Jeff Bezos spent his childhood summers working on his grandfather’s ranch in Texas, learning self-reliance and problem-solving skills. They fixed a broken-down bulldozer together, showcasing his grandfather’s resourcefulness. Transcript: Speaker 1 Most interesting. Most memorable. Most memorable. Most impactful. It was a real working ranch. All my summers on that ranch from age four to 16. And my grandfather was really taking me in the summers. In the early summers, he was letting me pretend to help on the ranch, because of course, a four-year-old is a burden not to help in real life. He was really just watching me and taking care of me. He was doing that because my mom was so young. She had me when she was 17. And so he was sort of giving her a break. And my grandmother and my grandfather would take me for the summers. But as I got a little older, I actually was helpful on the ranch. And I loved it. I was out there. Like, my grandfather had a huge influence on me, huge factor in my life. I did all the jobs you would do on a ranch. I’ve fixed windmills and laid fences and pipelines and done all the things that any rancher would do. Vaccinated the animals, everything. But we had my grandfather, after my grandmother died, I was about 12. And I kept coming to the ranch. Then it was just him and me, just the two of us. And he was completely addicted to the soap opera, the days of our lives. And we would go back to the ranch house every day around 1 p.m. Or so to watch days of our lives. Like, sands through an hourglass. So are the days of our lives. Just the image of that, the two of us. Watching a soap opera. He had these big crazy dogs. It was really a very formative experience for me.

The Apollo program demonstrated that seemingly impossible tasks can be achieved with enough resources and determination. It was pulled forward in time due to the space race, making it a technical marvel and a great human achievement. Transcript: Speaker 1 Well, I mean, there’s so much inspiring there. One of the great things to take away from that, one of the great von Braun quotes is, I have come to use the word impossible with great caution. Yeah. Yeah. Yeah. And so that’s kind of the big story of Apollo, is that things, you know, going to the moon was literally an analogy that people used for something that’s impossible. You know, oh, yeah, you’ll do that when men walk on the moon. Yeah. And of course, it finally happened. So, you know, I think it was pulled forward in time because of the space race. I think, you know, with the geopolitical implications and, you know, how much resource was put into it, you know, at the peak that program was spending, you know, 2 or 3% of GDP on the Apollo Program, so much resource. I think it was pulled forward in time. You know, we kind of did it ahead of when we quote unquote should have done it. Yeah. And so in that way, it’s also a technical marvel. I mean, it’s truly incredible. It’s, you know, it’s the 20th century version of building the pyramids or something. It’s, you know, it’s an achievement that because it was pulled forward in time, because it did something that had previously been thought impossible, it rightly deserves its place As, you know, in the pantheon of great human achievements.

Space exploration and industry are essential for preserving Earth. By moving heavy industry to space, we can continue to use more energy per capita while protecting Earth’s ecosystem. Transcript: Speaker 1 Exactly. This planet, we’ve sent robotic probes to all the planets. We know that this is the good one. Not to play favorites or anything. Earth really is the good planet. It’s amazing. The ecosystem we have here, all of the life and the lush, the plant life and the water resources, everything, this planet is really extraordinary. And of course, we evolved on this planet. So of course, it’s perfect for us. But it’s also perfect for all the advanced life forms on this planet, all the animals and so on. And so this is a gem. We do need to take care of it. And as we enter the Anthropocene, as we humans have gotten so sophisticated and large and impactful as we stride across this planet, that is going to, as we continue, we want to use a lot Of energy. We want to use a lot of energy per capita. We’ve gotten amazing things. We don’t want to go backwards. If you think about the good old days, they’re mostly an illusion. We can almost every way, life is better for almost everyone today than it was, say, 50 years ago or 100 years ago. We live better lives by and large than our grandparents did and their grandparents did and so on. And you can see that in global illiteracy rates, global poverty rates, global infant mortality rates, almost any metric you choose, we’re better off than we used to be. We get antibiotics and all kinds of life-saving medical care and so on and so on. And there’s one thing that is moving backwards and it’s the natural world. So it is a fact that 500 years ago, pre-industrial age, the natural world was pristine. It was incredible. We have traded some of that pristine beauty for all of these other gifts that we have as an advanced society. And we can have both, but to do that, we have to go to space. And all of this really, the most fundamental measure is energy usage per capita. And when you look at, you do want to continue to use more and more energy. It is going to make your life better in so many ways. But that’s not compatible ultimately with living on a finite planet. And so we have to go out into the solar system. And really, you could argue about when you have to do that, but you can’t credibly argue about whether you have to do that.

Blue Origin’s Blue Ring spacecraft is designed to transport payloads up to 3,000 kilograms to geosynchronous orbit or lunar vicinity. It provides various services to payloads, including thermal management, electric power, compute, and communications. Transcript: Speaker 1 The blue ring is a very interesting spacecraft that is designed to take up to 3,000 kilograms of payload up to geosynchronous orbit or in lunar vicinity. It has two different kinds of propulsion. It has chemical propulsion and it has electric propulsion. And so you can use blue ring in a couple of different ways. You can slowly move, let’s say, up to geosynchronous orbit using electric propulsion that might take 100 days or 150 days, depending on how much mass you’re carrying. And then reserve your chemical propulsion so that you can change orbits quickly in geosynchronous orbit. Or you can use the chemical propulsion first to quickly get up to geosynchronous and then use your electrical propulsion to slowly change your geosynchronous orbit. Blue ring has a couple of interesting features. It provides a lot of services to these payloads. So the payload, it could be one large payload or it could be a number of small payloads. And it provides thermal management, it provides electric power, it provides compute, provides communications. And so when you design a payload for blue ring, you don’t have to figure out all of those things on your own. So kind of radiation tolerant compute is a complicated thing to do. And so we have an unusually large amount of radiation tolerant compute on board blue ring. And your payload can just use that when it needs to. So it’s sort of all these services. It’s like a set of APIs. It’s a little bit like Amazon Web Services, but for space payloads that need to move about an Earth vicinity or lunar vicinity.

Jeff Bezos describes his thinking process as involving lots of wandering and allowing for inefficiency in invention. He emphasizes the importance of giving oneself permission to wander and not knowing where you’re going when solving problems. Transcript: Speaker 1 Well, that’s very kind. I have, I’m an inventor. If you, if you want to boil down what I am, I’m really an inventor. And I look at things and I can come up with atypical solutions and, you know, and then I can create a hundred such atypical solutions for something. 99 of them may not survive, you know, scrutiny. But one of those 100 is like, hmm, maybe there is, maybe that might work. And then you can keep going from there. So that kind of lateral thinking, that kind of inventiveness in a high dimensionality space where the search space is very large. That’s where my inventive skills come. That’s the thing I’m, if I, I self identify as an inventor more than anything else. Yeah. Speaker 2 And he describes in all kinds of different ways, Walter Isaacson does that creativity combined with childlike launder that you’ve maintained still to this day. All of that combined together. Is there like, if you were to study your own brain introspect, how do you think about what’s your thinking process like? We’ll talk about the writing process of putting it down on paper, which is quite rigorous and famous at Amazon. But how do you, when you sit down, maybe alone, maybe with others and thinking through this high dimensional space and looking for creative solutions, creative paths forward? Is there something you could say about that process? Speaker 1 It’s such a good question. And I honestly don’t know how it works. If I did, I would try to explain it. I know it involves lots of wandering. Yeah. So I, you know, when I sit down to work on a problem, I know I don’t know where I’m going. So to go in a straight line, to be efficient, efficiency and invention are sort of at odds because invention, real invention, not incremental improvement. And incremental improvement is so important in every endeavor and everything you do. You have to work hard on also just making things a little bit better. But I’m talking about real invention, real lateral thinking that requires wandering. And you have to give yourself permission to wander. I think a lot of people, they feel like wandering is inefficient and, you know, like when I sit down at a meeting, I don’t know how long the meeting is going to take if we’re trying to solve A problem because if I did, then I’d already, I know there’s some kind of straight line that we’re drawing to the solution. The reality is we may have to wander for a long time. And I do like group invention. I think there’s really nothing more fun than sitting at a whiteboard with a number, you know, a group of smart people and spitballing and coming up with new ideas and objections to those Ideas and then solutions to the objections and going back and forth.

When presenting a new idea, forewarn people that it may have many easy objections but encourage them to work with you to develop it further. This approach helps prevent premature rejection of potentially valuable ideas. Transcript: Speaker 1 Right. In fact, when I come up with what I think is a good idea and it survives kind of the first level of scrutiny, you know, that I do in my own head and I’m ready to tell somebody else about the idea. I will often say, look, it is going to be really easy for you to find objections to this idea. But work with me. There’s something there. There’s something there and that is intuition because it’s really easy to kill new ideas in the beginning because they do have so many, so many easy objections to them.

Rockets benefit from being larger due to physics and economies of scale. Larger rockets have less impact from parasitic mass, and components like turbopumps become more efficient as they increase in size. Transcript: Speaker 1 It sounds epic. What was it? I mean, when you look at the physics of rocket engines, and also when you look at parasitic mass, if you have, let’s say you have an avionics system, so you have a guidance and control system, That is going to be about the same mass and size for a giant rocket, as it is going to be for a tiny rocket. And so that’s just parasitic mass that is very consequential if you’re building a very small rocket, but is trivial if you’re building a very large rocket. So you have the parasitic mass thing. And then if you look at, for example, rocket engines have turbopumps. They have to pressurize the fuel and the oxidizer up to a very high pressure level in order to inject it into the thrust chamber where it burns. And those pumps, all rotating machines, in fact, get more efficient as they get larger. So really tiny turbopumps are very challenging to manufacture. And any kind of gaps are between the housing, for example, and the rotating impeller that pressurizes the fuel, there has to be some gap there. You can’t have those parts scraping against one another. And those gaps drive inefficiencies. And so if you have a very large turbopump, those gaps and percentage terms end up being very small.

Modern rocket engineering has made significant advancements in materials, such as carbon composites for fairings and friction stir welding for aluminum lithium, which provide improved structural efficiency and strength compared to traditional methods. Transcript: Speaker 1 Also because they’re humans are so small compared to it. We are building these enormous machines that are harnessing enormous amounts of chemical power, you know, in very, very compact packages. It’s truly extraordinary. Speaker 2 But then there’s all the different components and that, you know, the materials involved. Is there something interesting that you can describe about the materials that comprise the rocket so it has to be as light as possible, I guess, whilst withstanding the heat and the Harsh conditions? Speaker 1 Yeah, I play a little kind of game sometimes with other rocket people that I run into where say, what are the things that would amaze the 1960s engineers? Like what’s the change? Yeah. Because surprisingly, some of rocketry’s greatest hits have not changed. They are still, they would recognize immediately a lot of what we do today. And it’s exactly what they pioneered back in the 60s. But a few things have changed. You know, the use of carbon composites is very different today. You know, we can build very sophisticated, you saw our carbon tape laying machine that builds the giant fairings. And we can build these incredibly light, very stiff, fairing structures out of carbon composite material that they could not have dreamed of.

Rocket fairings are designed to maintain 100% integrity until they need to separate, at which point they must quickly achieve 0% integrity. This is achieved using explosive charges for a robust separation method. Transcript: Speaker 1 Yeah. Speaker 2 And what it’s supposed to do is kind of incredible because people don’t know it’s on top of the rocket. It’s going to fall apart. That’s its task, but it has to stay strong sometimes. Yes. Speaker 1 And then disappear when it needs to. That’s right. It’s a very difficult task. Yes. When you need something that needs to have 100% integrity until it needs to have zero percent integrity, it needs to stay attached until it’s ready to go away. And then when it goes away, it has to go away completely. You use explosive charges for that. And so it’s a very robust way of separating structure when you need to. Exploding. Yeah. It was a little tiny bits of explosive material. And it’ll sever the whole connection.

The biggest challenge for New Glenn is not just the first launch but achieving efficient rate production. This involves manufacturing upper stages, engines, and other components at a consistent pace to support regular launches. Transcript: Speaker 1 The first launch is one thing. And we’ll do that in 2024 coming up in this coming year. The real thing that’s the bigger challenge is making sure that our factory is efficiently manufacturing at rate. So rate production. So consider if you want to launch New Glin 24 times a year. You need to manufacture an upper stage since they’re expendable. Every twice a month you need to do one every two weeks. So you need to have all of your manufacturing facilities and processes and inspection techniques and acceptance tests and everything operating at rate. And rate manufacturing is at least as difficult as designing the vehicle in the first place in the same thing. So every upper stage has two BE3U engines. So those engines, if you’re going to launch the vehicle twice a month, you need four engines a month. So you need an engine every week. So that engine needs to be being produced at rate. And there’s all of the things that you need to do that, all the right machine tools, all the right fixtures, the right people, process, et cetera. So it’s one thing to build a first article. So that’s, you know, to launch New Glin for the first time, you need to produce a first article. But that’s not the hard part. The hard part is everything that’s going on behind the scenes to build a factory that

When making decisions, distinguish between ‘one-way door’ decisions (irreversible) and ’two-way door’ decisions (reversible). One-way door decisions should be made carefully and deliberately, while two-way door decisions can be made quickly by individuals or small teams. Transcript: Speaker 1 Up. That proves you’re a very curious explorer. All right. Back to rockets timeline. Speaker 2 You said 2024 as it stands now are both the first test launch and the launch of escapade explorers tomorrow is still possible. Speaker 1 2024. Yeah. Yeah, I think so. For sure, the first launch and then we’ll see if escapade goes on that or not. I think that the first launch for sure and I hope escapade too. Speaker 2 Hope. Speaker 1 Well, I just don’t know which mission it’s actually going to be slated on. So we also have other things that might go on that first mission. Oh, I got it. But you’re optimistic that the launches will still. Oh, the first launch. I’m very optimistic that the first launch of New Glenn will be in 2024 and I’m just not 100% certain what payload will be on that first launch. Are you nervous about it? Are you kidding? I’m extremely nervous about it. Oh, man. 100%. Every launch I go for New Shepherd for other vehicles too, I’m always nervous for these launches.

Avoid using compromise or war of attrition as resolution mechanisms for disputes. Instead, seek truth through data collection, escalation to higher management, or other more productive methods. Transcript: Speaker 1 So compromise, here’s we’re in a room here and I could say, Lex, how tall do you think this ceiling is? And you’d be like, I don’t know, Jeff. Maybe 12 feet tall. And I would say, I think it’s 11 feet tall. And then we’d say, you know what? Let’s just call it 11 and 1 half feet. That’s compromise. Instead of the right thing to do is to get a tape measure or figure out some way of actually measuring. But think getting that tape measure and figure out how to get it to the top of the ceiling and all these things, that requires energy. Compromise the advantage of compromise as a resolution mechanism is that it’s low energy. But it doesn’t lead to truth. And so in things like the height of the ceiling where truth is a noble thing, you shouldn’t allow compromise to be used when you can know the truth. Another really bad resolution mechanism that happens all the time is just who’s more stubborn. This is also, let’s say two executives who disagree. And they just have a war of attrition. And whichever one gets exhausted first, capitulates to the other one. Again, you haven’t arrived at truth. And this is very demoralizing. So this is where escalation, I try to ask people who on my team, say never get to a point where you are resolving something by who gets exhausted first. Escalate that. I’ll help you make the decision. Because that’s so de-energizing and such a terrible lousy way to make a decision. Speaker 2 Do you want to get to the resolution as quickly as possible? Because that ultimately leads to a high velocity of the system. Speaker 1 Yes. And you want to try to get as close to truth as possible. So you want like exhausting the other person is not truth seeking. And compromise is not truth seeking. So it doesn’t mean, now there are a lot of cases where no one knows the real truth. And that’s where disagreeing and commit can come in. But escalation is better than war of attrition.

Blue Origin is developing lunar landers (Mark 1 and Mark 2) that can be launched on a single New Glenn flight. They are working on solar-powered cryocoolers to make liquid hydrogen a storable propellant for deep space missions, which could be a game-changer for high-energy missions. Transcript: Speaker 1 Really much closer. I know it’s either the lander or Bill. OK. OK. I like Bill, but yeah. OK. Yes, the Mark 1 lander is designed to take 3,000 kilograms to the surface of the moon and to cargo, expendable cargo. It’s expendable lander. Lands on the moon stays there. Take 3,000 kilograms to the surface. It can be launched on a single, new-glenn flight, which is very important. So it’s a relatively simple architecture, just like the human landing system lander, they call the Mark 2. Mark 1 is also fueled with liquid hydrogen, which is for high energy missions, like landing on the surface of the moon. The high specific impulse of hydrogen is a very big advantage. The disadvantage of hydrogen has always been that it’s such a deep cryogen. It’s not storeable. So it’s constantly boiling off, and you’re losing propellant because it’s boiling off. And so what we’re doing as part of our lunar program is developing solar-powered cryocoolers that can actually make hydrogen a storeable propellant for deep space. And that’s a real game changer. It’s a game changer for any high energy mission, so to the moon, but to the outer planets, to Mars, everywhere. Speaker 2 So the idea with both Mark 1 and Mark 2 is the new glenkin. Carry it from the surface of Earth to the surface of the moon. Speaker 1 Exactly. So the Mark 1 is expendable. The lunar lander we’re developing for NASA, the Mark 2 lander, that’s part of the Artemis program. They call it the sustaining lander program. So that lander is designed to be reusable. It can land on the surface of the moon in a single stage configuration and then take off. So if you look at the Apollo program, the lunar lander in Apollo was really two stages. It would land on the surface, and then it would leave the descent stage on the surface of the moon. And only the ascent stage would go back up into lunar orbit where it would rendezvous with the command module. Here what we’re doing is we have a single stage lunar lander that carries down enough propellant so that it can bring the whole thing back up so that it can be reused over and over. And the point of

Jeff Bezos once demonstrated the importance of verifying data by calling Amazon’s customer service line during a meeting, revealing that the actual wait time was much longer than reported in their metrics. Transcript: Speaker 1 Is very early in the history of Amazon. And we were going over a weekly business review and a set of documents. And I have a saying, which is when the data and the anecdotes disagree, the anecdotes are usually right. And it doesn’t mean you just slavishly go follow the anecdotes, then it means you go examine the data. Because it’s usually not the data is being miscollected. It’s usually that you’re not measuring the right thing. And so, you know, if you have a bunch of customers complaining about something, and at the same time, you know, your metrics look like why they shouldn’t be complaining, you should doubt The metrics. And an early example of this was we had metrics that showed that our customers were waiting, I think, less than, I don’t know, 60 seconds when they called it a 1-800 number to get, you know, Phone customer service, the wait time was supposed to be less than 60 seconds. And but we had a lot of complaints that it was longer than that. And anecdotally, it seemed longer than that. Like, you know, I would call customer service myself. And so, one day we’re in a meeting or going through the WBR and the weekly business review, we get to this metric in the deck. And the guy who leads customer service is to fit in the metric. And I said, okay, let’s call. Picked up the phone. And I dialed the 1-800 number and called customer service. And we just waited in silence. What did it turn out to be? Oh, it was really long, more than 10 minutes, I think. Oh, wow. I mean, it was many minutes. And so, you know, it dramatically made the point that something was wrong with the data collection. We weren’t measuring the right thing. And that, you know, set off a whole chain of events where we started measuring it right. And that’s an example, by the way, of truth telling is like, that’s an uncomfortable thing to do.

Blue Origin is working on using lunar resources to manufacture solar cells and extract oxygen from lunar regolith. This could make long-term presence on the moon more sustainable by providing power and essential resources. Transcript: Speaker 1 Well, one of the things we’re working on is using lunar resources, like lunar regolith, to manufacture commodities and even solar cells on the surface of the moon. We’ve already built a solar cell that is completely made from lunar regolith stimulant. And this solar cell is only about 7% power efficient. So it’s very inefficient compared to the more advanced solar cells that we make here on Earth. But if you can figure out how to make a practical solar cell factory that you can land on the surface of the moon, and then the raw material for those solar cells is simply lunar regolith, Then you can just continue to churn out solar cells on the surface of the moon, have lots of power on the surface of the moon. That will make it easier for people to live on the moon. Similarly, we’re working on extracting oxygen from lunar regolith. So lunar regolith by weight has a lot of oxygen in it. It’s bound very tightly as oxides with other elements. So you have to separate the oxygen, which is very energy intensive. So that also could work together with the solar cells. But if you can, and then ultimately, we may be able to find practical quantities of ice in the permanently shadowed craters on the poles of the moon. And we know there is ice water or water ice in those craters. And we know that we can break that down with electrolysis into hydrogen and oxygen. And then you’d not only have oxygen, but you’d also have a very good high efficiency propellant fuel in hydrogen.

Jeff Bezos believes there’s room for multiple successful companies in the space industry, similar to how the internet has created many successful companies at various scales. Transcript: Speaker 1 Well, I would say, just like the internet is big and there are lots of winners at all skill levels, I mean, there are half a dozen giant companies that the internet has made, but they’re A bunch of medium-sized companies and a bunch of small companies, all successful, all with profit streams, all driving great customer experiences. That’s what we want to see in space, that kind of dynamism. And space is big. There’s room for a bunch of winners, and it’s going to happen at all skill levels. And so SpaceX is going to be successful for sure. I want Blue Origin to be successful. And I hope there are another five companies right behind us.

Jeff Bezos explains ‘Day One’ thinking as a mindset of constant renewal and rebirth, where every day is an opportunity to make new decisions and avoid being trapped by past practices or dogma. Transcript: Speaker 1 It’s really a very simple and I think age old idea about renewal and rebirth. And like every day is day one. Every day you’re deciding what you’re going to do. And you are not trapped by what you were or who you were or you need self consistency. Self consistency even can be a trap. And so day one thinking is kind of we start fresh every day and we get to make new decisions every day about invention, about customers, about how we’re going to operate. Even as deeply as what our principles are, we can go back to that. It turns out we don’t change those very often but we change them occasionally. And when we work on programs that Amazon, we often make a list of tenants and the tenants are kind of they’re not principles. They’re a little more tactical than principles but it’s kind of the main ideas that we want this program to embody whatever those are. And one of the things that we do is we put these are the tenants for this program and then we in parentheses we always put unless you know a better way. And that idea unless you know a better way is so important because you never want to get trapped by dogma. You never want to get trapped by history.

Be skeptical of proxy metrics in business. Understand why certain metrics were originally chosen and regularly reassess their relevance, as the world may have shifted, making them less valuable or even misleading. Transcript: Speaker 1 Well, you know, I’ll talk about because I think it’s the one that is maybe in some ways, the hardest to understand is the skeptical view of proxies. One of the things that happens in business, probably anything that you’re, where you’re, you know, you have an ongoing program and something is underway for a number of years, is you Develop certain things that you’re managing to. Like let’s say, the typical case would be a metric and that metric isn’t the real underlying thing. And so, you know, maybe the metric is efficiency metric around customer contacts per unit sold or something like if you sell a million units, how many customer contacts do you get or How many returns do you get and so on and so on. And so what happens is a little bit of a kind of inertia sets in where somebody a long time ago invented that metric and they invented that metric. They decided we need to watch for, you know, customer returns per unit sold as an important metric, but they had a reason why they chose that metric, the person who invented that metric And decided it was worth watching. And then fast forward five years, that metric is the proxy. The proxy for truth, I guess. The proxy for truth, the proxy for customer say, in this case, it’s a proxy for customer happiness. And, but that metric is not actually customer happiness. It’s a proxy for customer happiness. The person who invented the metric understood that connection. Five years later, a kind of inertia can set in and you forget the truth behind why you were watching that metric in the first place and the world shifts a little. And now that proxy isn’t as valuable as it used to be or it’s missing something and you have to be on alert for that. You have to know, okay, this is, I don’t really care about this metric. I care about customer happiness. And this metric is worth putting energy into and following and improving and scrutinizing only in so much as it actually affects customer happiness. And so you got a constantly beyond guard and it’s very, very common. This is a nuanced problem. It’s very common, especially in large companies, that they are managing to metrics, that they don’t really understand. They don’t really know why they exist. And the world may have shifted off from under them a little. And the metrics are no longer as relevant as they were when somebody 10 years earlier invented the metric.

In meetings, have the most senior person speak last to avoid influencing others’ opinions prematurely. This allows for more unfiltered and honest input from all participants. Transcript: Speaker 1 And I know from experience that if I speak first, even very strong willed, highly intelligent, high judgment participants in that meeting will wonder, well, if Jeff thinks that I came In this meeting thinking one thing, but maybe I’m not right. And so you can do little things like if you’re the most senior person in the room, go last. But everybody else go first. In fact, ideally, let’s try to have the most junior person go first and the second, then try to go in order of seniority so that you can hear everyone’s opinion in a kind of unfiltered way.

Jeff Bezos views large language models and AI as powerful discoveries rather than inventions. He is optimistic about their potential to help humanity, including possibly preventing self-destruction, while acknowledging the risks and challenges they present. Transcript: Speaker 1 You’re talking about generative AI, large language models, things like Chad GPT, and it’s soon successors. And these are incredibly powerful technologies to believe otherwise is to bear your head in the sand, soon to be even more powerful. It’s interesting to me that large language models in their current form are not inventions, they’re discoveries. The telescope was an invention. But looking through it at Jupiter, knowing that it had moons was a discovery. Like my god, it has moons. And that’s what Galileo did. And so this is closer on that spectrum of invention. We know exactly what happens with a 787. It’s an engineered object. We designed it. We know how it behaves. We don’t want any surprises. Large language models are much more like discoveries. We’re constantly getting surprised by their capabilities. They’re not really engineered objects. Then you have this debate about whether they’re going to be good for humanity or bad for humanity. Even specialized AI can be very bad for humanity. Just regular machine learning models can make certain weapons of war that could be incredibly destructive and very powerful. They’re not general AIs. They’re just very smart weapons. And so we have to think about all of those things. I’m very optimistic about this. So even in the face of all this uncertainty, my own view is that these powerful tools are much more likely to help us and save us even than they are to unbalance, hurt us, and destroy us. I think we humans have a lot of ways of we can make ourselves go extinct. These things may help us not do that. So they may actually save us. So the people who are overly concerned, in my view, over the course of this is a valid debate. I think that they may be missing part of the equation, which is how helpful they could be in making sure we don’t destroy ourselves.

Jeff Bezos views large language models and AI as powerful discoveries rather than inventions. He is optimistic about their potential to help humanity, including possibly preventing self-destruction, while acknowledging the risks and challenges they present. Transcript: Speaker 1 You’re talking about generative AI, large language models, things like Chad GPT, and it’s soon successors. And these are incredibly powerful technologies to believe otherwise is to bear your head in the sand, soon to be even more powerful. It’s interesting to me that large language models in their current form are not inventions, they’re discoveries. The telescope was an invention. But looking through it at Jupiter, knowing that it had moons was a discovery. Like my god, it has moons. And that’s what Galileo did. And so this is closer on that spectrum of invention. We know exactly what happens with a 787. It’s an engineered object. We designed it. We know how it behaves. We don’t want any surprises. Large language models are much more like discoveries. We’re constantly getting surprised by their capabilities. They’re not really engineered objects. Then you have this debate about whether they’re going to be good for humanity or bad for humanity. Even specialized AI can be very bad for humanity. Just regular machine learning models can make certain weapons of war that could be incredibly destructive and very powerful. They’re not general AIs. They’re just very smart weapons. And so we have to think about all of those things. I’m very optimistic about this. So even in the face of all this uncertainty, my own view is that these powerful tools are much more likely to help us and save us even than they are to unbalance, hurt us, and destroy us. I think we humans have a lot of ways of we can make ourselves go extinct. These things may help us not do that. So they may actually save us. So the people who are overly concerned, in my view, over the course of this is a valid debate. I think that they may be missing part of the equation, which is how helpful they could be in making sure we don’t destroy ourselves.