with trusted experience. Closing the books, getting your people paid, and bringing on new hires. Running a small or mid-sized business can be exciting and also a little chaotic. Workday Go makes simplifying your business, well, simple. Imagine all the important aspects of your company, HR, finance, and payroll, all on the AI platform.

3:06No more juggling multiple systems, no more worrying about growing too fast. Just the full power of Workday Go, helping small to mid-sized businesses like yours scale and run more smoothly. Think about what that means. Seamless onboarding for new team members, real-time insights at your fingertips, and payroll that works perfectly every single time. So you can focus on the big picture and go after your big ambitions. And with Workday, you can activate quickly in as little as 30 to 60 business days. So simplify your business. Go for growth. Go with Workday Go.

3:52Julia Black, welcome to Business Wars. Thanks so much for having me. You've been reporting on tech, politics, and AI for a while now. And I wonder what fascinates you most about the landscape right now? I think the thing that interests me most right now, and really forever, is going to be characters. But I do think we're in a moment right now where this kind of Mount Rushmore of Silicon Valley is more important than ever. And I think in the world of AI in particular, it's very important to understand these people, their beliefs, their personalities, their approaches, because they're shaping systems that mean that those very personalities will trickle down into the technologies that shape the rest of our lives.

4:33Well, too bad there aren't any characters in Silicon Valley. You know, that's the thing. Yeah, poor me. Let's talk about what happened with OpenAI. Most recently, back in March, an independent investigation commissioned by the OpenAI board found that Sam Altman was wrongfully terminated back in November. What were the other key findings of that report? You know, I hate to be cynical, but I do think that when these super large firms hire these kind of outside investigators, a lot of this is for show. They weren't super transparent about the process. They did say that they interviewed various OpenAI employees.

5:12They reviewed documents. But yeah, not a lot of the findings actually reached the public. They just kind of came away with the conclusion that there was no definitive reason to fire him. That doesn't mean that they didn't find issues there. That doesn't mean that they didn't interview people and have them confirm some of the stuff that the board was saying. But I do think that the so-called coup, whatever you want to call it, was so messily handled. It happened so fast and in such a way that I think, you know, the board alleges that their intention was not for it to end up with his firing. But it all seems to have gotten out of control so quickly.

5:56So, yeah, I do think that the investigation saying that there was no reason for his termination is not the same thing as them saying there were no issues here. Did the verdict or the outcome come as a surprise to you or to anyone else? Or was this, by and large, what folks were expecting, people who were following this story? Gosh, it's hard to say now or, you know, to remember how it felt that week. It was one of the few cases in my career where I've had more sources coming to me asking questions than I've brought to sources. So I think to say, oh, in retrospect, we all knew what was going to happen would be a lie.

6:36It was complete tumult for that week. You know, I had former open AI people contacting me. Have you heard anything? What do you think is going to happen? Everyone was speculating. It was that much of a closed box there. Oh, absolutely. Yeah, I was privy to, you know, open AI group chats where people were talking to each other, exchanging theories. People were really kept in the dark. And to this day, I think there are so many unknowns. When you think about those unknowns, I guess in a way, it seems like this report was prepared to provide a level of transparency so that people have more confidence.

7:10And this was a big part of our series. When Altman returns to the company, there are others on the outside saying, well, should we do business with a company given that kind of apparent instability? And I'm wondering, as you look at the report, what is it that it doesn't say? What's missing? If I had to pick the number one question currently on people's minds, it's what's the deal with QSTAR? What is the deal with QSTAR? You know, if only I knew. But QSTAR is a project, a secretive project that they were working on internally at OpenAI. And allegedly, what is currently understood publicly is that this development within the QSTAR project was advanced enough that it really alarmed certain folks on the OpenAI team who are part of this contingent who's really worried about AI taking over the world and committing mass murder and all these horrific things.

8:08And so that ties into probably the second biggest mystery right now, which is what's going on with Ilya. That's Ilya Sitzkever. And he was supposedly the person who escalated this concern. Sort of behind it all was the way he was kind of characterized in a lot of the media. Yeah, yeah. I mean, a lot of people were accusing him of being kind of a backstabber. But at the end of the day, I think that the problems that unfolded that week were things that were baked in at the very beginning of OpenAI. I mean, it is fundamentally such a strange corporate structure. The incentives and the motivations of the company are arguably inherently misaligned.

8:54It's impossible to both operate as a nonprofit and this capped profit company because, you know, it's really hard to be answerable to the world at large as well as shareholders who want to maximize profit. There's a reason most of our system is separated into one or the other, your nonprofit or your for-profit. Do you sense that there remains a divide within open AI's leadership now? And if so, what does that mean for the company and its goals? I think it reflects a divide that you'll find in just about any circle working on the AI problem. It's what makes AI so fascinating to write about and these characters so fascinating to write about.

9:35There's this inherent contradiction where the people who know the most about AI, who are developing it, are the researchers forming these models, often tend to be the very people who are most worried about it. So I think you have a significant contingent within the OpenAI team. A lot of them are on something called the super alignment team. And these are the people who are affiliated with a movement called effective altruism. Another sort of realm is rationalism. And these are conversations that started in online communities years ago. There were these message boards where basically, you know, sci-fi nerds would gather.

10:18And they had a lot of knowledge. These are really smart people. They had a greater understanding of AI at the time than obviously most of the world. But they really started to go off into some pretty outlandish directions. I think that the best way of understanding this group of people is to look at something called the paperclip maximizer. The paperclip maximizer? The paperclip maximizer. So this is a pretty wacky thought experiment developed by this community. And the basic principle is, what if you were to teach a super advanced AI system? What if you were to give it the input? I want you to make as many paperclips as possible.

10:54So, you know, any human would probably take that directive and say, okay, I'm going to have to gather the material for paperclips and then start a factory. But they worry that an AI system that hasn't been trained with that kind of sophistication could make the mistake of understanding the directive as, I must take all possible material in the universe. That includes this table in front of me. That includes ripping the carpet out. It also even includes using human bodies as material to make paperclips. And so in this doomsday scenario thought up by these pretty wacky people, an AI system, given a prompt that it could misinterpret, could accidentally turn all of matter in the universe into paperclips, and it would be the end of the world.

11:44Obviously, that's absurd, but that's the kind of thought experiment that shapes this group's thinking. And that very group that started on these online communities has really trickled out into Silicon Valley, is really like this is dinner party, cocktail party conversation in the AI community. And a lot of those people who came from these communities do work at OpenAI. I think Sam Altman has some of those concerns, but I actually was listening to a podcast he did the other day, and he said himself that he goes back and forth on the question of the danger of AI. But currently in charge of this company, OpenAI, his primary goal seems to be to protect the company, to protect shareholder value, and not to think so much about those doomsday scenarios.

12:41If Sam Altman says that this is where he's at, does that reflect the way the board is thinking from what you can tell? Certainly, I think given that he more or less handpicked the new board, they reflect his thinking. I think he put people into place who are much more aligned with him and his vision. And he kind of got rid of those pesky, effective altruists who are, you know, chicken little, the sky is falling. So he's purged a lot of that energy from the board and, frankly, from the ranks of OpenAI. It's just not as dominant, I think, in the makeup of the staff as it was probably OpenAI's founding.

13:24Now, you mentioned Ilya Sutskiver, a missing link, I suppose you might say. He was one of the board members advocating for Sam Altman's firing, but seemed to reverse his position shortly after it happened, at least based on a lot of the social media messaging at the time. Where is it believed Dr. Sutskiver is now in all of this? You know, I heard Sam Altman did a two-hour interview on the Lex Friedman podcast this week, and Lex asked him the very same question. He said, is, is Ilya in a nuclear facility somewhere? And Sam just wouldn't answer the question. He wouldn't answer where he is. He wouldn't answer what his role is at the company.

14:03So they continue to shroud that in mystery. I'm surprised Lex couldn't pry it out of him. He, he's usually pretty, pretty dogged about that sort of thing. But, but Sam Altman was saying nothing. Yeah. Yeah. Everyone's staying pretty tight lipped on Ilya. Um, and you know, As that was all unfolding in November, it seemed to be a pretty dramatic interaction between the two of them. Reports were coming out that there was a tearful kind of begging on his knees type interaction where Ilya said, I'm so sorry. I don't know what I was thinking. I take it all back. So, yeah, it's really impossible to overstate how messy that week was.

14:47It was just people running around in circles with their heads cut off. What does this do, this drama? And frankly, I can't think of anything like this in the public sector. The only thing that comes to mind when I think about this kind of drama with so much at stake, seemingly, might be back in the early 60s in the government during the Cold War days, you know, when the military and the civilian administration was, there was all this fighting. And then out in the real world, the stakes were so high, it was literally nuclear, you know. You look at the stakes here with AI and you think about the players and the core of the argument and the philosophy.

15:27And I'm thinking, OK, well, Sam Altman now been reinstated to OpenAI's board of directors in addition to serving as CEO. What does this mean for OpenAI's next steps? Yeah, you know, I think up until November, Open Eye was the undisputed dominant player in this space. And I would think a few months later, that's less the case. They have some competitors who've been nipping at their heels. And I think that this did probably shake public trust in their work and their ability to succeed. They no longer have this kind of near monopoly on the space. They have a lot of competitors who are getting a lot of funding.

16:15I would think that more and more it's going to become this crowded marketplace where, again, it goes back to the personalities. I think you're going to start thinking more about who's running each company, who's designing each model, and what they're going to give you. So rather than OpenAI being your only option, Chat2PT being the chatbot you use, you're going to have different consumers coming to this marketplace and saying, do I want an AI system designed by Elon Musk that's anti-woke? Do I want an AI system that's designed by the original dominant player? Or maybe you'll say, oh, OpenAI is old news.

16:52They already had their chance. Now I'm excited about Anthropic and Claude. So, you know, I think it's going to become more and more of a marketplace for different types of models, and they're going to have to start distinguishing themselves more and more from each other. Julia Black covers the intersection of technology, media, and politics for the information. Let's take a short break. Coming up, we're going to talk about OpenAI's origins and where they're going next. Stay with us.

17:19December can be a busy month for all of us. Finishing projects, planning for next year, maybe even wrapping a few presents. while your teams take a well-deserved break. Cyber threat actors don't. Stay protected this holiday season with Manage Engine. Their cybersecurity solutions ensure that your data is secure, even while your office lights are turned off. With 18 data centers around the world and over 20 years of experience securing global enterprises, Manage Engine can keep your business running safely, no matter how busy or hectic the world gets. This holiday, gift yourself and your business peace of mind.

17:58Visit cybersecurity.manageengine.com. That's cybersecurity.manageengine.com. This message is brought to you by Apple Card. Apple Card members can earn unlimited daily cash back on everyday purchases, wherever they shop. This means you could be earning daily cash on just about anything, like a slice of pizza from your local pizza place or a latte from the corner coffee shop. Apply for Apple Card in the Wallet app to see your credit limit offer in minutes. Subject to credit approval, Apple Card issued by Goldman Sachs Bank USA, Salt Lake City branch. Terms and more at AppleCard.com.

18:47hey welcome back to business wars our guest is julia black a reporter at the tech industry publication the information julia let's rewind back to open ai's founding this is a company that developed its technology very quickly how they managed to do that i mean was was this very much Sam Altman writ large, or did he have much of a hand in that? Yeah, I think that one of the reasons they have found themselves in this first place on the leaderboard among AI companies has been winning the talent wars. I mean, I think the fact that the company was founded by Elon Musk, Sam Altman, obviously set a high bar.

19:30And I think that AI researchers, in the last decade, there's been a very narrow talent pool. And I think Open AI was successful at snapping up a lot of that talent early. So I think that's part of it. I think the other thing that's kind of set them apart has been their brazenness. They've been absolutely audacious in their use of training data, just, you know, using anything and everything they can get their hands on, whether or not they had the rights to it, which, you know, now they're starting to see the blowback of lawsuits and controversy over the use of creative content that they haven't licensed.

20:15So, you know, I think that they just had really good access to talent, training data, and yeah. I'm really interested in hearing what you got to say about Sam Altman's place in this. You know, Sam Altman, he didn't have the kind of force of personality, I think, at least on the national stage, that someone like an Elon Musk might have. Have you gotten a chance to get to know him? And maybe there's some reason for that, you know, why he seemed to be such a prominent presence, you know, in Silicon Valley, but only recently has sort of started to pop into the national consciousness, if you know what I'm saying.

20:56Yeah, absolutely. I can't say I know Sam well. I have interviewed him. I have written about him extensively, thought a lot about him. And yeah, the experience of interviewing him was a fascinating one. He's kind of different from a lot of his Silicon Valley peers. It's a group that is largely known for its awkwardness socially, if I'm being honest. And he's a pretty charismatic guy. And so it's been interesting to watch him kind of enter the public eye and in real time kind of experiment with trying on different masks is the way I kind of see it. I think he's very careful and thoughtful about his communications.

21:40And, you know, as the board drama started to play out, that is something that was perceived as manipulation. That's one way of seeing it. Another is just, let's just say I've heard the word Machiavellian used about him more than once. Interesting. But yeah, I mean, I think he has this outward facing character that's very personable, very friendly. He always has had a good relationship with journalists, which is more and more rare these days. He writes his friendly little emails with smiley faces and lowercase letters and always sounds so eager to help. But then sometimes you look at the amount of power he's wielding and you wonder how real that can be or if it's all part of a larger goal.

22:26Let's talk about the personal philosophies, though, of the OpenAI board members. You know, at the beginning, it seemed like those board members and founders were very much aligned. But as the company expanded, there was a divergence. What was the tipping point? What seems to have been the tipping point? Sure. I think this brazenness really is fundamental to Sam's approach and to OpenAI's approach. The fact that they were willing to release a model into the wild, the way that the public started to interact with ChatGPT really before it had interacted with any other model, that was a really bold move.

23:07It was this kind of overnight release that suddenly felt like it changed the world, changed everything. And Sam's philosophy was very much the move fast and break things philosophy that, has become kind of infamous in Silicon Valley. So his idea was, let's release this model. It may be imperfect, but we have this advantage of allowing the public to interact with it. And then the public can kind of start to train it. So I think the willingness to go there was what set OpenAI apart from its competitors, but also what really worried this contingent of people who have concerns about AI safety. I think a lot of people said, whoa, whoa, whoa, we haven't even put the guardrails in place.

23:51We don't even know what that looks like. We don't know how to regulate this. And yet you're just releasing it into the wild and hoping for the best. So I think that the release of the models into the public was the turning point that started to alarm people. So based on your reporting, was it more Sam Altman that brought this about, that he had lost touch with the original mission? Or do you think the board was perhaps overreacting to what Altman was doing? It's a hard question to answer. I think that Altman probably started to feel pressure from his investors. And again, it just gets back to this fundamental kind of core problem with this company.

24:36It's like, who should he be answering to? A board of people who come from this effective altruism background or investors who just say, go, go, go, make that profit, grow, scale. That's very much typically how Silicon Valley companies operate. So that more restrained board, I'm not shocked that they got pushed aside. It's not, it's kind of no fun among the Silicon Valley folks who just want to accelerate at all costs. So let's talk about Microsoft's role here. The information reports that the company's pledged$13 billion in funding to OpenAI. How symbiotic is this relationship between Microsoft and OpenAI, or is there one party that benefits more here?

25:23Where is this going? I think everybody benefits. You know, one thing that Microsoft brings to the table is Azure, its cloud computing platform. So OpenAI gets access to those kind of resources for cloud computing. And obviously, Microsoft gets the enormous benefits of having, getting to associate with the most transformative company of the last few years. Because Microsoft was, it's hard to remember it now because they're in this, they're flying high at the moment. But they really went through this kind of deeply uncool moment in Silicon Valley where no one really wanted to work with Microsoft.

26:03No one wanted to use their products. So their willingness to go all in on AI has been transformative for them as a company. And I think Satya Nadella deserves a lot of credit for that. He's kind of this mastermind who's brought them into a new era. So I think everybody wins. I think the other interesting thing to keep an eye on is the increasing consolidation of the AI powers that be under the Microsoft umbrella. Mustafa Suleiman had been hired to work at Microsoft. And Mustafa had been working on his own company, Inflection AI, which had gotten a lot of funding from a lot of big names and was shaping up to be one of these competitors in the space.

26:48So for him to be absorbed by Microsoft is just another signal that they are becoming this absolute behemoth in this space. And really, it's less that no one can compete with OpenAI. The question is, can anyone compete with Microsoft, which is going to be gobbling up all of these relatively smaller companies? Let's turn to Elon Musk, famously an early investor of OpenAI, a former board member. Back in February, he sued the company for straying from its mission. But it also seems like Musk feels a sense of ownership, maybe, over the company, even though he hasn't been a part of it since 2018. This is Musk in a CNBC interview from May 2023.

27:31Let's listen. So you helped to the creation of OpenAI. You put in as much as$50 million. More than helped. It wouldn't exist without me. It wouldn't exist without you. That's a really large claim. It wouldn't exist without me. What do you make of Elon's perspective here? So first I'll say I think the lawsuit is pure theater. I don't think it's at all a serious lawsuit. Why? What's the purpose? What does this serve? So Elon Musk is currently launching his own AI competitor. So the reason I think the lawsuit is pure theater is because his incentive is, you know, to publicly put down this other competitor who is currently at the forefront so that Grok, his chatbot, can move in.