0: SAL KHAN: So first of all, I just want to thank Elon for coming-- hungry. You didn't even have dinner.

5: And we didn't even feed you properly. ELON MUSK: No, sorry to be a bit late. I just came from the Tesla factory in Fremont.

10: SAL KHAN: Yes. Was something wrong? ELON MUSK: There's always something. SAL KHAN: Did you have to like-- ELON MUSK: At any given point, there's always something wrong.

16: SAL KHAN: Yes. ELON MUSK: Because there's just too many things going on. So one of the trickiest things about a car

23: is that there's thousands of individual components-- there are thousands of unique components-- and even if one of those things is missing,

29: you can't make cars. So today's fiasco was-- I kid you not-- we were missing a $3 USB cable.

39: OK. So we could not complete cars, because-- SAL KHAN: So the whole line was stopped?

45: ELON MUSK: Yeah. So essentially, because it's part of the wiring harness. So you can't put the interior in without this cable.

53: And so we could either make a whole bunch of cars minus the interior, which means that you've got to stack them up in the yard.

59: SAL KHAN: The resale value would be no good. ELON MUSK: Well, it can be done, but if then things go out

64: of sequence, and it's way more inefficient--

70: you don't have a moving production line. Then you have to send people out to hundreds of cars that are sitting in the storage yard.

77: And so this happens to be a particularly pernicious cable. It's kind of routed under the carpet, in a difficult place.

87: And it's literally $3. And so we basically had to send people throughout the Bay Area

95: to go and buy USB cables. SAL KHAN: Like, literally, Radio Shack? ELON MUSK: Like Fry's. SAL KHAN: Oh, Fry's. That's better.

100: ELON MUSK: You're going to have a hard time getting a USB cable right now at Fry's, because we bought every one of them. SAL KHAN: That's good.

107: ELON MUSK: And so we're able to continue production. And I don't want to belabor the anecdote,

112: but essentially the supplier is in China. And we had plan A and plan B. And plan A

120: was like the normal supply chain process. But what the supplier did was instead

127: of sending our parts in their own package, they grouped it together with a bunch

133: of other stuff for other companies and sent that all via some extremely slow boat from China

140: to LA. And when it got to LA, the other stuff didn't pass customs.

145: And so they wouldn't let our stuff through, because-- SAL KHAN: They put it like a barrel fruit or something. ELON MUSK: I don't what they put it,

152: but something that customs didn't like. And the paperwork wasn't in order or whatever. So it got stuck there for like a couple weeks.

159: And then we had plan B. So we called and said, look you've got to air freight some of these cables--

166: cause they're just little cables-- to us. And we talked to their US subsidiary

172: and ordered from the US subsidiary, who then communicated to China. But then because this was another batch of parts,

180: so it was kind of double the order, it exceeded the credit limit that we had.

185: So it bounced off the credit limit, so they didn't ship it. SAL KHAN: Fascinating. So someone's losing their job now.

191: This is-- no, I'm kidding. You shouldn't fire anyone. ELON MUSK: I mean, it's pretty farcical.

197: And, anyway, so, it's coming like tonight at

660: PM or something.

203: SAL KHAN: Wow. And these things are happening like all the time? This was an unusual circumstance? ELON MUSK: Yeah.

208: That's like one example, but there's many things like that. SAL KHAN: I guess, I mean, that's actually

214: a really good example, because that leads into what I've always been fascinated by a lot of what you're doing.

219: Well, I'll start with, how did you get into this?

224: ELON MUSK: Into cars? SAL KHAN: Into cars. Into taking over NASA.

229: Well, not taking over NASA-- being a contractor for NASA. ELON MUSK: Just for the record, we are not taking over NASA.

236: SAL KHAN: You're not taking over NASA. They are an independent organization. But you are becoming a major provider of services for NASA.

244: Obviously, kind of internet payments and payments generally. I mean these are three completely different spaces.

250: I think a lot of people would not take someone seriously, if they had a business plan in one of these.

256: ELON MUSK: Right. Sorry to eat. SAL KHAN: Oh, yeah, take your time. What was your-- did you always think

262: you were going to be doing this or-- when did it dawn on you that you would try to revolutionize

268: three industries? ELON MUSK: Well, when I was in college-- I didn't actually expect to do it.

274: So it was not like this is some long-fulfilled expectation. But when I was in college, I thought

280: about what were the areas that would most effect the future of humanity, in my opinion. And the three areas were the internet, sustainable energy,

288: and space exploration, particularly if humanity becomes a multi-planet species. You know, there's kind of like a pretty substantial bifurcation

295: in our future, if we're either out there among stars

300: on multiple planets, or if we're confined to Earth until some obviously eventual extinction. Not Not that I'm pessimistic about live on Earth.

307: I mean, things are likely to be good. More likely to be good by far than bad. SAL KHAN: Yellowstone's due for an explosion

312: every several hundred thousand-- Shandra knows about that.

318: It's been 700,000, ELON MUSK: Right. Right. Yeah. SAL KHAN: Super volcano for those of you who don't know.

323: It would envelop, but well-- ELON MUSK: Yeah. Exactly. I know exactly what you're talking about. So-- SAL KHAN: We read the same books. I can tell.

329: ELON MUSK: Absolutely. I mean something bad is bound to happen if you give it enough time.

334: And civilization has been around for such a very short period of time that these time scales seem like very long,

342: but on an evolutionary time scale, they're very short. A million years on an evolutionary time scale

347: is really not very much. And Earth's been around for four and a half billion years, so that's a very tiny, tiny amount of time,

356: really. But for us that would be-- can you can imagine if human civilization continued at anything remotely like the current pace of technology ad

363: advancement for a million years? Where would we be? I think we're either extinct or on a lot of planets.

370: SAL KHAN: Yes. We should-- ELON MUSK: Those are the two options. SAL KHAN: But given that-- I mean, one, that's kind of as epic as one

375: can think about things, literally. How did you make that concrete?

381: How does that turn into SpaceX, Tesla and Paypal?

392: ELON MUSK: Well, so I thought about these things kind of in the abstract. Not from the expectation that I would actually

398: have careers in those arenas. But, I wanted to be involved in at least one of them.

405: And at first I thought the best bet was going to be electric cars. And so the area that I was studying

412: was advanced capacitors. So essentially capacitors that have an engine density

418: exceeding that of batteries. Because they have a very high power density, but a low energy density. Maybe you have lecture to that effect, I don't know.

423: SAL KHAN: Oh, yes, no. We should do that. We'll get to it later. ELON MUSK: Exactly.

429: So obviously, if you could make a capacitor that had anywhere near the energy density of a battery

434: with this incredibly high power density and this quasi-infinite cycle and calendar life, then you'd have an awesome solution for energy storage

441: and mobile applications. So I was going to try to work on that and try to leverage the equipment that was developed

449: for advanced chip making and photonics to create ultra-precise capacitors

457: at the molecular level. SAL KHAN: And this was when you were going to go into grad school? You had a brief stint at Stanford? ELON MUSK: That's right.

462: SAL KHAN: At a PhD in applied physics? ELON MUSK: Applied physics, material science. SAL KHAN: Right. So even then you were thinking of trying

468: to do something in the space? ELON MUSK: Actually, this was d to work on energy storage solutions for electric cars.

475: And I'd actually worked at a company in Silicon Valley called Pinnacle Research, which did advanced capacitors.

482: There were electrolytic capacitors.

489: And they actually were pretty good. They had like the energy density of a lead-acid battery, which

494: for a capacitor, that's a big deal. But they used ruthenium tantalum oxide.

500: And I think at the time, there was maybe like one or two tons of ruthenium mined per year in the world.

505: So it's not a scalable solution. But I thought there could be some solid-state solution,

511: like just using chip-making equipment. That was going to be the basic idea. But it was one of those things where I wasn't sure

517: if success was one of possible outcomes.

523: It's difficult to bound that problem exactly and say, OK-- SAL KHAN: So you're saying, I felt like this was a destined failure is another way

530: to parse that sentence. But anyway, sorry. ELON MUSK: No. I didn't think it would fail, but I wasn't sure that success was a possibility.

536: SAL KHAN: OK. Yes. ELON MUSK: And generally you want

542: to embark on something-- it's desirable to figure out if success is at least one of the possibilities.

548: SAL KHAN: Right, exactly. ELON MUSK: Because for sure failure is one of the possibilities.

553: But, ideally, you want to try to bracket it and say success is in the envelope of outcomes.

562: And I wasn't quite sure if that was the case. I think success on an academic level would have been quite

568: likely, because you can publish some useless paper-- and most papers are pretty useless--

574: SAL KHAN: We have a few-- don't take offense. ELON MUSK: I mean, how many PhD papers are actually used by someone ever?

582: SAL KHAN: That's a good point. ELON MUSK: Percentagewise it's not good. And so it could have been one of those outcomes

590: where you add some leaves to the tree of knowledge.

596: And that leaf is, nope, it's not possible.

601: And there goes seven years of my life. So that was one path.

606: And I was prepared to do that. But then the internet came along. And it was like, oh, OK, the Internet,

611: I'm pretty sure success is one of the outcomes, and it seemed like I could either do a PhD

617: and watch the Internet happen, or I could participate and help build in some fashion.

623: Like, I was just concerned with the idea of watching it happen. So I decided to put things on hold

630: and start an Internet company. And we worked on internet publishing software,

638: maps and directions, yellow pages, those kind of things. And we had as investors and customers the media companies.

646: So like the New York Times Company, Knight Ridder. SAL KHAN: And this is just at the early stages. I mean this was like--

652: ELON MUSK: '95. SAL KHAN: '95. So it's really early stages, so it's really out the gate. ELON MUSK: Yeah. Absolutely.

657: And so then we-- the reason we worked with the media companies was because we needed to have money.

663: There was no advertising money in '95. In fact, the idea of advertising on the internet seemed like a ridiculous idea to people.

671: Obviously, not so ridiculous anymore. But, at the time, it seemed like a very unlikely proposition.

676: And a lot of the media companies weren't even sure that they should be online. Like, what's the point of that? SAL KHAN: And did you all think that PayPal was just

682: going to be a simple, little internet way to-- or did you think it was going to turn into the major kind of transaction processing engine

689: that it is right now? ELON MUSK: I didn't expect PayPal's growth rate to be what it was.

694: And that actually created major problems. So we started Paypal on University Avenue.

700: After the first month or so of the website being active, we 100,000 customers. SAL KHAN: Really?

705: That fast. Wow, I didn't realize it was-- ELON MUSK: Yeah, it was nutty. SAL KHAN: And how did it start? How did people just even know to use it?

711: I mean, obviously, both buyer and seller have to be involved. ELON MUSK: Yeah. Well, we started off first by offering people $20

719: if they opened an account. And $20 if they referred anyone. And then we dropped it to $10.

727: And we dropped it to $5. As the network got bigger and bigger, the value of the network itself exceeded any sort of carrot

736: that we could offer. SAL KHAN: So much money did you all spend with that kind of $5, $10, $20 incentive to get that critical mass going?

741: ELON MUSK: It was a fair amount. I think it was probably $60 or $70 million. SAL KHAN: Oh, wow, OK. So it was substantial.

747: OK. So we're not talking peanuts here. ELON MUSK: It depends on your relative scale.

754: It's a peanut to Google. SAL KHAN: Yeah, no, that's right. That's right. ELON MUSK: Here's a peanut.

759: I mean, Google's got $50 billion. Apple's got $150 billion, some crazy amount of money. That's just cash.

765: SAL KHAN: Yeah. So it's not an outlandish-- I didn't realize that was so core. ELON MUSK: Like 1% of Google's cash would be $500 million.

772: So, you know, that's 0.1% percent of Google's cash. SAL KHAN: That's true. You're right, that's inexpensive. It's nothing. ELON MUSK: Relative to them, it's pretty inexpensive.

780: SAL KHAN: That's right. ELON MUSK: And then we did a bunch of things to decrease the friction.

789: It's just like bacteria in a Petri dish. So what you want to do is try to have one customer generate

801: like two customers. OK? Or something like that. Maybe three customers, ideally. And then you want that to happen really fast.

809: And you could probably model it just like bacteria growth in a Petri dish. And then it'll just expand very quickly

817: until it hits the side of the Petri dish and then it slows down. SAL KHAN: And then after Paypal, then I

823: mean-- to some degree, especially us in Silicon Valley, we kind of understand the Internet. We know people.

829: PayPal's obviously of the scale that is noteworthy, but then SpaceX just seems really, you know-- well,

837: one, how did you decide that I'm definitely going to do that? And then like what's the first thing that you do?

843: How do you even go out-- I don't even know how to start trying to make a rocket company.

849: ELON MUSK: Well, neither did I really. And in fact, the first three launches failed. So it's not as though it was like spot on.

857: It's like, did not hit the bull's eye. But SAL KHAN: But even getting to the point where you're launching rockets. I don't even how do you get there?

864: One, how did you decide? And then what did you do on day one? Like, who did you call? Did you write a plan? Did you start-- I don't even know.

871: ELON MUSK: Actually, the origin of SpaceX is that I was trying to figure out

877: why we'd not sent any people to Mars. Because the obvious next step after Apollo

885: was to send people to Mars. But what in fact happened was that we sent a few people

890: to the moon and then we didn't send anyone after that to the moon or Mars or anything.

895: But if you'd asked people in 1969, what would 2013 look like, they would have said, there

901: will be a base on the moon. We would have least sent some people to Mars. And maybe there'd even be a base on Mars.

906: There'd be like orbiting space hotels. And there'd be all this awesome stuff in space. And that's what people expected.

912: And if you said, well, actually, the United States in 2013 will not be able to send anyone to orbit.

918: But I'll tell you what will exist is that there'll be this device in your pocket that's like the size of-- smaller than a deck of cards

924: that has access to all the world's information, and you can talk to any one on planet Earth. And even if you're like in some remote village somewhere so

932: long as there's something called the Internet-- they wouldn't know what that means, of course-- then you would you be able to communicate with anyone

939: instantly and have access to all of humanity's knowledge. They would have said, like bullshit.

944: There's no way that that's going to be true. SAL KHAN: Right. Right. ELON MUSK: And yet we all have that.

949: And space is not happening. So I was trying to figure out like what was the deal here. And this was 2001.

956: And it was just a friend of mine asked me, what am I going to do after Paypal. And I said, well, you know, I've always been interested

961: in space, but I don't think there's anything that an individual could do in space, because it's the province of government, and usually a large government.

967: But, I am curious as to when we're going to send some one to Mars. So I went to the NASA website to try to figure out where is the place that tells you that.

974: And I couldn't find that. So I was like, either I'm bad at looking at the website, or they have a terrible website, because surely there

980: must be a date. SAL KHAN: That should be a big date. ELON MUSK: Yeah. This should be on the front page. And then I discovered actually that NASA

987: had no plans to send people to Mars, or even really back to the moon.

992: So this was really was disappointing. I thought well, maybe this is a question of national will.

1002: Like do we to get people excited about space again?

1007: And try to get NASA a bigger budget, and then we would send people to Mars.

1012: And so I started researching the area,

1019: becoming more familiar with space, reading lots of books. And I came up with this idea to do so-called Mars oasis,

1027: which was to send a small greenhouse with seeds in dehydrated gel that upon landing, you hydrate the gel.

1034: You have green plants on a red background. The public responses to precedents and superlatives.

1041: So it would be the first life on Mars. The furthest that life's ever traveled. And you'd have this money shot of green plants

1047: on a red background. So that seemed like it would get people pretty excited.

1054: So I started getting into this. And trying to figure out, OK, well can I afford to build a spacecraft?

1062: Because I had some money as a result of PayPal, but it had to fit within that budget. And I figured we had to do two missions, because if we only

1071: did one and it failed, then it might have like the opposite effect. SAL KHAN: But you were willing to bet the farm, so to speak,

1077: on this? ELON MUSK: Yeah. Well, I figured I was willing to spend half the money that I got

1084: from PayPal with no expectation of return. Because I thought this was just something

1089: that was pretty important and yeah, it seemed like I could spend half the money

1095: I made on PayPal on this, and if that got NASA a bigger budget and resulted in us going to Mars, that

1100: would be a pretty good outcome. SAL KHAN: And when your friends or your family came up to you and said, look there's nations that can't do this.

1107: You're a guy, I mean you have some resources, what did you say or do or think? ELON MUSK: Well, so I had a lot of friends of mine

1117: try to talk me out of starting a rocket company, because they thought it was crazy. And one friend of mine made me watch

1122: a video of rockets blowing up. And there were just lots of people

1128: that thought it was a really crazy idea. And there was some people that had tried to start rocket companies, not succeeded. And they tried to talk me out of it.

1135: But the thing is that-- their premise for talking me out of it was, well, we think you're going to lose the money that you invest.

1141: I was like, well, that was my expectation anyway, so I don't really mind if I lose-- you I mean, I mind,

1147: but I mean it's not like I was trying to figure out the rank-ordered best way to invest money

1153: and on that basis chose space. It's not like that's-- I thought, wow--

1159: SAL KHAN: You weren't looking at like money-market bonds, AAA bonds, rocket company. You weren't like-- ELON MUSK: I could do real estate.

1166: I could invest in shoe making. Anything. And, whoa, space is the highest ROI.

1173: That is not what-- it wasn't the premise. I just thought that it was important

1179: that humanity expand beyond Earth, and we weren't doing that, so maybe there was something I could do to spur that on.

1186: And then I was able to compress the costs of the spacecraft and everything down to a relatively manageable number.

1194: And I got stuck on the rocket. The US rockets were way too expensive. I ended up going to Russia-- I flew to Russia three times

1201: to negotiate a purchase of an ICBM. I tried to buy two of the biggest ICBMs in the Russian fleet in 2001 and 2002.

1210: And I actually negotiated a price. SAL KHAN: I'll just let that statement stand. I'm not even going to-- Well, actually,

1219: I have to-- like who did you call? ELON MUSK: You open the yellow pages.

1225: Go to ICBMs. Oh! SAL KHAN: How does this-- I don't want to get too much in to it but I'm

1230: curious about this one particular thing. You decide at some point you need to buy an ICBM? ELON MUSK: Yeah. Well, actually at first I tried to buy just a normal launch

1237: program that they use to launch satellites, but those are too expensive. SAL KHAN: I see. I see. ELON MUSK: The Boeing Delta II would have cost $65 million

1245: each, so two would have been $130 million. And then I was like, woah, OK, that breaks my budget right there.

1254: And I tried to negotiate with them. And that was not-- I did not make progress. SAL KHAN: How much does an ICBM go for? I'm curious what's the market rate for one of those?

1261: ELON MUSK: Well-- SAL KHAN: This is right after the fall, it might have gone up. ELON MUSK: Yeah, it's gone up a lot since then.

1267: But in 2001, it would've been about $10 million each. So two would have been $20 million.

1273: And then I thought I could get the rest of the mission

1278: down to also around $10 million per, so we'd have a dual mission with like two identical launches,

1284: two identical spacecraft for roughly $40 million.

1289: And so I thought, OK, I can do that. SAL KHAN: But you must have had some like rocket scientists advising you at this point?

1297: This sounds like you were serious. I mean you were-- ELON MUSK: Yeah. I engaged a bunch of consultants and started

1305: to get familiar with the space industry. But then after the third trip to Russia,

1312: I came to realize that I was actually wrong about my first premise, that there was a lack of will.

1318: In fact, I think that there's a tremendous amount of will in the United States for space exploration.

1323: Because the United States is essentially a nation of explorers. I mean, it's a distillation of the human spirit of exploration.

1330: So of course it was quite silly of me to think that people lacked motivation.

1336: But what people don't want to think is that, OK, sending people to Mars is going to be so expensive that they'll

1342: have to give up health care or something. They're not going to do that. So it's got to be that going to Mars

1348: is not going to cause some meaningful drop

1353: in their standard of living. So if it's like maybe a quarter of a percent or half a percent of GDP-- something

1359: like that is palatable. Anyways, so I thought, OK, it's not really

1366: going to maybe matter that much if I do this mission, because what really matters is having a way.

1374: So I was wrong-- I thought there wasn't enough will, but there actually was plenty of will,

1379: if people thought there was a way. So then I said, OK, well, I need to work on the way.

1385: How hard is it really to make a rocket? Historically, all rockets have been expensive, so therefore,

1390: in the future, all rockets will be expensive. But actually that's not true. If you say, what is a rocket made of.

1396: And say, OK, it's made of aluminum, titanium, copper,

1402: carbon fiber, if you want to go that direction. And you can break down and say, what is the raw material

1408: cost of all these components. And if you have them stacked on the floor and could wave a magic wand so that the cost of rearranging

1415: the atoms was zero, then what would the cost of the rocket be. And I was like, wow, OK, it's really small.

1421: It's like 2% of what a rocket costs. So clearly it would be in how the atoms are arranged.

1430: So you've got to figure out to OK, how can we get the atoms in the right shape

1436: much more efficiently. And so I had a series of meetings on Saturdays

1444: with people some of whom were still working at the big aerospace companies, just to try to figure out is the some catch here

1450: that I'm not appreciating. And I couldn't figure it out. There doesn't seem to be any catch. So I started SpaceX.

1456: SAL KHAN: And you ended up-- you had some failures, but obviously some huge successes. What was the cost that you were able to build this rocket for

1463: relative to what they were being built for before? ELON MUSK: So let's see. For the Falcon 1, which is the first rocket we built.

1471: And the first three flights did not make it. In fact, we got progressively further.

1478: But like the first rocket came in and landed maybe

1483: a couple hundred yards away from the launch site, and tiny fragments.

1490: So, yeah, anyway, that rocketed ended up costing around $6 million compared

1500: to other rockets in that class, which were about to $25 million. SAL KHAN: Wow.

1505: So significant? ELON MUSK: Yeah, like a quarter.

1511: But there's an even better step beyond that which is to make rockets reusable.

1516: Right now that is around what our comparison price is-- excluding the refurbished ICBMs.

1523: So, if you say building a rocket from new, how does the SpaceX rocket compare to a rocket from Boeing or Lockheed?

1530: It's about a quarter of the price. However, if we make it reusable, then it

1537: can be two orders of magnitude cheaper. SAL KHAN: Two orders of magnitude cheaper. A 100th of the price?

1542: ELON MUSK: That's right. For you. SAL KHAN: Only today.

1548: Memorial day sale. And I've seen some-- you all are doing these vertical landings,

1555: like literally out of like the 1950s Sci-Fi movies. And that's what you're talking about?

1561: ELON MUSK: Yeah. Essentially, the rocket needs to come back and land at the launch site, and then reload propellant

1566: and take off again. Like an airplane in its reusability. SAL KHAN: How far do you think we are from that?

1572: When do you think-- your best guess, when we'll actually see that happening? ELON MUSK: Well, I'm hopeful we can do it next year.

1577: SAL KHAN: Oh, OK. Yeah. That's-- we've got some ambitious stuff at Khan Academy for the next year, too.

1582: So we can compare. We're redesigning the site. ELON MUSK: Right.

1590: We've been working on it for a long time. I should say, SpaceX has been around for 11 years, and thus far we have not recovered any rockets.

1597: We recovered the spacecraft from orbit. So that was great.

1602: But none of our attempts to recover the rocket stages have been successful. The rocket stages have always blown up essentially

1609: on reentry. Now, we think we've figured out why that was the case.

1615: And it's a tricky thing, because Earth's gravity is really quite strong. And with an advanced rocket, you can do maybe 2% to 3%

1624: of your lift-off mass to orbit, typically. And then reusability subtracts 2% to 3%

1633: So then you've got like nothing to orbit or negative. And that's obviously not helpful.

1640: And so the trick is to try to shift that from say 2%, 3% in an expendable configuration

1649: to make the rocket mass efficiency, engine efficiency, and so forth, so much better that it moves to maybe around

1657: 3.5% to 4% in expendable configuration. And then try to get clever about the reusability elements

1663: and try to drop that to around the 1.5% to 2% level. So you have a net payload to orbit of about 2%.

1669: SAL KHAN: But you're doing it at one, two orders of magnitude cheaper. ELON MUSK: Yeah. Absolutely, because our Falcon 9 rocket cost about $60 million.

1677: But the propellant cost-- which is mostly oxygen-- it's two-thirds oxygen, one-third fuel--

1685: is only about $200,000. SAL KHAN: Wow. ELON MUSK: And it's much like a 747.

1691: It costs about as much to refuel our rocket as it does to refuel a 747 within-- well, pretty

1700: close, essentially. SAL KHAN: So assuming you all are successful, and you all have proven yourself to be successful on these audacious things in the past, I mean,

1706: what happens? I mean that seems like it's-- what happens in the next 5, 10 years in the space industry, if you all are successful

1712: there? I mean do we get to Mars? Do we have kind of market forces, commercialization

1718: of space starting to happen? ELON MUSK: Yeah. Let's see.

1725: Well, the first step is that we need to earn enough money to keep going as a company. So we have to make sure that we're launching satellites.

1735: Commercial satellites like broadcast communications,

1740: mapping, government satellites that do scientific missions.

1746: Earth-based or space-based missions. GPS satellites. That kind of thing. And then also servicing the space station.

1753: Transferring cargo to and from the space station, which we've done a few times. And then taking people to and from the space station.

1760: So we've got to service the sort of Earth-based needs

1766: to launch satellites and that pays the bills. But in doing that keep improving the technology

1774: to a point where we can make full reusability work. And we have sufficient scale and sophistication

1782: to be able to take people to Mars. SAL KHAN: Wow. So you think this is going to be a reality? What's your best guess of when we're

1789: going to have someone on Mars? ELON MUSK: I think probably about 12 years. SAL KHAN: That's nothing. And you think it'll be a round trip?

1796: It won't just be some type of permanent colony on Mars? ELON MUSK: I think it's probably a round trip.

1803: It's not for sure. SAL KHAN: I could talk about this for-- people know, I'm-- ELON MUSK: Aspirational it'd be a round trip.

1809: SAL KHAN: This is mind blowing. And then on Tesla. I mean Tesla's obviously, from my vantage,

1816: it's a huge success. What do you think in that industry-- well, one,

1823: I'll ask kind of the same question. What did you think-- this is something that GM and Toyota and these massive multi-billion dollar

1829: organizations have been trying. What gave you the confidence to pursue it? And now that it seems to be a huge success, where

1836: do you think this industry's going to be the next 5, 10 years? ELON MUSK: Yes. So with Tesla, the goal is try to accelerate

1844: the advent of sustainable transport. I think it would happen anyway, just out of necessity.

1850: But because we have an un-priced externality in the cost of gasoline.

1855: We weren't pricing in the environmental effects of CO2 in the oceans and atmosphere.

1861: That's causing the normal market forces to not function properly. And so the goal of Tesla is to try

1869: to act as a catalyst to accelerate those sort of normal forces.

1874: The normal sort of market reaction that would occur. We're trying to have a catalytic effect on that.

1881: And try to make it happen, I don't know, maybe 10 years sooner than it would otherwise occur.

1889: That's the goal of Tesla. So that's the reason we're making electric cars and not any other kind of car.

1895: And we also supply powertrains to Toyota and to Mercedes

1900: and maybe to other car companies in the future to accelerate their production of electric vehicles.

1909: So that's the goal there. And so far, it's working out pretty well. SAL KHAN: I mean, I just saw a news report earlier today

1915: that you all sold more Model S's than-- you all are leading that segment of the industry. The Mercedes S class, the BMW 7 Series, or the Lexus LS400,

1925: or whatever it is. ELON MUSK: Yeah, actually, that seems to be the case.

1931: I didn't realize they sold so few cars in that segment. Because we don't sell that many cars.

1937: We sell 5,000 a quarter, or 12,000-- SAL KHAN: Well, out here they seem like you know every-- ELON MUSK: Well, this is our home team.

1946: So it's-- we better sell a lot in the Bay Area. Because otherwise we're like--

1951: SAL KHAN: And, well, I mean, similar thing. How did you start? What gave you the confidence?

1957: And do you see yourselves as kind of a major automotive, mainstream brand in 5,10 years,

1962: all the way down to competing with the Honda Accords and Civics?

1967: ELON MUSK: I mean, yeah. Our goal-- it's not to become a brand big brand

1973: or to compete with Honda Civics, rather

1979: to advance the cause of electric vehicles. And so we're just going to keep making

1985: more and more electric cars and driving the price point down until the industry is very firmly electric.

1993: Like maybe half of all cars made are electric or something like that. Which is not to say that we expect to make half of all cars.

1999: We want to just have that catalytic effect until at least that occurs. And I think the point at which we're approaching half

2007: of all new cars made are electric, then I think I would consider that to be the victory condition.

2013: And so the faster we can bring that day, the better. SAL KHAN: When would be your guess when that happens?

2022: ELON MUSK: Well, I made a bet with someone about three years ago that it would be sooner than 20 years.

2027: So it's 17 years from now. But that's conservative.

2034: I think it's probably maybe 13 or 14 years.

2042: SAL KHAN: Wow. Right when we're going to Mars. ELON MUSK: Right. SAL KHAN: It'll be exciting times.

2050: ELON MUSK: True. Exactly. I was just thinking about that. It was like, oh, those time frames are kind of coincident.

2060: The nature of new technology adoption is it tends to follow an S-curve. So what usually happens is people under-predict it

2066: in in the beginning, because people tend to extrapolate in a straight line. And then they'll over-predict it at the midpoint,

2072: because there's late adopters. And then it'll actually take longer than people think at the mid-point,

2079: but much shorter than people think at the beginning.

2084: But I'm pretty excited about how things are going. And, in fact, I think that the pace of technology improvement

2091: in electric energy storage is really moving faster than anyone thinks.

2097: SAL KHAN: Wow. I got one more-- how are we doing on time? Where's Ester? Oh, 9 o'clock. So how much time do you have?

2103: I want to make sure we don't go over. ELON MUSK: Well, I guess maybe another 15 minutes. SAL KHAN: OK. So I'll finish with one last question and then we'll open it up.

2109: What advice do you have for us at Khan Academy? ELON MUSK: I don't know. You guys seem to be doing really great. So I was wondering if you had advice for me.

2116: SAL KHAN: Oh, no, well. ELON MUSK: Yeah.

2121: It seems like you're doing an amazing job of-- really super leveraged. I mean, obviously, a small team, and you're

2127: having a dramatic effect-- SAL KHAN: Yeah, half these people don't even work here. There just like-- so it's like it's even--

2132: ELON MUSK: Right. So it's, I think very impressive thing you're doing to spread knowledge and understanding

2141: throughout the world. SAL KHAN: The universe soon, if you hold up your end of the bargain.

2148: ELON MUSK: It's actually kind of funny. If you think, what is education? Like you're basically downloading data and algorithms

2154: into your brain. And it's actually amazingly bad in conventional education.

2161: Because like it shouldn't be like this huge chore. So you're making it way, way better.

2170: But I think a lot of things that I would say, you've probably heard 100 times.

2176: And, in fact, are if not doing. The more you can game-ify the process

2183: of learning, the better. For my kids, I do not have to encourage them to play video games.

2188: I have to like pry them from their hands, like crack. SAL KHAN: Yes. ELON MUSK: It's like, drop that crack needle.

2195: SAL KHAN: You have that problem at your house, too. The crack is addictive.

2201: ELON MUSK: So to the degree that you can make somehow learning

2206: like a game, then it's better. And I think, unfortunately, a lot of education

2212: is very vaudevillian. You've got someone standing up there kind

2217: of lecturing at people. And they've done the same lecture 20 years in a row, and they're not very excited about it.

2223: And that lack of enthusiasm is conveyed to the students. They're not very excited about it.

2228: They don't know why they're there. Like why are we letting this stuff. We don't even know why. In fact, I think a lot of things that people

2234: learn that probably there's no point in learning them. Because they never use them in the future. SAL KHAN: Because who's going to launch a rocket into space?

2241: I mean, that's just like-- exactly, that never happens. ELON MUSK: Well, you have to say-- people don't stand back and say, well,

2248: why are we teaching people these things. And we should tell them, probably, why we're teaching these things.

2253: Because a lot of kids are probably just in school, probably puzzled as to why they're there.

2260: I think if you can explain the why of things, then that makes a huge difference to people's motivation.

2266: Then they understand purpose. So I think that's pretty important.

2273: And just make it entertaining. But I think just in general conventional education should be massively overhauled. And I'm sure you pretty much agree with that.

2281: I mean the analogy I sometimes use is, have you seen like Batman, the Chris Nolan

2287: movie, the recent one. And it's pretty freaking awesome. And you've got incredible special effects, great script,

2295: multiple takes, amazing actors, and great sound,

2301: and it's very engaging. But if you were to instead say, OK--

2307: even if you had the same script, so at least it's same script. And you said, OK, now that script, instead of having movies, we're going

2313: to have that script performed by the local town troop. OK, and so in every small town in America,

2319: if movies didn't exist, they'd have to recreate The Dark Night. With like home-sewn costumes and like jumping across the stage.

2330: And not really getting their lines quite right. And not really looking like the people in the movie.

2336: And no special effects. And I mean that would suck.

2342: It would be terrible. SAL KHAN: That's right. Very-- ELON MUSK: That's education.

2349: SAL KHAN: So with that-- and I apologize to all of you guys for hogging up all of the time, because, obviously, I could talk for hours about this stuff.

2355: But we do have time, probably 5 or 10 minutes for a handful of questions. If none of you all have any, I have about nine more.

2362: But, yes. SPEAKER 1: So I noticed-- I picked up two kind of themes from what you were discussing.

2367: One was somewhat audacious goals. And the other was I don't think I heard you use the word

2372: profit in anything that you spoke about. You seem to be-- each thing is pointed at like re-invigorating

2381: an industry or bringing back space missions. How much of your success do you attribute

2387: to having really audacious goals or versus just not being focused on the short term, money coming

2396: in, or I don't know, investors? ELON MUSK: Unfortunately, one does have to be focused on the short time and money coming

2404: in when creating a company, because otherwise the company will die. So I think that a lot of times people

2411: think like creating company is going to be fun. I would say it's really not that fun.

2417: I mean there are periods of fun. And there are periods where it's just awful.

2423: And, particularly, if you're the CEO of the company, you actually have a distillation of all the worst

2428: problems in the company. There's no point in spending your time on things that are going right.

2434: So you're only spending your time on things that are going wrong. And there are things that are going wrong that other people can't take care of.

2442: So you have like the worst-- you have a filter for the crappest problems in the company. The most pernicious and painful problem.

2451: So I think you have to feel quite compelled to do it.

2457: And have a fairly high pain threshold. There's a friend of mine who says, starting a company is like starting into the abyss

2463: and eating glass. And there's some truth to that.

2470: The staring into the abyss part is that you're going to be constantly facing the extermination of the company.

2477: Because most start ups fail. It's like 90% percent-- it could be 99% of start ups fail.

2485: So that's the staring into the abyss part. You're constantly saying, OK, if I don't get this right,

2492: the company will die, which can be quite stressful.

2497: And then the eating glass part is you've got to work on the problems

2504: that the company needs you to work on and not the problems you want to work on. And so you end up working on problems that you really

2513: wish you weren't working on. And so that's the eating glass part.

2519: And that goes on for a long time. SPEAKER 1: So how do you keep your focus on the big picture when you're constantly faced with,

2525: we could be out of business in a month? ELON MUSK: Well, it's just a very small percentage

2532: of mental energy is on the big picture. Like you know where you're generally heading for

2538: and the actual path is going to be some sort of zigzaggy thing in that direction.

2544: You're trying not to deviate too far from the path that you want to be on, but you're going to have to that to some degree.

2553: But I don't want to diminish the-- I think the profit motive is a good one, if the rules of an industry

2559: are properly set up. So there's nothing fundamentally wrong with profit. In fact, profit just means that people

2565: are paying you more for whatever you're doing than you're spending to create it.

2570: That's a good thing. And if that's not the case, then you'll be out of business.

2576: And rightfully so. Because you're not adding enough value.

2582: Now there are cases, of course, where people will do bad things in order to achieve profit,

2589: but that's actually quite unusual. Because usually the rules are set up mostly correctly.

2596: Like not completely, but mostly correctly. SAL KHAN: I think we have time for one more question.

2602: Joel. JOEL: Yeah, I have an important one. SAL KHAN: OK, very good. Yes, please. SPEAKER 3: No. JOEL: OK, so few months ago, you teased Hyperloop,

2609: and we haven't heard anything since. So, first of all, a few of us engineers were talking about it, and I think

2615: we have a few ideas, if you need help. But, if you feel comfortable, maybe you

2621: could tell us a little bit more. ELON MUSK: I was reading about the California high-speed rail, and it was quite depressing.

2629: Because California taxpayers are going to be on the hook to build the most expensive high-speed rail

2638: per mile in the world-- and the slowest.

2643: Those are not the superlatives you want.

2648: And, it's like, damn, we're in California, we make super high-tech stuff. Why are we going to be spending--

2654: now the estimates are around $100 billion-- for something that will take two hours to go from LA to San

2661: Francisco? I'm like, OK, well, I can get on a plane and do that it 45 minutes. It doesn't make much sense.

2668: And isn't there some better way to do it than that. So if you just say, OK, well what would

2677: you ideally want in a transportation system? You'd say, OK, well you'd want something

2682: that relative to existing modes of transportation is faster-- let's say twice as fast-- costs half as much

2690: per ticket, can't crash, is immune to weather, and is--

2697: you can make the whole thing like self-powering with like solar panels or something like that. That would be pretty-- SAL KHAN: That would be great, yes.

2702: ELON MUSK: --a good outcome. And so what would do that?

2707: And what's the fastest way short of inventing teleportation that you could do something like that? And some of the elements of that solution are fairly obvious,

2716: and some of them are not so obvious. And then the details-- the devil's

2721: in the details of actually making something like that work. But I came to the conclusion that there

2727: is something like that that could work. And would be practical.

2732: SAL KHAN: Is this around the evacuated tubes? The vacuum tubes?

2738: Like the old bank-- ELON MUSK: It's something like that. SAL KHAN: But you haven't been more public with what this is?

2744: ELON MUSK: No. Although I did say that once Tesla was profitable that I would talk more about it.

2751: But, we haven't done our earnings call yet. So I think I'll probably do it after the earnings call.

2758: And the thing is I'm kind of strung out on things that I'm already doing. So adding another thing-- it's like doesn't-- it's a lot

2767: SAL KHAN: Learning the guitar You could pick up all sorts of things. ELON MUSK: Right. I tried learning the violin.

2773: That's, by the way, a hard thing to learn. SAL KHAN: Yeah. Launching rockets, electric cars,

2779: revolutionizing transportation. Yeah, it's easy. ELON MUSK: I cannot play the violin at all. Very horrible.

2785: If you think about the future, you want a future that's better than the past, and so if we had something like the Hyperloop,

2792: I think that would be like cool. You'd look forward to the day that was working. And if something like that, even if it was only

2801: in one place-- from LA to San Francisco, or New York to DC or something like that-- then it would be cool enough

2810: that it would be like a tourist attraction. It would be like a ride or something. So even if some of the initial assumptions didn't work out,

2817: the economics didn't work out quite as one expected, it would be cool enough that like, I want to journey to that place just to ride on that thing.

2825: That would be pretty cool. And so that's I think how-- if you come with a new technology,

2831: it should feel like that. You should really-- if you told it to an objective person, would they look forward to the day

2836: that that thing became available. And it would be pretty exciting to do something like that.

2842: Or an aircraft. Like I thought it was really disappointing when the Concorde was taking out of commission, and there was no supersonic transport available.

2851: And of course the 787 has had some issues.

2857: But the thing is, the 787 even in the best case scenario is only a slightly better version of the 777.

2864: And it's like, OK, not that exciting. SAL KHAN: So this is something that you are working on?

2869: ELON MUSK: I wouldn't say working on it. SAL KHAN: And one day in the not-too-far future-- or there's some plans or consultants involved or something?

2877: SPEAKER 4: You called Russia. SAL KHAN: You made some phone calls to Russia. ELON MUSK: No, every now and then, it's percolating away.

2884: I'm not actually thinking about it. But then they'll be some new element of that I'll think of. Oh, this would make it better.

2889: SAL KHAN: Fascinating. Well, I think I'm speaking for everyone. This is like the most epic possible conversation

2897: one could have over about the course of an hour. And I think all of us would love to chat with you for hours on end, but thank you so much.

2904: I know you have a lot of free time, so it probably wasn't that big of a deal for you to come here. But, it was a huge honor.

2909: And I think it's inspired all of us to go out and change the world and the universe. ELON MUSK: Cool. All right.

2914: SAL KHAN: Thank you very much. [APPLAUSE]