Conversation with Elon Musk | World Economic Forum Annual Meeting 2026 (YouTube Video Transcript)

(00:00:00) Your YouTube transcript will appear here (00:00:00) Heat. Heat. (00:00:20) Heat. Heat. (00:00:27) [music] (00:00:45) [music] (00:00:47) Heat. Heat. (00:01:08) Heat. Heat. [music] (00:01:28) Heat. Heat. (00:01:32) [music] (00:01:48) Heat. (00:02:10) Heat. (00:02:25) Heat. Heat. (00:02:35) >> [music] (00:02:44) [music] (00:02:54) >> Yeah. (00:02:56) Heat. (00:03:06) >> [music] (00:03:18) [applause] (00:03:28) >> That was not a That was Not a large (00:03:32) applause. Start again. (00:03:34) [applause and cheering] (00:03:40) >> That's better. Thank you. (00:03:42) >> Yeah, we're going to make this (00:03:43) interesting. (00:03:45) [laughter] (00:03:46) How many how many quotes are you going (00:03:47) to want that are going to be after this (00:03:50) session? [snorts] (00:03:52) >> Uh I don't know. I mean, (00:03:54) five. (00:03:55) >> Okay. [laughter] (00:03:57) So, uh good afternoon everybody. It's (00:03:59) great to see everybody here. Uh it's (00:04:00) been an amazing week here in Davos. Um (00:04:04) hopefully everybody (00:04:06) saw that we are having conversations (00:04:09) here. Hopefully everybody agrees. There (00:04:12) [snorts] are some conversations that we (00:04:13) may disagreed. There's many (00:04:15) conversations we may have agreed but (00:04:17) through those conversations and I think (00:04:19) today's result with a peace agreement (00:04:21) earlier today (00:04:23) um the world economic forum is here to (00:04:26) have those conversations to have (00:04:27) understandings and also resolution. So (00:04:31) um uh it's an important component of who (00:04:35) we and what we are and I'm thrilled uh (00:04:38) to have Elon Musk here. Um (00:04:41) he came all the way from California to (00:04:43) be here to see all of you. So uh thank (00:04:46) you Elon. (00:04:46) >> Uh you're most welcome. (00:04:49) [applause] (00:04:50) >> Um (00:04:51) >> I mean I heard I heard about heard about (00:04:53) the formation of the the peace summit (00:04:56) and I was like is that uh is that PIC? (00:05:00) Uh (00:05:02) [laughter] (00:05:04) you know little piece of Greenland, a (00:05:06) little piece of Venezuela. [laughter] (00:05:09) We we got one. (00:05:11) >> All we want is peace. [laughter] (00:05:12) >> Okay. I'm going to uh as I said, I'm a (00:05:14) pretty proud uh CEO of Black Rockck (00:05:16) since we went public. Um uh the (00:05:19) compounding return of Black Rockck to (00:05:21) our shareholders with 21%. (00:05:23) Uh since Elon took Tesla uh public, his (00:05:28) compounded return (00:05:30) is 43%. (00:05:34) This is I just another advertisement for (00:05:36) everybody, especially for Europeans. (00:05:38) This is why more citizens should be (00:05:42) investing with growth, investing with (00:05:45) their countries. Imagine if a lot of (00:05:47) pension funds invested with Elon when (00:05:51) Tesla went public and how much (00:05:55) return with the all the pension funds (00:05:58) that invested side by side with Elon and (00:06:01) the growth. So, um, a spectacular (00:06:03) return. There's very few companies, (00:06:06) well, I don't think there's any other (00:06:08) company as large as Tesla today that has (00:06:10) that compounded return. So, (00:06:11) congratulations. Thank you. It's a good (00:06:13) measurement. (00:06:14) >> Well, we have an incredible team at (00:06:15) Tesla and that's the reason. (00:06:17) >> So, I want to get into uh the dirt, the (00:06:20) the meaningful component about (00:06:22) technology, the possibilities. Um I want (00:06:26) to talk about AI and robotics, energy, (00:06:29) space and the progress ultimately coming (00:06:31) down to engineering engineering (00:06:33) discipline scale execution. (00:06:36) Um and few few people if not anyone has (00:06:40) the experience and the fortitude to (00:06:43) confront these issues headon. Not just (00:06:46) the ideas but the execution across so (00:06:49) many different technologies. Elon and (00:06:51) that's why I thought it was important (00:06:53) for us to have this dialogue here uh in (00:06:55) Davos. So [clears throat] you're (00:06:58) presently building on AI, on robotics, (00:07:01) on space, on energy all at the same (00:07:04) time. When you look across those (00:07:07) efforts, (00:07:08) what do they have in common from an (00:07:10) engineering standpoint? (00:07:12) >> Uh well, they're all very difficult (00:07:14) technology challenges. Um but the (00:07:18) uh the overall goal of my companies is (00:07:20) to (00:07:22) maximize the future of civilization like (00:07:25) basically maximize the probability that (00:07:27) civilization has a great future um and (00:07:31) uh to [clears throat] (00:07:32) expand consciousness beyond earth. So (00:07:35) take SpaceX for example that SpaceX is (00:07:38) about build advancing rocket technology (00:07:40) to the point where we can extend life (00:07:42) and consciousness beyond earth. uh to (00:07:45) the moon, to Mars, uh eventually to (00:07:48) other star systems. And uh and I I think (00:07:51) we should always view consciousness uh (00:07:54) life as we know it as as precarious and (00:07:57) delicate. Um because to the best of our (00:08:01) knowledge, we we we don't know of life (00:08:03) anywhere else. You know, I'm often asked (00:08:06) um are there aliens among us? And I'll (00:08:10) say that I am one, but (00:08:12) >> or you're from the future. They don't (00:08:14) believe me. (00:08:15) >> Okay. (00:08:15) >> Um, so, uh, but I I I [clears throat] (00:08:20) think if anyone would know if there are (00:08:21) aliens among us, it would be me. Um, (00:08:24) and, uh, we we have 9,000 satellites up (00:08:28) there, and not once have we had to (00:08:30) maneuver around an an alien spaceship. (00:08:33) So, I'm like, I don't know. It's bottom (00:08:36) line is I think we need to assume that (00:08:39) life and consciousness is extremely rare (00:08:41) and it might only be us. And if that's (00:08:45) the case, then we need to do everything (00:08:46) possible to to ensure that the the light (00:08:51) of can the light the light of (00:08:52) consciousness is not extinguished (00:08:55) because we're effectively the way I view (00:08:58) it is the image in my mind is of a a (00:09:01) tiny candle in a vast darkness tiny (00:09:05) candle of consciousness that could (00:09:07) easily go out. Um and that's why it's (00:09:10) important to make life multilanetary (00:09:12) uh such that if there is a natural (00:09:14) disaster or a man-made disaster on earth (00:09:17) that consciousness continues. That's the (00:09:18) purpose of SpaceX. Um (00:09:22) te Tesla is obviously about sustainable (00:09:24) technology and uh and and also at this (00:09:29) point we've we've sort of added to our (00:09:31) mission sustainable abundance. (00:09:34) So with robotics and AI (00:09:38) um this this is really the path to (00:09:42) abundance for all. If you say you know (00:09:44) people often talk about (00:09:46) uh solving global poverty or essentially (00:09:50) how how do we make give everyone a very (00:09:54) high standard of living? I I think the (00:09:56) only way to do this is AI and robotics. (00:10:00) um which which doesn't mean that it is (00:10:02) uh (00:10:04) without its issues. I mean there we need (00:10:06) to be very careful with AI. We need to (00:10:08) be very careful with robotics. We don't (00:10:11) want to find ourselves in a James (00:10:13) Cameron movie. Uh (00:10:16) you know Terminator (00:10:18) [laughter] (00:10:18) >> he's he's great movies. Love his movies (00:10:20) but but we don't want to be in (00:10:22) Terminator obviously. Um but but if you (00:10:25) have um ubiquitous AI that is (00:10:29) essentially free or close to it and (00:10:31) ubiquitous robotics uh then uh you will (00:10:35) have an an an (00:10:38) explosion in the in the global economy (00:10:39) an expansion in the global economy that (00:10:41) is truly (00:10:42) beyond all precedent. (00:10:44) >> Elen can that expansion be broad? Yes. (00:10:48) >> Or is it narrow and how can that be (00:10:51) created? How can it broaden the global (00:10:53) economy? (00:10:55) >> Yeah, it's (00:10:58) I mean I mean a way to think of it is (00:11:00) that if you have a large number of (00:11:02) humanoid robots um the economic output (00:11:05) is (00:11:07) the average productivity per robot times (00:11:10) the number of robots, (00:11:12) >> right? Um and and actually my prediction (00:11:16) is in the in the benign scenario of the (00:11:19) future that we will the robots we will (00:11:21) actually make so many robots in AI that (00:11:23) they will actually saturate all human (00:11:26) needs. Meaning you won't be able to even (00:11:28) think of something (00:11:30) to ask the robot for at a certain point (00:11:33) like like there will be such an (00:11:36) abundance of goods and services (00:11:38) because the [clears throat] (00:11:40) my prediction is there'll be there'll be (00:11:42) more robots than people. So (00:11:44) >> but how do you then have human purpose (00:11:47) in that scenario? Yeah, I mean, you (00:11:50) know, there nothing's perfect, you know, (00:11:52) but [laughter] (00:11:53) um (00:11:55) but I mean I mean it is it is a a (00:11:58) necessary um (00:12:01) like you can't have both. You can't have (00:12:03) work that has to be done uh um and uh (00:12:07) amazing abundance for all. (00:12:09) Um because if it's if it's work that has (00:12:11) to be done then then then you and and (00:12:13) only some people can do it then then you (00:12:16) you you can't have abundance (00:12:17) >> then it's narrow. (00:12:18) >> It's narrow. Exactly. (00:12:20) >> So (00:12:21) >> um but if if you if you have billions of (00:12:24) humanoid robots and I think there will (00:12:25) be um then I think I think everyone on (00:12:29) earth is going to have one and going to (00:12:30) want one. Um because uh you would who (00:12:34) wouldn't want a robot to you know um (00:12:38) assuming it's very safe uh watch over (00:12:40) your kids, take care of your pet. Uh if (00:12:43) you have elderly parents uh a lot of (00:12:46) friends of mine said they have elderly (00:12:48) parents it's it's uh very difficult to (00:12:49) take care of them. (00:12:50) >> Expensive. (00:12:51) >> Yeah, it's expensive and it's expensive (00:12:53) and there just aren't enough people to (00:12:55) take care of the there aren't enough (00:12:56) young people to take care of the old (00:12:57) people, (00:12:58) >> right? (00:12:58) >> Um so if you if they Um, if if you had a (00:13:02) robot that could take care of and and (00:13:05) protect an an elderly parent, I think (00:13:08) that would be great. That would be an (00:13:09) amazing thing to have. Um, and and that (00:13:12) I think we will have those things. So, I (00:13:15) mean, overall, I'm I'm very optimistic (00:13:17) about the future. I think we're headed (00:13:18) for a future of amazing abundance, uh, (00:13:22) which is very cool. Um, and, uh, and and (00:13:26) definitely [clears throat] we are in the (00:13:28) most interesting time in history. (00:13:30) Um, I don't think there's more a more (00:13:33) interesting time in history. (00:13:36) >> Can we uh can you and I reverse aging in (00:13:38) this new history or or or are we going (00:13:41) to see it? (00:13:44) >> You know, I haven't I haven't put much (00:13:45) time into uh the aging stuff. I I do (00:13:49) think it is a very solvable problem like (00:13:51) like you can (00:13:53) I think when when when we figure out (00:13:56) what causes aging, (00:13:57) >> I think we'll find it's incredibly (00:13:59) obvious. It's not a subtle thing. Um, (00:14:02) the reason I say it's not a subtle thing (00:14:03) is because all the cells in your body, (00:14:06) you know, with some pretty much age at (00:14:08) the same rate. I've never seen someone (00:14:11) with with an old left arm and a young (00:14:14) right arm ever in my life. (00:14:16) >> Um, so why is that? It that means that (00:14:19) there must be a clock, a synchronizing (00:14:20) clock, (00:14:21) >> right, (00:14:21) >> that is synchronizing across 35 trillion (00:14:24) cells in your body. Um and uh (00:14:30) you know the but there is some benefit (00:14:34) to death by the way. [laughter] It's (00:14:36) like there's there's a reason why we (00:14:38) don't actually have a longer lifespan. (00:14:41) Uh because if you if you have if people (00:14:43) do live forever for very long time I (00:14:45) think there's some risk of an oification (00:14:48) of society of of things just getting (00:14:51) kind of locked in place. Um and uh (00:14:57) you know it it just may become um (00:15:02) stalifying just not uh lack lack (00:15:05) vibrancy. Um but but that that said, do (00:15:08) I think we'll figure out ways to extend (00:15:12) life and um and maybe even reverse (00:15:15) aging? I think that's highly likely. (00:15:19) >> I'm looking forward to that. (00:15:21) >> Yeah. Uh so um in the future that you (00:15:25) talk about AI models, autonomous (00:15:28) machines, rockets depends on massive (00:15:31) increases of compute, massive increases (00:15:34) in energy, expensive energy, (00:15:37) manufacturing scale, (00:15:40) what are the bottlenecks to to get (00:15:42) there? And then once again with all that (00:15:45) expenditures again, how can we make sure (00:15:48) that it's broadened, not narrow? (00:15:54) Um [clears throat] (00:15:56) I I just think the natural thing is it's (00:15:58) going to be very broad because uh AI (00:16:01) companies will seek as many customers as (00:16:03) they possibly can and the cost of AI (00:16:06) will get is already very low and it's (00:16:08) it's plummeting every year. I mean you (00:16:11) almost the cost of AI is almost ch (00:16:14) meaningfully changing on a (00:16:15) month-to-month basis. (00:16:16) >> There's open there's open models now (00:16:18) everywhere. Yeah. (00:16:19) >> Yes. very there's open models um and the (00:16:22) open models only lag they're maybe a (00:16:25) year behind right the the private the (00:16:27) sort of closed models (00:16:29) >> um so so I I think the the the ai (00:16:34) companies will seek as as many customers (00:16:36) as possible which means they'll seek (00:16:38) they'll provide AI to the world (00:16:40) >> but the cost of getting to there the (00:16:42) compute the chips (00:16:44) um the fab um the powering (00:16:50) that to me what are the what are the you (00:16:52) know those are huge (00:16:53) >> the limiting factor yeah I think the (00:16:54) limiting factor for um AI deployment is (00:16:58) fundamentally electrical power (00:17:00) >> it's just [clears throat] right it's (00:17:01) energy (00:17:02) >> yeah yeah um I mean we're seeing the the (00:17:06) rate of AI chip production increase (00:17:08) exponentially but the rate of (00:17:10) electricity being brought online is uh (00:17:13) >> 3% 4% a year max (00:17:15) >> yeah it's clear that we're we're We're (00:17:18) we're very soon, maybe even later this (00:17:20) year, uh we'll be producing more chips (00:17:22) than we can turn on (00:17:25) except for China. China China is China's (00:17:28) growth in electricity is is tremendous. (00:17:30) >> They're building 100 gawatts of nuclear (00:17:32) as we speak. (00:17:33) >> Uh actually solar is the biggest thing (00:17:35) in China. So China's I believe China's (00:17:38) production capacity on solar is 1500 (00:17:40) gawatts a year. Um and they're deploying (00:17:43) over 1,000 gawatts a year of of solar. (00:17:47) Um now you know for continuous solar (00:17:51) load you divide that by roughly I don't (00:17:53) know four or five uh call it that's (00:17:57) around uh 250 gawatts of steady state (00:18:00) power um paired with batteries (00:18:04) um and that that's a very big number (00:18:07) that's half of the average power usage (00:18:10) in the US (00:18:10) >> right (00:18:11) >> so US US power uh usage on average is is (00:18:14) 500 gawatt uh China just in solar just (00:18:18) like just in in in in solar like that (00:18:21) can provide steady state power uh and (00:18:25) batteries can do half of the US (00:18:27) electricity output per year just with (00:18:29) solar. Solar is by far the the the (00:18:33) biggest source of of of energy. Um and (00:18:36) actually when you look beyond or even (00:18:38) even on Earth but certainly beyond Earth (00:18:42) u the sun rounds up to 100% of all (00:18:45) energy. This is an important thing to (00:18:48) consider. Um so this the sun is 99.8% (00:18:53) of the mass of the solar system. Jupiter (00:18:56) is about.1% and everything else is (00:18:58) miscellaneous. Um now even if you were (00:19:01) to uh burn Jupiter in a in a (00:19:05) thermonuclear reactor uh the sun the (00:19:09) amount of energy produced by the sun (00:19:10) would still round up to 100% because (00:19:13) Jupiter is only.1%. If you teleported (00:19:16) teleported three more Jupiters into our (00:19:18) solar system, the and and burnt three (00:19:21) more Jupiters and everything else in the (00:19:24) solar system, the sun's (00:19:27) energy would still round up to 100%. (00:19:31) So, it's really all about the sun. Um (00:19:34) and that's that's why (00:19:36) uh one of the things we'll be doing with (00:19:37) SpaceX uh in within a few years is (00:19:40) launching um solar powered AI (00:19:43) satellites, (00:19:44) >> right? Um because the space is really (00:19:48) the source of immense power and then you (00:19:51) don't need to take up any room on earth. (00:19:53) uh there's so much room in space and you (00:19:56) can scale to uh enormous uh I mean you (00:20:00) can you can scale to (00:20:02) I think ultimately hundreds hundreds of (00:20:05) terowatts a year. (00:20:10) You and I have had these conversations (00:20:11) before, but why don't you tell the (00:20:14) audience, what would it take for the (00:20:15) United States and what type of geography (00:20:18) would it take to have that solar field (00:20:21) to electrify the United States? And and (00:20:23) then let me ask a question. Why aren't (00:20:24) we doing it? (00:20:26) >> Yeah. So I mean I guess rough way to (00:20:29) think about it is um 100 miles by 100 (00:20:32) miles we'll call it I60 kilometers by (00:20:35) 160 kilometers of of [clears throat] (00:20:37) solar is enough to power the entire (00:20:40) United States. (00:20:42) So you 100 mile by 100 mile area is is I (00:20:46) mean you could take basically a small (00:20:48) corner of Utah, (00:20:49) >> Nevada, (00:20:50) >> Nevada, New Mexico. [clears throat] (00:20:52) Obviously, we wouldn't want it all in (00:20:53) one place, but you can it's it's it is a (00:20:56) very small percentage of the area of of (00:20:59) the US to generate all of the (00:21:02) electricity that the US uses. Um, and (00:21:04) the same is true actually, I mean, for (00:21:06) for Europe. You you could take a small (00:21:08) part you could take uh relatively (00:21:12) unpopulated areas of say Spain and (00:21:14) Sicily and generate all of the (00:21:16) electricity power that Europe needs. So (00:21:18) why don't you think that there's a (00:21:20) movement towards that here and in the (00:21:22) United States? (00:21:24) >> Uh well there is (00:21:25) >> as it is in China. (00:21:27) >> Well unfortunately in in the US the the (00:21:30) the tariff barriers for solar panel are (00:21:33) are extremely high. [clears throat] Um (00:21:35) and that makes the economics of (00:21:37) deploying solar uh (00:21:40) so artificially high because China makes (00:21:43) almost all the solar um and uh and the (00:21:46) that (00:21:46) >> what would it take for Europe or the US (00:21:48) to build it commercially if it's that (00:21:51) scale? (00:21:53) Yeah, I I think I think uh well, I can (00:21:57) tell you what what we're going to do, (00:21:59) you know, at SpaceX and Tesla was is (00:22:00) we're we're building up um large scale (00:22:03) solar, (00:22:04) >> right? (00:22:04) >> So, the the SpaceX and Tesla teams both (00:22:07) separately are working to build to 100 (00:22:09) gawatts a year of solar power in the US (00:22:12) of manufactured solar power. Um and um (00:22:17) that'll probably take us I don't know (00:22:19) about three years or something. (00:22:21) >> But that's these are pretty big numbers. (00:22:23) Mhm. (00:22:23) >> Um and um you know I'd encourage others (00:22:27) to the to do the same. Um (00:22:30) we obviously don't control the you know (00:22:32) US tariff policy. uh but uh for for for (00:22:38) other countries uh I would re you know (00:22:41) that there's China makes solo cells that (00:22:44) are incredibly low cost and I think uh (00:22:47) it would be worth uh doing large scale (00:22:49) solo. (00:22:54) So, [clears throat] (00:22:58) I know you are you're going to be having (00:23:00) a couple big announcements on robotics (00:23:02) and what it can do. I mean, when I went (00:23:05) to the factory, you showed me those (00:23:06) robots. (00:23:07) >> Yeah. (00:23:07) >> Um, (00:23:11) how quickly you talked about the (00:23:13) billions of robots, but how quickly and (00:23:15) how quickly can they be deployed in a (00:23:17) manufacturing setting? How quickly can (00:23:19) they be utilized and be functional and (00:23:22) be uh create that that abundance that (00:23:25) you talked about? (00:23:27) [clears throat and cough] (00:23:29) >> Well, (00:23:31) humanoid robotics will advance very (00:23:33) quickly. I think uh we we we do have (00:23:37) some of the Tesla Optimus robots doing (00:23:40) simple tasks in the factory. (00:23:43) um except probably later this year by (00:23:47) the end of this year I think they'll be (00:23:48) doing um more more complex tasks um and (00:23:53) and but still deployed in an industrial (00:23:56) environment [clears throat] and uh and (00:23:58) probably sometime next year I'd say that (00:24:02) by the by the end of next year I I think (00:24:05) uh we'd be selling humanoid robots uh to (00:24:09) the public. um that that's when we are (00:24:12) confident that the it's very high (00:24:14) reliability, very high safety um and the (00:24:17) range of functionality is uh is also (00:24:20) very high. You can basically ask it to (00:24:21) do anything you'd like. (00:24:23) >> You're [clears throat] already seeing (00:24:24) that in Tesla cars. This the software (00:24:26) changes that you're doing and what is it (00:24:28) every quarter now a software change that (00:24:30) upgrades the the ability of the robot (00:24:33) within the car. (00:24:35) >> Uh yes, the Tesla full self-driving (00:24:37) software we update it sometimes once a (00:24:39) week. Um and um [clears throat] recently (00:24:43) some of the insurance companies have (00:24:44) said that uh it is actually so safe uh (00:24:47) when where Tesla full full self-driving (00:24:50) is so safe that uh they're they're (00:24:53) offering uh customers half price (00:24:56) insurance if they if they use Tesla full (00:24:59) self-driving in their car (00:25:00) >> and that could be monitored by the (00:25:02) insurance company. Can they is that part (00:25:03) of the agreement then? (00:25:04) >> Yeah. (00:25:07) Um but I I think uh self-driving (00:25:11) [clears throat] cars is essentially a (00:25:14) solved problem at this point, (00:25:16) >> right? Um (00:25:18) and te Tesla's r rolled out uh sort of (00:25:21) robo taxi service in a few cities and (00:25:23) >> right (00:25:24) >> will uh I be very very widespread by the (00:25:28) end of this year within the US and then (00:25:31) we we hope to get supervised full (00:25:33) self-driving approval in Europe (00:25:36) hopefully next month. (00:25:37) >> Really that quickly? (00:25:38) >> Yeah. And then [clears throat] uh maybe (00:25:41) a similar timing timing for China (00:25:43) hopefully. I want to move to space (00:25:45) because historically space is very (00:25:48) capital intensive. It historically been (00:25:50) done by governments. Obviously SpaceX (00:25:53) changed the whole model. Uh but we've (00:25:56) seen it slow to scale and now I'm (00:25:58) starting to see it ramping up in what (00:26:00) you're doing and other things. Um (00:26:03) talk to us about the resil you know the (00:26:05) automation and AI how it's changing the (00:26:07) economics in building uh and preparing (00:26:10) for us in operating in space. (00:26:13) Uh sure. Um (00:26:18) [clears throat] well the key (00:26:19) breakthrough that tells that that's the (00:26:21) the major breakthrough that SpaceX is (00:26:23) hoping to achieve this year is full (00:26:25) reusability. [clears throat] (00:26:27) Um so no one has ever achieved full (00:26:30) reusability of a rocket which is very (00:26:32) important for the cost of access to (00:26:34) space. Um we've achieved partial (00:26:36) reusability with Falcon 9 by landing the (00:26:38) boost stage. We've we've now landed the (00:26:40) boost stage over 500 times. Um but uh we (00:26:44) we have to throw away the upper stage. (00:26:47) The upper stage burns up on re-entry for (00:26:49) Falcon 9. So and that that the cost of (00:26:51) that is equivalent to a small to (00:26:53) mediumsiz jet. So um but with with (00:26:58) Starship, which is a giant rocket, it's (00:27:00) it's the largest flying machine ever (00:27:02) made. (00:27:02) >> That's a rocket that you using for the (00:27:04) idea of going to Mars, right? (00:27:06) >> Yeah. Mars and the moon. um as well as (00:27:08) for uh high volume satellite stuff. So (00:27:12) Starship um hopefully this year we (00:27:15) should prove full reusability for (00:27:18) Starship which will be um a a profound (00:27:23) invention uh because (00:27:26) uh the cost of access to space will drop (00:27:29) by a factor of 100 (00:27:31) when you achieve full reusability. Um, (00:27:34) >> right. (00:27:35) >> It's it's the same sort of economic (00:27:36) difference that you would expect that uh (00:27:38) between say a reusable aircraft and a (00:27:41) non-reusable aircraft. Like if you have (00:27:43) to throw your aircraft away after every (00:27:46) flight, that would be a very expensive (00:27:47) flight. Um, but if you only have to (00:27:50) refuel uh then it's the cost of the (00:27:53) fuel. And so that's really the uh (00:27:58) the the fundamental breakthrough that (00:28:00) gets the cost of access to space uh we (00:28:04) think uh below the cost of uh of freight (00:28:09) on aircraft. (00:28:11) Uh so (00:28:13) you know under under $100 a pound type (00:28:16) of thing easily. Um so it it it makes uh (00:28:20) [clears throat and cough] (00:28:21) putting large satellites into into space (00:28:25) very low very very cheap. Um and then (00:28:28) when you have solar in space you you get (00:28:31) uh five times more effectiveness maybe (00:28:33) even more than that than solar on the (00:28:35) ground because it's it's always sunny. (00:28:37) >> It's cold and (00:28:39) >> it yeah it's it's it's always well it's (00:28:41) always sunny so you don't have a dayight (00:28:43) cycle or seasonality (00:28:44) >> or weather. Um and you get about uh 30% (00:28:48) more power in space uh because uh you (00:28:51) don't have atmospheric attenuation of (00:28:53) the power. (00:28:54) >> Right? (00:28:54) >> The the net effect is solar is five (00:28:56) times more uh any given solar panel will (00:29:01) do five times more uh energy in space (00:29:04) than uh on the ground. (00:29:06) >> Is there any capacity in doing that and (00:29:08) and then taking that power and bringing (00:29:10) it back to earth? Is there any way of (00:29:12) doing that or or you're just taking that (00:29:14) power and utilizing it for the needs (00:29:16) like building um u AI data centers in (00:29:21) space? I I I think the the case it's (00:29:24) it's a no-brainer for building uh AI (00:29:27) solar powered AI data centers in space. (00:29:29) Um because as you mentioned it's it's (00:29:31) also very cold in space. If you're if (00:29:32) you're if you're in the shadow uh then (00:29:35) it's it's very cold in space just 3° (00:29:37) Kelvin. So you just have you have solar (00:29:39) panels facing the sun and then uh a uh (00:29:43) radiator that's like point like pointed (00:29:46) away from the sun. Um so it has no sun (00:29:48) incident and then it's and then it's (00:29:49) just cooling. It's a very efficient (00:29:51) cooling system. So (00:29:54) uh net net effect is that the lowest (00:29:57) cost place to put AI will be space and (00:30:00) that and that'll be true within two (00:30:03) years maybe three three at the latest. (00:30:06) >> Wow. [clears throat] So looking 10 or 20 (00:30:08) years out um what would how would you (00:30:11) describe success with AI or space (00:30:14) technology and where do you see it is (00:30:16) that can you are you more certain what's (00:30:19) going to happen in the next three years (00:30:20) or or five or 10 (00:30:24) >> I don't know what's going to happen in (00:30:25) 10 years but the rate at which AI is (00:30:28) progressing (00:30:30) I think we [clears throat] (00:30:32) we're we're (00:30:35) AI that is smarter than any human by the (00:30:37) end of this year. Um and and I would say (00:30:39) no later than next year. (00:30:40) >> Wow. (00:30:42) >> Um and then probably by 2030 or 2031, (00:30:46) call it 5 years from now, (00:30:49) uh AI will be smarter than (00:30:50) [clears throat] (00:30:51) uh all of humanity collectively. (00:30:55) >> We only have a number of minutes left, (00:30:57) but I want I want to humanize you for a (00:30:59) second (00:31:00) >> so there's no speculation that you're (00:31:02) talking about peace. (00:31:03) >> Right. Right. I want to [laughter] (00:31:06) I mean I would frame this question by (00:31:08) you are the most successful entrepreneur (00:31:10) industrialist in the 21st century maybe (00:31:13) beyond. Um I want to so I want to really (00:31:15) get this you know what inspired you who (00:31:18) inspired you (00:31:20) what was the foundation of of your (00:31:22) curiosity (00:31:24) and and importantly what was the what (00:31:27) was it was there a aha moment epiphany (00:31:30) at any time in your life and career? (00:31:35) Well, um, (00:31:37) I mean, as a kid, I read a lot of (00:31:38) science fiction, sci-fi, fantasy books. (00:31:41) >> Yep. We talked about (00:31:43) >> and, uh, comic books. Uh, and I always (00:31:46) like technology. Uh, I didn't expect to (00:31:48) be where I am today. Seems incredibly (00:31:51) implausible. Um, but, uh, yeah, I was I (00:31:55) was inspired by reading about books (00:31:57) about the future, about science fiction. (00:31:58) And uh and I guess I want to make (00:32:02) science fiction not fiction forever at (00:32:04) some point turn science fiction to (00:32:06) science fact. Um, and uh, (00:32:10) you know, we want to have like Starfleet (00:32:14) and Star Star Trek really for for real, (00:32:16) like where we actually have giant (00:32:19) spaceships traveling through space, (00:32:21) going to other planets, traveling to (00:32:23) other star systems. (00:32:25) >> Be beamed up to go back to New York. (00:32:27) [laughter] You know, I I'd like to just (00:32:29) be beamed back to New York instead of (00:32:31) flying. (00:32:34) >> Yeah. (00:32:35) um (00:32:37) >> talk about Star Trek. (00:32:38) >> No, I guess my my central I what I would (00:32:41) call the phil philosoph philosophy of (00:32:43) curiosity. Um I'd like to understand um (00:32:46) the meaning of life, you know, the is (00:32:50) the standard model of is a standard (00:32:51) model of physics correct (00:32:53) >> regarding the beginning of life, (00:32:54) beginning of existence and the end of (00:32:56) the universe. Um what what questions do (00:32:59) we not know to ask that we should ask? (00:33:02) Um, and AI will help us with these (00:33:04) things. Um, so I'm just trying to (00:33:07) understand how do we get here? What's (00:33:09) going on? What's real? Are there aliens? (00:33:11) Maybe they are. Um, and if if we've got (00:33:14) if we got spaceships that are traveling (00:33:15) to other star systems, uh, we may find (00:33:18) we may encounter aliens and or we may (00:33:20) find many long dead alien civilizations. (00:33:24) But I I I'm just I just I just want to (00:33:26) know what's going on. kind of curious (00:33:27) about the the universe and (00:33:30) um that's my philosophy. (00:33:35) >> Do you see yourself ever going to Mars (00:33:37) in your lifetime? (00:33:40) >> Uh yeah, I mean I would say like I you (00:33:41) know I (00:33:42) >> that's a long commitment. (00:33:43) >> I've been asked (00:33:44) >> isn't that three years each way? (00:33:46) >> Uh it's six months. (00:33:48) >> Six months. That's all it is? (00:33:49) >> Yeah, six months. But the planets only (00:33:51) align every every two years. (00:33:53) >> Okay. So, [clears throat] uh, yeah, I've (00:33:56) been asked a few times like, "Do I want (00:33:58) to, you know, die on Mars?" And I'm (00:34:00) like, "Yes, but just not on impact." (00:34:04) [laughter] (00:34:08) That's a good That's a good answer. (00:34:09) Anyway, uh, we're out of time. I (00:34:11) hopefully everybody enjoyed this. Um, I (00:34:15) mean, there's so many myths around Elon (00:34:17) Musk. I could tell you he's a great (00:34:19) friend and I constantly learn so much (00:34:22) from him. Um, and I'm totally inspired (00:34:26) by what he's what he has done. (00:34:30) I've been inspired who he is and I'm (00:34:34) totally inspired by his vision of the (00:34:36) future and I don't think it's such a bad (00:34:39) future and I agree with his optimism. So (00:34:41) Elon, thank you. Any last words? Um (00:34:46) >> um well I think generally I think my (00:34:49) last words would be I would encourage (00:34:50) everyone to be optimistic and excited (00:34:53) about the future. (00:34:54) >> Good. (00:34:55) >> Um and and and generally I think for (00:34:57) quality of life it is actually better to (00:35:00) on the side of being an optimist and (00:35:03) wrong rather than a pessimist and right. (00:35:07) On that note, [applause and cheering] (00:35:14) [applause] (00:35:21) you know, I think you needed (00:35:27) [music] (00:35:38) Heat. Heat.