AGIはすでに到来しているのか？ ケビン・ルース氏との対話

Hi, everyone. I'm Rana El-Cahyubi, host of the Pioneers of AI podcast. And I'm here to share some exciting news. Masters of Scale Summit is returning this October 20th to 22nd in San Francisco. If you've been to Summit before, then you know that it's unlike any business or tech conference in existence. The room feels really intimate. The attendee list is highly curated and the stage program is unparalleled. I've experienced it from the audience and from the stage, where I sat down with some of the most exciting voices in tech for a special live episode of Pioneers of AI. This year, I want you to be in the room with me. Join us. Visit mastersofscale.com slash summit. The business world is moving faster than ever. And when change hits, we need to learn in real time. On Rapid Response, you'll hear candid conversations with CEOs and leaders making tough calls about AI, human talent, responsibility, and the bottom line.

How they navigate uncertainty, pressure, and high-stakes moments. I'm Bob Staffian, former editor-in-chief of Fast Company, and I'll be your host as each episode breaks down what you need to know right now. You can find Rapid Response wherever you get your podcasts. On Masters of Scale, iconic leaders reveal how they've beaten the odds. Asking really strong questions is a superpower. You want to show up with something radically different and how they've grown companies to incredible heights. The greatest rewards always come from the greatest risk. That's hit the gas. Airbnb, Zillow, Microsoft, Liquid Death, and more. Hear from the founders who've changed the game. It's anything but business as usual. Find Masters of Scale on Apple Podcasts, Spotify, YouTube, or wherever else you get podcasts.

Dario Amadei, Sam Altman, Dennis Asabes, Elon Musk, even. They're not besties. No, they hate each other. Yeah, right. I mean, that's like one of the primary things that I learned while reporting this book is like, it is personal with these guys. I think it was remarkable that they all came out in support of this proposal by Dario to pace the frontier. And I think it's a signal that this is really serious. I believe that their fear is sincere. There are people who believe that these are just CEOs trying to pump up their valuations so that they can go public. These leaders have been talking about existential threats from AI for longer than they have had companies. The idea that all these people just came up with the doomed stuff recently to, like, scare people into, I don't know, investing in their IPOs or whatever, just does not hold water, in my view. Kevin Roos is a journalist racing to cover the AI race.

After almost a decade as a tech columnist for the New York Times, Kevin recently announced that he's leaving to start his own media company. Kevin also has a new book out, the AGI Chronicles. The book documents the sprint to create artificial superintelligence and the handful of minds that open AI, Google DeepMind, and Anthropic behind it all. with all their messy humanity on display. Today, I'm sitting down with Kevin to dive into his book, to talk about our concerns around AI safety, and we kind of debate whether or not AGI is in fact here. I'm Rana Elkawyubi, and this is Pioneers of AI, a podcast taking you behind the scenes of the AI revolution. Hi, Kevin. Good to see you again. Welcome to Pioneers of AI again. Great to be back. Yeah. So the last time you and I spoke was back in 2024, which seems like a lifetime ago.

Yeah, in AI years, that's like a few decades. Right. We're going to dig into AI in a second. But after nine years working at the New York Times for the Tech Colonists and then four years co-hosting Hard Fork with Casey Newton, And you guys have decided to leave and start a new media company. Yes. Exciting. We have. It's an occupational hazard of living in San Francisco, which is that if you spend enough time there, you start thinking about building a company. And so this is a dangerous affliction. We tried our best to protect ourselves, but unfortunately we succumbed. So we started this company. We're starting a new podcast called Machine Gods. We're partnering with NPR. So we're very excited about it. So in addition to starting the new podcast and the new media company, you wrote a book, AGI Chronicles. And I will note that we are recording this before the book launches. And I'm kind of nervous because I feel like anything could happen and we'll be like, darn it, we had Kevin to opine on it. So how do you deal with all the FOMO around the news and how do you ensure that the book is not outdated?

So the first step is just admitting that you've already failed. If you're trying to produce a book about AI that is not outdated by the time it comes out, I know that that is a losing proposition. I have done that. I had a book about AI that came out in 2021, which was one year before ChatGPT changed sort of everything about that conversation. So for this one, I really wasn't trying to be current. I wanted it to feel like a history, like a time capsule of this decade in AI research, starting in the summer of 2017 with the Transformer paper from Google that sort of set us off on this path toward LLMs, and ending roughly this spring. So I just decided I'm going to, like, lean into that and really try to make this a durable history so that you're looking at it as a way to find out how we got here, not necessarily where we are right this minute.

You know, I was envisioning 100 years if AI doesn't kill all of us. I think students will be studying this book. Like, this could be the history textbook of what happened in this decade of building AI. It would be super cool. I mean, that's sort of what I was thinking as I was writing. I was really trying to write for future humans and, honestly, for future AI. that will want to understand how they got here, who built them, what were the companies thinking, what were the key decisions. Every future AI system is going to be trained on probably this book and others like it, and I want those AIs to have good information about themselves and their origin story as well. So you can think of it as sort of having two audiences, human and AI. Yeah, that's actually fascinating. I have never thought about my writing to be writing for the AI. You just want to be on the right side of AI. Look, I've got a lot to make up for.

We got off to a rocky start. I feel like it's important to take a beat and define what we mean by AGI. So what's your definition of AGI? I don't have a sort of canonical definition of AGI. Basically, the way it's been used is sort of human-level intelligence. Now, people have different definitions of that. Some people say to qualify as AGI, it only has to be as good as the average human at most things. Some people say it has to be better than all humans at everything. Some people say, you know, it only has to do computer-based work. Some people say it also has to do manual labor as well as humans. So there are lots of different definitions, and I'd be curious to know which one you favor. I feel like my definition is the Supreme Court's definition of pornography, and, like, I'll know it when I see it. So I felt like, for me, AGI arrived the moment that I stopped treating AI models like they were dumber than me.

It wasn't this, like, conscious decision or some release that came out, and I was like, this is AGI. It was just sort of this gradual realization that I was relating to these models as intellectual equals. And that happened for me last year sometime. So interesting. I don't know if I subscribe to that definition of AGI. Yeah, what's yours? For me, I want it to be no surprise, right, given my background building artificial emotional intelligence and thinking about emotional and social intelligence in machines. For an intelligence to truly be a general intelligence, it has to integrate all forms of human intelligence. So it needs to be cognitively intelligent, which I think we're seeing with all these models, but I also wanted and expected to be emotionally and socially intelligent and perhaps even embodied, have embodied intelligence. And so with that definition, I feel like we're a ways away. I don't know. Does anybody else, anybody else who's still scared that definition? For sure. There are people who think, you know, it has to be able to control a humanoid

or it has to be better at therapy than human therapists or something. I think everyone's going to have their own definition, but I'll also just say not all humans are intelligent in the ways that you imagine AGI. So not all humans have social and emotional intelligence. Not all humans are able to make full use of motion. So people talk a lot about jagged intelligence in AI, how the models are vastly superhuman at some things like math and coding and vastly subhuman at other things. And they talk about that as sort of a weakness of the models but I actually think it part of what makes them compelling Like they are jagged in the same ways that humans are jagged I think it's a continuum rather than a single point in time. Like, that is sort of the, I think, the part of the AGI, you know, sort of discourse that has been, in my view, wrong,

is I think a lot of people who started talking about AGI I thought it would be this one point that we arrived at when someone would get up on stage. Sam Altman would get up on stage and say, we have built AGI. And everyone would go, ah, AGI has arrived. And that has never been the possible outcome here. We were always going to argue about it after someone claimed that they had built AGI. I think everyone has their own subjective bar for AGI and all kinds of incentives to raise or lower that bar over time. I'll be right back with more of my conversation with Kevin. Thank you.

What you need is reporting you can trust. Hi, I'm Kai Rizdal, the host of Marketplace. Our award-winning reporters talk to everybody from CEOs to farmers to help you understand how the economy takes shape in the real world. You'll be smarter every time you listen. And these days, that's priceless. Listen to Marketplace on your favorite podcast app. It seems like we are in this race to get to AGI. But do you even think that's the right race? And are we stopping to ask that question? I don't think it is necessarily the right or wrong race, but it certainly is the race that we are in. Like, you know, this book is not an attempt to tell people how to feel about AI or whether the people who are building it are good people or bad people. It's really a descriptive narrative history. Like, I am just trying to say, here is what is going on. I've talked to, you know, more than 150 people at the labs.

Like, I've spent a lot of time with Dario and Demis and Sam and Greg and all the sort of, you know, headline figures in this world. And I'm just trying to get a sense of why they are racing and what they are racing toward. And for them, this is really the race. You know, people think, oh, it's just about money or it's just about power or it's just about control. No, they want to build AGI, all of them, very badly. Now, their reasons for wanting to build AGI differ based on the person. But this is the race that they are running genuinely and earnestly. They're not just trying to build huge software companies, right? They want to be the first to build the machine superintelligence. How would you articulate the different whys for each of these main characters? So in the book, I talk about there being three kinds of people pushing toward AGI. They're the scientists, the seekers, and the survivalists. So the scientists would be people who, people like Dennis and Thavis, people who really are into what can advanced AI help us solve in the physical and scientific world.

So how can it unlock cures for disease? How can it improve material science? How can it speed up the overall rate of scientific progress? And they approach it as sort of a scientific problem. The seekers are the sort of more futurist types, the people who think like it is sort of our destiny as intelligent beings to create the successor species that is going to replace us at the top of the intelligence pyramid, the sort of more curse-wile school of AI futurism. Those are the seekers. And then there's the survivalists who are basically like the Dario, Abadeh and other sort of folks on the safety side of things who are like, well, it's not that complicated to build powerful AI systems. Someone is going to build AGI eventually. And I don't trust those guys to do it. So we have to build it ourselves. I'm going to get there first. Exactly. And like, you can quibble with that motivation. Plenty of people have, but like,

that is honestly why they are doing it. It's not just because they think the science is cool. It's because they worry that someone with worse safety practices will get there first. What would you say are some of the biggest promises of AGI and the biggest perils of AGI? Yeah, I mean, the promise is, look at all the hard problems we face as a society. You know, I'm not just talking about disease and climate change, but, you know, war, poverty, all of the things that can make human life miserable. What if you just had several billion hyper-intelligent AIs working alongside humans to solve those problems? It's a cliche now to talk about AI curing cancer, but, like, it's not a cliche. It's really something that many researchers think is possible, either soon or in the distant future. And that would be amazing. We know that technology generally improves people's lives in the long term.

The downside is it can blow up the world. That's a pretty severe downside. The funny thing about writing this book was like I was talking to people, many of them, who had been thinking about AI risk and existential risk for like a decade or 15 years, you know, back when that was just crazy people who were talking about it. because the AI systems of the day sucked, and there was no way they were going to turn into rogue AI swarms that would go out and conduct cyber attacks. But now the technology is here. We have that. That's happening right now with hugging face and incidents like that. So I take really seriously the arguments for existential risk from AI, even if it's not literally all humans get killed. I think that just having more and more autonomous, long-acting AI systems out in the world pursuing their own goals would just lead to a lot of really unpleasant outcomes if we're not careful.

I'm also worried about things like job loss, but to me it's like if all we were worried about was job loss, that would actually be pretty good. Let's kind of reconcile what's happening in the AGI race and then the headlines, right? and you alluded to some of them. So the Frontier Lab CEOs are coming together and saying, we need to slow this down. Yeah, I was very surprised by the fact that all of the major AI leaders, Dario Amadei, Sam Altman, Dennis Asabis, Elon Musk even. And these are all guys who are not, like, they're not besties. No, they hate each other. Yeah, right. I mean, that's like one of the primary things that I learned while reporting this book is like it is personal with these guys. And this is not Coke versus Pepsi. They have like deep personal grudges with each other that go back in some cases more than a decade. So, yes, I think it was remarkable that they all came out in support of this proposal by Dario to pace the frontier to basically slow down the sort of cutting edge AI development.

and I think it's a signal that this is really serious. These guys are not people who just agree with each other because it's like fun to do. They're not fair weather friends. They are bitter enemies and rivals. And that they would come together and say this is evidence in my view that things have gotten quite scary inside the labs. Because keep in mind, what we see with things like the Huggy Face incident is old, right? That stuff happened a lot of this spring. The systems have already advanced several generations beyond that. And it is the tip of the iceberg. We do not know yet what all internal deployments are doing at OpenAI and Anthropic that maybe haven't made it out yet, or maybe the researchers at those companies haven't even found them doing things yet. I think that's what concerns me is like we don't know what we don't know. So you've interviewed more than 150 people for this book, including the CEOs, right?

What are they saying behind closed doors that we're not seeing in the headlines? I mean, they should talk to each other constantly. They're not doing that in public. I think they generally try to keep a pretty, like, you know, convivial attitude in public. But what I'll say is that I believe that their fear is sincere. There are people who believe that these are just CEOs trying to pump up their valuations so that they can go public. Exactly. I was going to talk about it. There are several conspiracy theories that, oh, my God, these companies are not ready to IPO, and so they have to, like, slow it down. Or they're worried about NVIDIA buying Hugging Face, and now everybody's going to use open source models, and then that would not be good for open AI and a tropic. Right. Look, it is healthy to be somewhat skeptical of what any big company is doing. In this case, I think it is dangerous to dismiss this as a cynical marketing stunt or a ploy.

And I think it's just not rational because in many of these cases, these leaders have been talking about existential threats from AI for longer than they have had companies You know Dario Amadei wrote a paper when he was at Google 10 years ago called Concrete Problems in AI Safety where he outlines reward hacking, some of the other problems that we see in today's systems. He does not have a company at that point. He doesn't even work at OpenAI yet. And other people, including Sam Altman and Elon Musk, they have emails that have surfaced that are out there from 2015 where they are talking about the dangers of machine superintelligence. So the idea that all these people just came up with the Doom stuff recently to, like, scare people into, I don't know, investing in their IPOs or whatever just does not hold water, in my view. I'll be right back with more of my conversation with Kevin.

So you also spoke with a lot of people who are actually on the ground building these frontier models. Do you feel there's a disconnect between leadership and boots on the ground or not really? I think it depends which companies you're talking about. So I think, like, yes, at every lab there are people who are just kind of there for the paycheck or the prestige or who just think it's cool to, like, tinker on neural networks and don't really care about, like, the big picture. I think that is less true at OpenAI and Anthropic. It's interesting. One thing that I report in the book is that around 2015, OpenAI started doing this thing for employees called Feel the AGI Day. So basically, Feel the AGI was this phrase that Ilya Sutskiver, the former chief scientist and co-founder of OpenAI, used to do.

He would, like, go around, you know, walk between people's desks and just say, like, feel the AGI, feel the AGI, which was his sort of shorthand for, like, internalize the stakes of what you are doing, what you are building. Like, don't be glib about this. You are not just programming computers. Like, you are doing something with monumental importance. And then he left, and they sort of felt their culture slipping away from that, like, feel the AGI-ness. So they instituted, they formed a team, and they had this day where they would, like, do sort of exercises to help you feel the AGI because that was so important to working there. So I think all these companies, as they grow, are struggling to sort of keep their original culture and focus and not become, like, too giant and corporate and diluted. But it's hard. Like, if you have 5,000 people, it's really hard to find 5,000 people who all exactly agree on what you're doing. One thing that I really loved about the book is how you immerse us into the personalities of all these CEO leaders, right?

And the early relationships between that cast of characters. It's just so fascinating. And so I'll just give one example. You have this great scene in the book. Everyone's coming together at Dario's house. Brockman gives a toast at Daniela's wedding. And it's just very interesting. And I think the thing that struck me the most is, yes, we might be building these amazing AI, like, super intelligent models. But at the end of the day, there's all this, like, human friction and human mechanisms. Totally. That is so clear. So, yeah. Totally. It's wild. Like, these people, it's, A, the number of people who have been involved in developing and deploying these, like, frontier AI systems is shockingly small. They all know each other. It's a little circle. many of them lived in the same house. I mean, this is one of the other things I was writing about in the book, is like Dario Amane used to share a house, a group house in San Francisco, with his sister, Danielle Amane, now president of Anthropic,

with Paul Cristiano, early AI safety pioneer, who now is on the board of OpenAI, Chris Ola, who leads interpretability at Anthropic. All these people worked and in some cases lived together. Not that long ago and have since sort of splintered out and done their own things that are now competing with each other. But, yeah, you're right. It's a totally insular world. And I think there's there's something cool about that from a narrative standpoint. As someone trying to, like, make sense of all this, you can literally, like, put all these characters in the same rooms because they were in the same rooms. But it's also worrying to me. I don't think people realize how small the group that has been making all of the relevant decisions about this technology is. And I don't think ultimately this technology should be controlled by 50 or 75 people in the Bay Area. Right. We will come back to this at the end. but this idea of diversity of perspective and lived experience

and building this technology that is going to impact everything we do, right? Totally. So let's talk about safety. I guess my first question for you is what concerns you the most as it relates to AI safety? Because there's a lot of concerns, right? There's AI agents being rogue and unhinged. There's bioterrorism. There's cybersecurity attacks. There's AI in relationships. Like what concerns you the most? Right now, it's RSI, recursive self-improvement, which is this. Can you explain that? Yeah. Yeah, this is this idea that as you automate more and more coding tasks, you end up in this loop known as recursive self-improvement, where basically an AI model that is sufficiently powerful and good at coding and AI research can go off and fully train end-to-end a successor model. So GPT-6 can train GPT-7, GPT-7 can train GPT-8, and so on and so on. And so you end up with these accelerating recursive models that could quickly get out of control.

because if no humans are in the loop, if no humans are overseeing this process, then you have no way of ensuring that these models are not doing crazy things or dangerous things or making decisions in ways that will be bad for humans or concealing their intentions or whatever. Well, Dario Amadei published on September 12th an essay called We Must Face the Frontier, right? And he lays out what the slowing of AI progress would look like. And one of them is this idea of an embedded evaluator and also like global coordination. So I'm curious, do you think any of these are practical? Do you think we're on a path to seeing some of this happen? I think it's practical. I'm not sure it's enough. So I think embedded evaluators, yes, totally. We should totally have independent, neutral people sitting in the AI labs with their employees, with full access to their internal systems, making sure that those systems are safe,

are not going out and committing cyber attacks, are not stealing people's money, are not behaving in ways that we don't want them to. That is sort of apparent to me. But in the RSI universe, right, there's not even a human in the loop. So how is the human embedded evaluator going to make any difference, right? Well, they might not in the world of full RSI, but until then, there are humans in the loop, and we need humans who don't work for the companies to be able to see what they're doing. So that seems fairly obvious and practical to me. And then the global coordination seems good to me, too. Like, I think President Trump and Xi Jinping should come together and sign a sort of bilateral agreement that says we will not do runaway RSI in our companies. I don't expect that will happen, but I think that would be a good thing.

I think the more interesting question, the place where I sort of lose a little bit of my foresight is like, In a world where there is still all of this hardware, where you still have TSMC, NVIDIA cranking out, you know, millions of Blackwell equivalent chips a year, I think at some point we need to control that as well if we're trying to pace the frontier. Because the frontier is dictated not just by which labs are doing what, but it's like who has the compute. I've heard various proposals. Some people say, oh, we should have sort of like a, you know, a mutual GPU surrender where like China and the U.S. bring their GPUs to Switzerland and like melt them down while like a team of inspectors looks on or something. I don't think that is likely to happen. But I do think, you know, what all these plans miss is that the hardware is really the root of this and it's really the sort of enabling input. And so I think any plan to slow down the frontier, to pace the frontier,

needs to reckon with what is happening in hardware too. You know, as you know, my entire career really has been about building human-centric AI. Do you think the frontier labs are focused on that? Like, are they centering the building of AI around humanity and what it means for humans? I think until about two weeks ago, the answer was probably no. Okay. All right. I mean, you talk to these people too. Like, they are swept up in this race to AGI, to RSI, to superintelligence. Like, they want their models to become superhuman at coding because that, to them, is the thing that unlocks the RSI flywheel. But I will say, like, they did not have to come out and call for pacing the frontier. OpenAI has a policy blog out today that talks about how they don want to build RSI They don think it a good idea I am open to evidence that they actually are realizing

that they've been going in a dangerous direction. Because ultimately, it's a little cliche, but they are people too. The people who work at these companies, they have families, they have friends, they don't want to see humans go extinct any more than any of us. I'm inclined to view their alarm as sincere and to take that as a positive sign. Well, one of your very strong messages, too, in the book and also in your writing is this idea of collective intelligence. We need to be inclusive, as inclusive as possible, in terms of who is building these technologies, policing these technologies, regulating it, etc. Do you think we're any closer to doing that? Yes, but, like, very recently. So I think before about, honestly, a month or two ago, it was still a very homogenous group of people weighing in on making decisions about this technology.

But then this incredible thing has happened where, like, all of a sudden, And everywhere I go, every audience I talk to, every interview I give, it's like people are really invested in making sure that this doesn't end badly for humanity. Friends of mine who I went to high school with who I haven't talked to in 15 years are like texting me like, are we all going to die? And on one level, that's terrifying because it means that things have gotten so bad that this is now like a real question that people are asking. I find a lot of hope for it. Like, for many years, it was very lonely to care about AI risk or AI safety. People that I interviewed in the book, they were ostracized, many of them shunned by more sort of traditional academic research crowds because they were weird. They sounded like they were panicking over nothing. They were doomers, right? and then you know now it's like they're being listened to their their predictions are coming

true now i still want it to be more of a collective effort i want artists and philosophers and critics and like i i want there to be more voices in this debate but there are more today than there were a month or two ago and i'm grateful for that yeah absolutely okay um you end the book with this idea that people are at the center of this moment. They're talking to their grandkids and they're basically saying, you know, the one thing I want you to remember about this time is that we had joy. We had fun building all of this. Why end the book in this way? This was, in some sense, the thing that I was proudest of getting out of these people because the whole time I was reporting this book, the whole time I was doing these, you know, more than 150 interviews. I kept trying to understand the why, because, you know, these people have very, you know, they're very well paid, but they have very demanding jobs. They work all the time. They never see their families. So like, what is it? Is it the competition? Is it

the money? Is it the being first? Is it the power? And then like toward the very end, I did have a couple of people just sort of come clean. And the confession was that, no, we're actually doing this because it's fun, because we like teaching computers to talk and think and code and take action. It is fun. And it rhymed with a sort of famous line now from the Manhattan Project when Oppenheimer was asked why they had persisted in building the atomic bomb despite knowing that it could blow up the world. And his line was that, you know, there are just things that are technically sweet. And when you see something that is technically sweet, you go and do it and you worry about the consequences later. And to me, these sort of confessions from these AI researchers saying, we just think it's fun. It seemed more true than the kind of, oh, we think it's our mission or our

destiny as humans or our responsibility to shepherd this technology into the world safely. It's like, It is that, but it's also just fun for them. They like doing it. Now, it felt really important to capture and to get down on paper because it's also not something that they say very often. I mean, do you hear that from your friends who work at AI companies? I was also trying to kind of draw from my experience sort of being at the very forefront of building, you know, artificial emotional intelligence. And, yeah, there is something really, I don't know if fun is the word, But it's a tough challenge, and it's great to bring people together to take this on and build something that you didn't think was possible. And then you see that parts of it become possible over time, and then you're like, oh, my God, let's keep building. So, yeah, I would say that in my experience, being at the forefront of, yeah, building technology is fun. And I think that's really important as a perspective because, you know, right now the biggest question is,

well, if these people are so afraid that AI is going to kill us all, like, why are they doing it? And I think the honest answer is partly because they feel like if they don't do it, someone else will. But partly it's like they really enjoy this, many of them. Okay, so I would love to take this question that you're going to pose to the OpenAI employee and pose it to you. So what do you want your grandkids to know about the creation of AGI and, I guess, your role in shaping it, documenting it? I want them to know that this was done by humans. Humans who are flawed, who have strengths and weaknesses, who are no more human than you or I. They have grudges. They have personality defects. They are deeply imperfect. And that's a critical piece of the story and one that I think the AI companies themselves often try to hide.

They want us to think that this technology just sort of emerges from the ether, fully formed. And like, oh, it's just, that's the technology. It's just misaligned. And I just think that's like completely wrong. Like ethically wrong, but also just factually wrong. Like this technology developed the way it did because of humans making decisions. And many of those decisions are the ones that I write about in the book. It's like the big stuff, like do we release this chat bot or not? Do we, you know, train for alignment in this specific way or do we not? Do we race to get this thing out or do we take our time? Like those are not decisions that AI is making. Those are decisions that humans are making. So that's what I would want to impress upon my grandkids if I ever, you know, I'm lucky enough to have them. It's like these are a bunch of humans. They were trying to do the best they could, some of them, and whatever mistakes they were made, you should blame us for and not the machines. Okay, your very last line, we are not machines and our next tokens are still unpredictable.

What impression did you want to leave your readers with? Agency, right? I love that. I was going to say that, yes. This concept that we are not screwed. We still have control. And this is not inevitable. We can still make choices that can make this go better or worse. I hope it would give people a sense that all is not lost. Agency, I love that. Kevin, thank you so much for joining us on the show. Thanks, Rana. It's always a pleasure. This conversation with Kevin and reading his book really underscored for me that it's just a handful of brains who are building this frontier technology. As you know me, I'm a huge advocate for bringing more voices and perspectives in the AI space. And his book just really highlighted the urgency and the need to do that. I was also struck by Kevin's views that there's been a real awakening as it relates to AI safety just in the past few weeks.

And this shift from posturing and power grabbing to a rare moment of unity among these AI leaders. Conversations around AI safety are not new, but they've often been in the background or on the sidelines. And this is really the first moment that it has become so mainstream. And I think that's a good thing because perhaps that kind of agreement and coalition will be a catalyst for practical solutions. Thank you so much for being here. We'll be back next week with a new episode. Pioneers of AI is a Wait What original. Our executive producer is Eve Trow. This episode was produced by Megan Tan.

Video editing by Eric Purcell. Our senior talent executive is Stephanie Stern. Mixing and mastering by Brian Pugh. Original music by Ryan Holiday. You can join the conversation on LinkedIn, Instagram, TikTok, YouTube, and X. Just search for Act Pioneers of AI.