马斯克Code Conference完整采访

0:00You know let me start by saying we're very glad you're here safe and sound thanks. I had to start with lead. I was

0:07Flank I flew here with the landing gear down because it was like some kind of landing gear issue

0:11Lending air was stuck. There's some kind of warning light

0:15my

0:16Pilot said that if they were attracted to landing gear it may not go back down again, so

0:21So this happens to you. Yeah

0:24Yeah.

0:25Okay.

0:26And we're glad you're here today.

0:27I'm glad you're here today.

0:28I'm glad you're here today.

0:29So, anyway, we're happy to wait for you.

0:32And we're glad you got here safe.

0:33It's great to see you guys.

0:34Thank you for coming.

0:35I really appreciate it.

0:36You kept your promise, which was nice.

0:38Of course.

0:39I think you were drunk when you promised me, but that's okay.

0:40I'll take it.

0:42So, you know, in the couple of years since you've come, you've done some astonishing things

0:47in terms of substantive stuff with both your companies.

0:52Let's start talking about space and what you've been doing there.

0:55Obviously you've had some success landing the rocket.

0:59You've done a bunch of other things

1:01that were people thought you weren't going to be successful.

1:03Talk a little bit about the progress you think you've

1:06made with SpaceX.

1:07Sure.

1:08Well, there's a lot of things where I think I

1:09didn't think would be successful.

1:11So the most significant thing is being

1:16able to land an orbit class rocket, a brew stage, and bring it both back to Cape Canaveral,

1:26land and land, and be able to land on a drone ship out in the ocean.

1:34There is a bit of an education process that's needed to understand orbital dynamics because

1:40a lot of people are confused of like why the heck are you landing a ship, landing a rocket

1:45in a ship in the ocean.

1:46That seems pretty inconvenient.

1:49And the reason is because that going up and staying up

1:53is actually about velocity horizontal to the Earth's surface.

1:58So there's a huge difference between space and orbit.

2:04It's like space you could think of as like,

2:06say being the international waters boundary

2:08for the Pacific Ocean.

2:09Like if you go, you know, 100 miles or shore,

2:13you're technically out of coastal waters now,

2:16you're in the Pacific.

2:17So it's like technically you're in the Pacific.

2:19But orbit is like circumnavigating the globe.

2:23It's a really giant difference.

2:25And the reason that things go up and stay up

2:29is because you're zooming around the Earth so fast

2:32that your outward radial acceleration is

2:34equal to the inward acceleration of gravity.

2:37And so those balance out and you have a net zero gravity.

2:41So when you see the space station, the thing that's a little sort of counterintuitive is

2:46that the space station is actually zooming around the earth at 17,000 miles an hour.

2:52Even though it seems like it's just...

2:53It seems really still, you know, but it's moving really, really fast.

2:58I mean to put that into perspective, a bullet from a 45 gun, you know, a handgun is just

3:08below the speed of sound.

3:09So the space station is going more than 25 times faster than that.

3:14And that's what's needed actually to go up and stay up.

3:18And that's why there's the term escape velocity, not escape altitude.

3:23There's no such thing as an escape altitude.

3:25There's only an escape velocity.

3:26You need to be a certain speed to escape the gravity of the Earth.

3:29Yeah, you can think of gravity as kind of a funnel in space time.

3:32So think of like a coin funnel.

3:36final. It's very much like that, but it's obviously a sort of four-dimensional coin

3:42funnel. If you spin a model, or a coin, on a coin funnel, when it's far out, it sort of

3:52spins slowly, and then as it gets closer, it spins faster and faster. If you were to

3:58start at the bottom of the coin funnel and you wanted to exit, you'd spin it horizontally,

4:04and it would spin out, and that's really how you get to orbit.

4:11Yeah, so gravity well is like the final.

4:14Why you wanna land on a ship in the ocean?

4:17Because in order to get to orbit,

4:20all that matters is your horizontal velocity.

4:22Your altitude doesn't really matter.

4:25In fact, the force of gravity at, say,

4:30the phenomenal boundary of space, 100 kilometers,

4:35is almost exactly the same as it is on the surface of the earth.

4:40It's a few percent lower than the surface of the earth.

4:45So in order to go up and stay up,

4:48the only thing that matters is how fast are you

4:49going horizontal to the earth's surface?

4:52So you have that outward radial acceleration,

4:54or think about like maybe like tether ball

4:55or something like that.

4:56It's really that outward acceleration

4:58is the thing that matters.

5:00And so when the rocket is going to orbit,

5:03the only reason it's going up is to get out

5:05of the thick part of the atmosphere,

5:07because at high velocity, the atmosphere is thick as molasses.

5:12And so it goes up very briefly.

5:13But if you look at a long exposure of the rocket's

5:18trajectory, you'll see it goes up, but immediately curves over

5:21and starts going horizontal.

5:24And so at the point at which the stages separate,

5:34those two stages, the primary boost stage, which is the most

5:37expensive part of the rocket, the point at which that staging

5:41occurs can be as high as Mach 10.

5:47So it's going away from the launch site 10 times the speed

5:50of sound.

5:52So in order to get back to the launch site,

5:55you would have to have enough fuel and oxygen

5:59to reverse out that velocity and boost back

6:04all the way to the launch site.

6:05And you just don't have the physics of it

6:08don't really allow you to have that much.

6:10It's not about saving money on fuel or anything.

6:13It's just physically impossible.

6:18Because another sort of thing about,

6:20if you're in space is that there's nothing to react against.

6:24So like whereas an aircraft can circle very easily

6:28because it's reacting against air in vacuum,

6:32there's nothing to react against.

6:34So the only way to go back the other direction is to apply

6:36just as much energy as it took you to go,

6:38if you want to go backwards,

6:40you have to apply just as much energy as it took you

6:42to go forwards.

6:43In fact, well twice as much really,

6:45because you go to zero it out and then you've got to.

6:47You've got to land elsewhere.

6:49Yeah, so bottom line is this thing is zinging out to zinging out to see it super at 10 times

6:54fast enough for it.

6:55That point, it may well be over the ocean because the ocean covers most of the...

6:58Oh, it's actually...

7:00The point of separation, it's not that far away.

7:02It's maybe 100 kilometers away from the launch site, but it is going like hell in the opposite

7:08way from the launch site.

7:10So the only way to really land it is to have it continue on that arc, that ballistic arc,

7:19and then land far out to sea on a ship that's pre-positioned to a particular latitude and

7:26longitude, very precise to within about a meter.

7:30And then the rocket will then go from vacuum through rarefied air at hypersonic velocity.

7:40So when it's in vacuum it has to, obviously you can't use aerosol surfaces, you have to

7:44use nitrogen jets to control the attitude and position.

7:50And then as it starts to encounter the air,

7:56we use grid fins because grid fins looked like a waffle.

8:02That they work quite well across a wide regime

8:05from both very high velocity hypersonics

8:08through supersonic, transonic, and subsonic.

8:12So it's hard to have arrow surfaces that work well

8:15across that entire regime.

8:17And then, so once the arrow forces become high,

8:21it uses the four grid fins to sort of control its attitude.

8:26To land it, so.

8:27Yeah, it's controlling pitch your end roll

8:31with the grid fins.

8:34And then once, and those grid fins will then

8:38position it to where it's fairly close to the ship

8:42and then it will light, in this case, three of the nine

8:44engines to arrest the velocity and then drop to one engine

8:49for precision right before landing.

8:51Right, OK.

8:52Super hard.

8:53Super hard.

8:54Maybe a bit of a long explanation.

8:55OK, what we're going to get to is that's super fucking hard.

8:59There's a video.

9:00So why is that important?

9:02Why has that this moment been important for you?

9:04Well, so in order to reuse the boost age, which is about 70% of the cost of the rocket,

9:15so that.

9:16Which cost is that?

9:17How much is that?

9:18Well, I mean, it's sort of on the order of $30 to $35 million.

9:22Right.

9:23So you want to save that?

9:24Yeah.

9:25I mean, it's like, I try to tell my team, it's like, imagine there was a pallet of cash that

9:28was formatting through the atmosphere.

9:33it was going to burn up and smash into tiny pieces, would you try to save it?

9:37Right, right, right.

9:38Probably yes.

9:39Yes, okay.

9:40Yeah, that sounds like a good idea.

9:42Right, okay.

9:43Uh-huh.

9:44So, so yeah, so we want to get a bag.

9:48Right.

9:49And that way we don't have to make another one.

9:52Right.

9:53And I think it's quite tragic if rockets like get smashed into tiny pieces and land the

9:57part of the aircraft.

9:58So can I ask a question?

9:59We've been going to space for, what, 50 years or something like that.

10:06Nobody, until you started doing this and Jeff Bezos' company has done it, the government

10:11never sort of saved the rockets.

10:13They never saved the palette of cash.

10:16Why not?

10:17And the Russians didn't either.

10:18I mean, well, it wasn't a real air.

10:20Yeah, I mean, there was some attempt to do that with the space shuttle, but there was no

10:24return.

10:25It's the first time that a rocket beast has returned to launch site from an orbital mission,

10:31and certainly the first time that there's been a landing on a ship out to see.

10:34But the regular rockets that weren't designed like planes never tried to do this.

10:41Right.

10:42The plane thing is not a good idea in my view.

10:48So the plane, and the reason I think it's intuitively it seems like a plane should work, but actually

10:54If you consider that really every mode of transport

10:57has a design that is appropriate to its medium.

11:01And if you're in space, wings are not very useful,

11:06because there's no air.

11:11And then if you want to go somewhere other than Earth,

11:14there's also no runways.

11:17So these are important considerations.

11:20So that's why when they went to the moon, they used propulsive landing.

11:26Right. But what I'm saying was when they built the space shuttle,

11:29it sort of was like a, looked like a kind of bulb.

11:32I think that appealed to Congress. Yeah.

11:34Yeah.

11:35Yeah.

11:36In my school.

11:37It's cool.

11:38Yeah.

11:39It's like an airplane.

11:40Can you explain the, you know, Jeff Bezos.

11:42Good one.

11:43Jeff Bezos was here last night and I asked him, what's the difference

11:46from what you're doing and what Elon Musk is doing?

11:48And he said, well, I think you were like-minded

11:54in the general sense of it,

11:56and then he went on to explain some differences.

11:57Do you?

12:00But he talked about, and correctly if I'm misquoting him,

12:04but I think he was saying,

12:06this is all about laying the foundations

12:09of being able to do greater things

12:11by getting the basic infrastructure

12:14of being able to reuse these rockets down right.

12:18Do you, is that correct?

12:20Do you have a similar starting point from him

12:23and your thinking process?

12:25I think there's certainly some similarities of opinion.

12:27I think both Jeff and I believe that is important

12:30to the future to be a space-braying civilization.

12:33And ultimately, you'd be out there among the stars.

12:35And I think that's the exciting, inspiring future

12:39that I think certainly people in this room want.

12:43Well, particularly after seeing that the asteroids

12:45are going to destroy this planet.

12:46I mean, I mean, I don't.

12:48I mean, I don't view it as, you know, we want,

12:50we, I mean, I think what, when I say, you know,

12:54multi-planets species, like that's really what we want to be.

12:57It's not like, you know, still being a single planet species,

13:00but moving planets.

13:01It's really being a multi-planets species

13:05and having civilization and life as we know extend beyond us

13:09to the rest of the solar system

13:10and ultimately to other star systems.

13:13I think that's the thing, that's the future

13:15that's exciting and inspiring.

13:16I think that's what, you know, I think, you know,

13:20any kind of, any things like that to make,

13:22to be glad to wake up in the morning.

13:24You know, like life, life can't be just about solving problems.

13:27Like they have to be things that are inspiring

13:28and deciding to make you glad to be alive.

13:31So what, in the immediate timeframe,

13:33what is your goal for SpaceX, now that you've done this,

13:37which is a huge accomplishment,

13:38what is the plan for you in the immediate time

13:42and then the longer range?

13:43Sure, so we plan to re-fly one of the landed rocket boosters hopefully in about two or

13:56three months, something like that.

14:02And then that will be an important milestone.

14:04So far the stages are looking quite good.

14:07Even though they come through quite,

14:10there's a really difficult entry situation,

14:14but they're looking, like they're in good shape.

14:18And we now have four of them.

14:19So we wanna start reflaying them

14:23towards the end of summer.

14:25And then hopefully by the end of the year,

14:28we'll be launching Falcon Heavy,

14:30which will be the most powerful rocket in the world

14:35by more than a factor of two.

14:36So Falcon Heavy is, will be on the order of five million

14:39pounds of thrust on lift torque, which is about two thirds

14:42the size of a seven five.

14:44Oh, really?

14:45Yeah, so that's the rocket that took the astronauts

14:48to the moon.

14:48Right, exactly.

14:49So in fact, launching from the same pad.

14:52From the Apollo 11 pad.

14:54Wow.

14:55That's amazing.

14:56So you're going to launch that Falcon Heavy by the end

14:59of this year, you see?

15:00Yeah, that's our aspiration.

15:02Now, that's somewhat of a delay from when you first

15:04hope to launch it, right?

15:06Yeah, but the I mean it's not like we had a lot of pressing customers who wanted us to launch it

15:17In fact the first launch will will not have any operational satellites it'll be a demonstration launch and

15:22The first operational flights where customers actually want us launch it or next year

15:28Where is there's there's a lot of customers who want us to launch Falcon 9?

15:33So about a quarter of our flights are for NASA, but three quarters are for commercial satellites,

15:40of broadcasting communication satellites, or science missions for other countries.

15:47And there's quite a backlog.

15:50And we had an issue with the rocket last year, so that part of that is six-month hold in

15:54our schedule.

15:55So we sort of backlogged on our launches, and we're trying to get them out as quickly

15:59we can and service our customers.

16:06So the launches will take place every two to four weeks.

16:10It was quite a high launch cadence.

16:11That's a much faster cadence than NASA had, right?

16:15Yeah, it'll be more launches than anything else in the world.

16:20So more than Russia, more than Europe,

16:22more than China by next year, certainly.

16:25And largely to deliver customers,

16:27Yeah, it's, it's, it's a lot of broadcasting communication satellites that are going to

16:33geosynchronous orbit.

16:36And, and then there's, we're, we'll also be launching the new iridium constellation.

16:43So that radium's got a next generation constellation of satellites, I think 60 or 70 satellites,

16:49quite, you know, decent sized satellites.

16:53That'll be, I think many orders of magnitude improvement over the current radium system.

16:57So we'll be able to have global broadband.

17:00So that'll be a whole bunch of launches.

17:03And then next year, we'll be flying Dragon Version 2,

17:10which is the one that's capable of taking up

17:12to seven astronauts to the space station.

17:16And in fact, Dragon 2 really is a propulsive lander as well.

17:21And it'll be the, it's intended to carry

17:26Ashland's Space Station, but it's also capable of being

17:29a general science delivery platform

17:32to anywhere in the solar system.

17:34So where are you going with it?

17:37We're gonna send one to Mars in 2018.

17:40Now, let's talk about 2018, that's for sure.

17:43Yeah, a couple years.

17:44Now, will you be on that flight?

17:47No.

17:48You have talked about this, you said you don't want it,

17:50You want to die on Mars just not on landing, right?

17:52Is that correct?

17:54Well, I mean, I think if you're going to choose a place to die,

17:57then Mars is probably not a bad choice.

18:01All right.

18:03But you're not ready.

18:04It's not some sort of Martian death wish or something.

18:07But yeah, I mean, you're going to be born on Earth,

18:10die on Mars.

18:11So sending this up to Mars 2018.

18:14Sending this up to Mars 2018, when

18:18will someone like you get there from your plans?

18:21Sure, so the 2018 mission would be a Dragon version two.

18:26And that I wouldn't recommend traveling to Mars in that.

18:32Because it has the interior volume of a large SUV.

18:35OK.

18:36The trip for a dragon would be on the order of six months.

18:40It's a long time to spend in an SUV, I think.

18:42Mm-hmm.

18:43It can be done, but not probably not ideal.

18:51And it also doesn't have the capability of getting back to us.

18:57That seems more important than the space of...

18:59Yeah, we can put that in the fine print.

19:01Yeah.

19:02Sorry.

19:03It's like the side effects in a drug ad.

19:05By the way, we cannot get back to Earth.

19:08Yeah.

19:09We saw the movie.

19:10We saw what happened.

19:11He got back.

19:12Yeah.

19:14So I actually enjoyed the movie.

19:17So do you think he could have gotten back like that?

19:19There was that plausible?

19:19Well, I thought there was some connection.

19:22It was mostly, it was like 80% scientifically correct.

19:27They did connect a series of improbable events.

19:29Such as?

19:31Well, I mean, I don't think you can sort of just

19:33take off from Mars on an unguided rocket, really.

19:37And then hurt your finger on the space suit

19:40and navigate to a space ship.

19:41Right?

19:42Yeah.

19:43Not impossible, just extremely unlikely.

19:49So the standard volume.

19:51But if you're Matt Damon, maybe?

19:53Maybe.

19:54You have some mad skills, for sure.

19:57So when will people like yourself get there?

20:00And I assume you'll be first in line for that.

20:02Yeah, so I later this year in September at the IAC, which is the big sort of world space

20:10conference, industry-based conference, I'm going to be presenting the architecture for

20:16Mars colonization.

20:19So I think what really matters is being able to transport large numbers of people and ultimately

20:26millions of tons of cargo to Mars.

20:29And that's what's necessary in order to create a self-sustaining, not really self-sustaining,

20:36but a growing city on Mars.

20:40I'm curious. Have you bent a space yet? No. Why?

20:43And you could just go up right for a little bit or not. I could I suppose yeah why haven't you like walk around the world something or

20:51Yeah, probably will well you do a moon test before you go to Mars

20:57Yeah, I've never like probably I'm gonna go to orbit in four or five years

21:00Is it like that but again space and orbit are very different things, but on the Mars thing would you send up?

21:07two or three

21:08Whether it's you or not, I kind of would prefer it if you tried it frankly, but

21:15Because it would be exciting but would you send up some people before you do this whole architecture for

21:23Colonizing Mars just a handful of people to kind of see what I mean the basic game plan is like we're gonna send

21:31Emission to Mars with every Mars opportunity from 2018 onwards, so and they occur approximately every 26 months

21:38So, you know, we're establishing cargo flights to Mars that people can count on for cargo.

21:49And it's like I said, the Earth Mars orbital rendezvous is only every 26 months.

21:53So it's one in 2018, that would be another one in 2020.

21:57And I think if things go according to plan, we should be able to launch people probably

22:08in 2024 with arrival in 2025.

22:11Is that a more certain schedule than United Airlines?

22:18Well, I don't know.

22:22Well, he's been on that idea.

22:24Certainly he's associated with that.

22:27So, I'm going to show...

22:29Anyway, that's the game plan.

22:30Like, approximately 2024 to launch the first of the Mars on your transport system, where

22:39I want to get back to what you said earlier about a multi-million.

22:41This will be a very big rocket.

22:43Okay.

22:44Very big, bigger than the Saturn fire.

22:46Yes.

22:47Twice as big or what?

22:49September.

22:50I'll tell you.

22:51Not going to say anything until September?

22:53Come on.

22:54Very big.

22:55Come on.

22:56Has to be very big.

22:57How big is very big?

22:59So big.

23:05Do you think we should abandon the Earth at something?

23:07No.

23:08No, I think that's great.

23:09But you have said things.

23:12Oh, do you have an Earth?

23:13It's really nice here.

23:14You've said things about we may have to abandon the Earth,

23:17so it's good to have a Plan B instead of the Earth.

23:19No, that was amazing.

23:21That's amazing.

23:22I think that's amazing.

23:23I don't know, but it wasn't me.

23:25All right, OK.

23:26It wasn't me, like Shaggy.

23:28So let's move to things on this Earth.

23:31Let's move to Hyperloop Tesla, other things.

23:34Let's talk about Tesla first.

23:37Where do you feel like the company is at at this point?

23:40And there's been lots of activity in self-driving cars,

23:43autonomous, autonomous.

23:45How do you look at how everybody's jumped in?

23:48Google, Apple, others.

23:50And all the car manufacturers.

23:54Yeah, I mean, there have been so many announcements of like

23:56autonomous EV startups.

23:57I'm waiting for my mom to announce one.

23:59It's like mom, you two.

24:06I mean, there's a lot.

24:07So yeah, I mean in the US alone, there are four, I think maybe five China-funded EV startups

24:20at the billion dollar plus level, like, serious funding.

24:26And there's a bunch of startups and then of course the car industry as a whole seems to

24:31be moving in that direction.

24:34Volkswagen just I think announced a huge battery factory that they're going to build.

24:39And I think these are all good.

24:41It's good for the industry to be moving towards sustainable transport as quickly as possible.

24:48We open-source our patents to try to be helpful in that regard.

24:51And it's encouraging to see all this activity.

24:56From a Tesla standpoint, we want to take the set of actions that are likely to accelerate

25:06the event of sustainable energy.

25:08So scale up production as fast as we can.

25:13So we accelerated plans for the Model 3 by two years.

25:19So we want to try to get to half a million cars in total in kind of the 2018 timeframe,

25:24which is an aggressive schedule, but I think achievable.

25:29And there may be a million cars a year by 2020.

25:34And, you know, I can see it, like I think a pretty clear path to get there.

25:40Autonomy is obviously extremely important.

25:43People are going to want to autonomy.

25:46It's going to be odd to have a car without autonomy in the future.

25:52So I think that's where it's scaling up.

25:56How do you look at all these efforts, not your mother?

26:00But she's not going to do it.

26:02She may do a rocket situation.

26:04But how do you look at each of them?

26:06Let's go through them.

26:07What Google is doing?

26:09How do you assess what they're doing when you're looking at it?

26:12Because there'll be competitors at some point.

26:13These are all eventual competitors.

26:17What Google's done a great job of showing

26:22the potential of autonomous transport.

26:26But they're not a car company, so they would potentially

26:30license their technology to other car companies.

26:32And I think they announced something with the Fiat.

26:36And so I wouldn't say, you know, Google's a competitor,

26:40because they're not a car company.

26:42We would compete with somebody, perhaps,

26:43with their license technology, too, but not to them directly.

26:47Apple?

26:50Yeah, that'll be more direct.

26:52That'll be more direct?

26:53Yeah.

26:54You can tell by the hiring pattern and that kind of stuff.

26:58So what are you, OK, so they're going to be more direct.

27:00I'm going to hire a lot of people.

27:01How do you assess it?

27:06I mean, I say like, I think it's great that they're doing this.

27:10And I hope they have a work style.

27:20What's the time frame from them, do you think?

27:23I don't know.

27:24I think they should have embarked on this project sooner,

27:31actually.

27:38But I don't know when they have, they don't share with me

27:42the details of their production plans.

27:45But I don't think it's going to be,

27:48I don't think there'll be an evolving production

27:50sooner than maybe 2020.

27:54I don't be like the soonest.

27:55And that's, is that too late?

27:56We say they should have embarked sooner.

27:58Is that because 2020 would be too late to stop you

28:01or beat you or beat with you?

28:03What?

28:04It's just like it's a missed opportunity.

28:05It's just that they, it's, it'll be over by 2022.

28:11No, no, it's just like it's a couple of years.

28:15I think they'll probably make a good car

28:17and probably be successful.

28:19The car industry is very big.

28:20So it's not as though there's, you know,

28:23one company to the exclusion of others,

28:27and there's like a dozen car companies in the world

28:30of significance.

28:31So the most that any company has is approximately 10%

28:37market share.

28:39So it's not like somebody comes up with a car

28:42and they're suddenly like, they kill everyone else.

28:44It's not that way.

28:47And the sheer scale of automotive manufacturing

28:50is just hard to appreciate until you see the plants.

28:54I mean, they're gigantic.

28:56Yeah, I have seen the plants.

28:58Yeah, I mean, the sheer size of the industrial infrastructure

29:01is staggered.

29:03Not just the assembly plant, but everything else that goes.

29:05Yeah, the supply chain, exactly.

29:06The assembly plant is just a little tip of the iceberg,

29:09really.

29:09The assembly plant is literally a tip of the iceberg.

29:11Phenomenal.

29:12The supply chain is, once you go to tier two, tier three,

29:17two or four suppliers.

29:19That's probably an order of magnitude more than the final assembly.

29:24So you think Google will not be a competitor after probably

29:27will be a direct competitor?

29:29Yeah, sure.

29:30What about the car companies?

29:33I think they'll be competitors.

29:35Yeah, sure.

29:36Who do you see out there that has done a nice job so far?

29:39Mercedes or a flop?

29:41G&M?

29:41Of what?

29:42A competitive car.

29:44Of the incumbents.

29:46potentially competitive car.

29:50I mean, I don't think anyone's, any of the car companies

29:54thus far have made a really great electric car.

29:58I mean, you tell me if you disagree.

30:00But I don't think yet that any of them

30:04have made a great electric car.

30:06They presumably will continue to improve

30:09on what they've done so far.

30:10And then at some point, they may make a car that's a great car.

30:17They haven't done that yet.

30:18Can I ask you about batteries for a second?

30:21Oh, yeah sure.

30:22So you're building this gigafactory, right?

30:24It's built.

30:25Well, it's not completely built.

30:27OK, but part of it's up and running.

30:29A big truck of it's built, yeah.

30:30Part of it's up and running.

30:31It's a really gigantic thing.

30:33Like when the gigafactory's done, it'll be the largest

30:35footprint building of any kind in the world.

30:38Of any kind, not just factories, it literally of any kind.

30:40What is this?

30:40The largest rocket, the largest building.

30:42I mean, I think it's not scale for scale sake.

30:46It's just like if you say, well, we want to accomplish these goals, then you kind of have

30:51to make a big thing.

30:54Okay.

30:55You've got this big thing.

30:57It's this big giant building.

30:59It's going to make batteries.

31:00The batteries is going to make our lithium.

31:03We're going to have an opening.

31:04Well, it's not taking an opening party since it's been operating for a little while, but

31:07we're going to have a party soon.

31:08You guys maybe want to come.

31:09Okay.

31:10All right.

31:11It's good.

31:12We'll come to the battery party.

31:13I love it.

31:14Right just this is crazy. No, this is like an alien dreadnought. It's really nutty because I love a battery party, but yeah

31:22Right, but but but talk about where it's got these lithium ion batteries. Yeah, sure so they're the same batteries that's in our phone. No

31:30Explain please explain. Yes

31:32So let me have you made a battery breakthrough is something I'm interested in

31:38Yeah, I mean generally the I mean there's there's so much nonsense out there about batteries like about you can believe about 1% of what you

31:44read on a, you know, maybe.

31:49Lithium ion covers a very broad range of technologies.

31:54And you can have an enormous difference in the power density

31:58and the energy density and the cycle life between one

32:04chemistry and another.

32:05They can be really enormously different.

32:08So what you really actually want to ask is what is the cathode

32:12cathode and what is the anode?

32:14Right.

32:16So in our case...

32:17That's right.

32:18Okay.

32:19I just put it in the...

32:21The lithium is actually 2% of the cell mass.

32:23So it's like the salt in the salad.

32:25It's a very small amount of the cell mass and a fairly small amount of the cost.

32:33It sounds like it's big because it's called lithium ion, but it really...

32:37Our battery should be called nickel graphite,

32:40because it's mostly nickel and graphite.

32:43OK.

32:44It's nickel, cobalt, aluminum.

32:46But battery.

32:46You know the little things in graphite

32:48with a silicon oxide layer.

32:49Battery efficiency or power, the power

32:53that you can store in a certain mass

32:55seems to be moved very slowly, at least

32:59compared to where used to be Moore's law pushing.

33:02Integrated circuits faster.

33:04Batteries kind of are always in our consumer devices,

33:07always lagging behind.

33:09In your, you've built this giant thing,

33:11the biggest building in the world has ever seen.

33:13It's not fully built, but yes, it's.

33:15You're building it.

33:15Pretty big chunk is built, though, yeah.

33:17The biggest building in the world, yeah, yeah.

33:18To make batteries.

33:20Your whole business depends on batteries in these cars.

33:23Have you figured out a way to do some significant increase

33:29in the yield of energy from a given amount of space

33:34in the battery?

33:36Well, yeah, I mean, the energy density is increasing,

33:41sort of maybe on the order of like 5ish percent per year.

33:46And it doesn't sound like much, but you add that up

33:49over a number of years with compound interest,

33:50it ends up being quite a significant number.

33:55And a lot of people sort of think that, oh, we just sort

33:58of cobbled together some laptop batteries

34:00and somehow made a great car.

34:02And if it was that easy, then I

34:04think we would have quite a few competitors

34:05who did the same thing.

34:07But it's really quite a lot harder than that.

34:11It's a cylindrical form factor, but the internals of the

34:14battery are quite different from what you would find in

34:17Lata.

34:20And it will be increasingly different with what's built

34:23at the gigafactory, which is highly optimized for

34:25automotive and has improved energy density.

34:31But mostly, it's not the energy density that's the

34:33issue.

34:34You know, you can buy, if you buy a model yesterday,

34:37the range is around 300 miles.

34:42And that's quite a lot.

34:45So it's pretty rare that people really need to go

34:48more than 300 miles at a time without stopping.

34:51Right.

34:54So I don't think we really have a range issue.

34:56And we could make a 400 mile range car today.

34:59Like that wouldn't be too big of a deal.

35:01What really matters is decreasing the cost per unit of energy of the battery packs so

35:10you can make the car affordable.

35:11That's actually the important thing.

35:14So, there's really two main dimensions along which cost optimization and making something

35:21available to the National Market can be achieved.

35:23One is design iteration, going through multiple versions of something, and then the other is

35:27economies of scale.

35:29You kind of need both of those things in order to make a compelling mass market product.

35:34You look at cell phones and how many design iterations have we gone through with cell phones.

35:40And look at the scale at which they're made, which is enormous.

35:46And that's what enables everyone to have a supercomputer in their pocket.

35:51So speaking of that, the sales, when we're talking about the sales, you have booked how

35:55many orders for?

35:56It's on the order of 400,000.

35:58400,000.

35:59So obviously, consumer interest and a promise.

36:02A lot of it around you, around the idea of you and Tesla and the excitement.

36:06They're not booking.

36:07It was quite surprising actually.

36:10We didn't do any advertising or there was no guerrilla marketing or anything.

36:16It was just basically like, yeah, we're going to have this webcast.

36:19There were only about 1,000 people in the audience.

36:23And it really caught us by surprise.

36:25I think when you have a product that really resonates

36:29with customers, the word of mouth grows like wildfire.

36:34And that seems to be what it is.

36:35Yeah, but it's a little bit.

36:36I mean, honestly, in some groups, especially men

36:39in Silicon Valley, if you show up in red,

36:41like a label of a peanut jar, they'd

36:43be thrilled with the situation.

36:44So I mean, a lot of this does base around you,

36:48like the idea of you and the excitement around this exciting

36:51entrepreneur, is that enough to get it to this massive company

36:57you've been hoping to, the idea of this is the Elon promise?

37:01Well, I think actually it's not too much.

37:03I mean, I sort of, I'm not sure, I think I deserve less

37:10credit than that actually.

37:12The, I mean, what Tesla's done, we have a phenomenal team.

37:17It was like 15,000 people at the company,

37:20worked super hard to create compelling products,

37:22to create great cars.

37:24And we set up with the Roadster, and then the Model S.

37:26And the Model S was rated by Consumer Reports

37:28as the best car ever.

37:31The Model X, which has had some teething issues.

37:34But I think it's now at the point where it's really starting

37:38to, I think it's really, I think, quite sublime

37:43at this point.

37:46And so people look at that and say, OK, well,

37:49If Tesla's made these cars, then probably the next car they make is going to be, you know,

37:54the less expensive one.

37:56Also a great car.

37:57And yeah, so it'll be a great car but it'll be affordable and it's like great, okay, that

38:01sounds like something I want.

38:03So this car, this next car, the price is, the-

38:06It's starting at $35,000.

38:07Okay, affordable.

38:08Okay, when is-

38:09If you're eligible to this-

38:10When do you get to the point of-

38:11When do you get to the really affordable, the way down much lower than that?

38:15Yeah, I mean, it's important to point out that $35,000, particularly when factoring

38:18the lower cost of electricity versus gasoline,

38:21and that the maintenance cost is much less.

38:23You don't have to have oil changes.

38:25You never need to replace your brakes,

38:27because the car uses regenerative braking,

38:29so the brakes last as long as the car is.

38:32Basically, you just need to replace the tires,

38:34like that's about all.

38:37So the operational cost of the car

38:38is much lower fundamentally than a gasoline car.

38:41And so,

38:43and I think the average price for cars

38:47for guessing cars is around 30, 32, something like that.

38:52Yeah.

38:52I mean, there's starting prices that are lower,

38:54but when people pick options, I believe it's

38:56in around 32 or so in the US.

38:59So we're pretty close to that.

39:01But that's your base price, right?

39:02I mean, that would be a great car.

39:04It would be not be the ASP for the car.

39:06No, but it's going to be a great car, even at 35.

39:09So even if you order nothing, no options at all,

39:11it'll be great.

39:12But you're likely to have a mix where the average car

39:17that you actually sell, sells for a little more than that?

39:19Yes, probably, it's probably gonna be,

39:22yes, probably probably gonna be some higher number.

39:25But it's really important to have size like 35,

39:27if somebody orders a 35,000 car, they'll be very happy.

39:30Like it's not like you need to order a bunch of options

39:32in order without which the car is, you know, not good.

39:35That car will have autonomous for 35?

39:40I have a, I'm gonna do another Tesla event,

39:46Maybe at the end of the year, talk more about that.

39:50And say it.

39:51Well, so you could start here.

39:55It will be real big news if I start here.

39:58We don't mind that.

40:01Let me just say that we're going to do the obvious thing.

40:06OK.

40:07Got it.

40:08It's really obvious.

40:12So cup holders, good.

40:14OK.

40:15Brilliant.

40:18Those things are nuanced.

40:20Absolutely.

40:21Let's talk about two more things.

40:23I want to talk about AI because we've been talking about it a lot here.

40:26Which, I want to get it clear what your thoughts are.

40:30Because it's mostly Elon scared of robots.

40:32I mean, that kind of thing.

40:34How do you?

40:35I don't scared of robots.

40:37Artificial intelligence.

40:38Can you like clarify exactly what the issue you have now?

40:42You deserve the background.

40:43We've been talking to Jeff Bezos, Sundar Pichai.

40:51We talked to Mark Fields from Ford about it.

40:54My chat for you.

40:55Yeah, the Facebook folks.

40:57There certainly seems to be, in the tech companies,

41:02a big, tremendous new drive or interest to believing

41:07there will be all competing for intelligent assistance.

41:11It'll make your life better.

41:13It's going to suddenly get smart.

41:17Microsoft's one is going to get smart.

41:19Google is going to cream them all.

41:21It's largely a happy version of this.

41:23Sometimes technology hurts you, but not as much as it helps you.

41:26That's really.

41:27Yeah. So there's been a lot of conversation here about that.

41:30And you've staked out a slightly different position.

41:33So can you talk about that?

41:35Well, I think my full position would require quite a long explanation.

41:42I am concerned about certain directions that AI could take that would be not good for the future.

41:55I think it would be fair to say that not all AI futures are benign.

42:01And so if you have something, if we create some digital super intelligence that exceeds

42:08us in every way by a lot, it's very important that we have been in.

42:18So actually with a few others, I created OpenAI, which is an AI, it's a nonprofit actually.

42:29there's no, I think the governance structure here is important.

42:33So you want to make sure that there was not some fiduciary

42:35duty to generate profit off of the AI technology that's developed.

42:43So we created this 501c3.

42:48But it's quite different from a lot of 501c3s

42:55are, you know, they don't have a high sense of urgency.

43:00Like they're not like, you know, they're not really sort of

43:08developing technology at a fast pace.

43:11But OpenAI is.

43:12So OpenAI has a very high sense of urgency.

43:15I think that the people that have joined are really, really amazing.

43:22And the intent with OpenAI is to democratize AI power.

43:30There's a quote that I love from Lord Acton.

43:33He was the guy that came up with power corrupts

43:35and absolute power corrupts absolutely,

43:38which is that freedom consists of the distribution of power

43:41and despotism in its concentration.

43:44And so I think it's important, if we have this incredible

43:46power of AI that it not be concentrated in the hands

43:49a few and potentially lead to a world that we don't want.

43:53And what world is that?

43:54What do you see for see that when you sit?

43:57It's difficult, I mean it's called the singularity because it's difficult to predict what, exactly

44:03what future that might be except I don't know a lot of people who love the idea of living

44:08under a despot.

44:10You know I don't think people generally choose live in a democracy over a dictatorship.

44:16And the despot would be the computer?

44:19people controlling the computer.

44:21And do you worry specifically about any of these companies

44:26I mentioned, who've all seem to now be pivoting toward this?

44:31Is the battleground in the next 10 years?

44:34I wouldn't name a name, but there is only one.

44:37There's only one you're worried about.

44:40And they're not preoccupied with making a car that

44:42will compete with you, I assume.

44:47There's only one.

44:50And what is an interesting thing?

44:53It's not about competing.

44:55This is sort of like what we put at the point of competing

44:57for mutual destruction.

45:01It's not about competing.

45:02It's really just about trying to increase the probability

45:06that the future will be good.

45:09That's all.

45:10So the goal of OpenAI is really just

45:15to take the set of actions that are most likely

45:17to improve the positive futures.

45:20You can think of the future as a set of probability streams

45:24that branch out and then converge,

45:26collapse down to a particular event,

45:28and then branch out again.

45:30And there's a certain set of probability associated

45:33with the future being positive,

45:34and different type of believers of that.

45:36And at OpenAI, we want to try to do whatever we can

45:40to guide, to increase the probability

45:43of the good futures happening.

45:44I think that's really what we're trying to do there.

45:46worry that by making this open, some bad actors may use some of what has been developed to

45:53do bad stuff with the power of AI.

45:56Yeah, I mean that is certainly the good rebuttal to that.

46:01However, I think if AI power is widely distributed, then, and there is not, say, one entity that

46:10has some super AI that is a million times smarter than anything else.

46:16If instead the AI power is broadly distributed and to the degree that we can link AI power

46:23to each individual's will, it's like you would have your AI agent, you would, like everybody

46:30would have their sort of AI agent.

46:33Then if somebody did try to do something really terrible, well then the collective will

46:37of others could overcome that bad actor.

46:40Which you can't do if there's one AI that's a million times better than ever.

46:48And it's proprietary.

46:49And it's, yeah, it either has its own will or more likely at least in the beginning is

46:54controlled by some small set of people.

46:58So I think that's really the risk.

47:01I mean, you know, those are those means arguments.

47:04Like what's the best form of government?

47:06You know, big fan of, I think it was Churchill,

47:10like, you know, democracy is the worst form

47:12of government except for all the others.

47:13Right.

47:15So speaking of that, this election, you are from?

47:18Oh, no.

47:19No, no, yes, yes.

47:21How does that strike you what's happening now?

47:25You come to this country, you're naturally so.

47:27You know, I think I'm glad that the framework

47:30is the Constitution so fit to ensure that the president was

47:34someone who was captain of a large ship with a small rudder.

47:38OK.

47:40And?

47:42There's a limit to how much harm can you give him, President?

47:44Are you sure about that?

47:45Oh, yeah.

47:46Yeah.

47:47Yeah.

47:47So you're not worried about are you backing in either

47:50of the candidates at this point?

47:52Try to stay out of this situation.

47:54Because?

47:56I don't think that's the finest moment in our democracy.

47:59Well, given that it's not the finest moment in our democracy,

48:02do you think the best thing is to stay out?

48:05Or to get in?

48:06I'm not sure what I could do to the worst.

48:09I'm not sure how much influence I could have

48:11as one person on the outcome.

48:19I mean, if I think I could make a difference,

48:20I would probably do something.

48:22But like I said, I think I'm just glad that being

48:28the US president is like being captain of a large ship

48:30with a small rudder.

48:32And so that's just a limit to how much good or bad

48:35a president can actually do.

48:38Obviously, if a president could make the economy great,

48:42and there was a button he could press,

48:43they're repressing that button at the speed of light.

48:50But they can't.

48:52They can't just magically make the economy good.

48:55No president wants the economy bad ever.

48:57but they, you know, like there's just a limit

49:00to how much they can do.

49:03And yeah, I guess there is the nuclear thing,

49:08which is the nuclear thing.

49:09Yeah, the nuclear thing.

49:11I guess there is the nuclear thing.

49:13Yeah, but I don't know, I think you'll,

49:15I think I don't think we would like

49:17just arbitrarily long nuclear missiles.

49:19Yeah, one would help.

49:21President can do that.

49:22I don't think so.

49:23I don't think so.

49:23I mean, I think that.

49:25He's, no, he's the commander chief.

49:26I still don't think that means you can just launch

49:28nuclear missiles whenever you want.

49:30I think Congress would be quite upset about that.

49:34And they might not be consulted.

49:37Yeah, but I think the military would be like,

49:39yeah, we really think Congress should be consulted

49:41on before you launch a...

49:43That might happen.

49:44We have to be a strike.

49:45Are you willing?

49:46You're basing your faith in that, though.

49:48I'm quite confident that the military would not just,

49:51you know, randomly agree to launching

49:54your glen muscles at somebody.

49:56Well, that's calming.

49:58This is fine.

49:58Yeah.

49:59So we're going to put up just very quickly

50:01and on hyperloop.

50:03You've been involved with it.

50:04Your level involvement is what at this point?

50:06Yeah.

50:07Yeah.

50:07I know it's a bit confusing because I am.

50:12You talked about it when you were here last time.

50:14Yeah.

50:15I actually came up with the idea.

50:19I came up with an initial idea which would turn out

50:21to be wrong, it wouldn't work several years ago.

50:25And then, but I sort of shot my mouth off

50:30and said, I click, I have an idea that would work.

50:32And turned out that didn't work,

50:33but with a lot of iteration I was

50:35able to come up with something that

50:37where the physics hangs together.

50:39And then published the paper and just said, look,

50:41anyone who wants to do this is great.

50:43Be my guest.

50:46Because I sort of have a plateful running Tesla and SpaceX.

50:49Yeah, yeah.

50:50And so I think it'd be great to have any interesting new

50:56transport solutions, anything that gets people to their

50:59destination in a way that's safer, cost less, it's more

51:04convenient.

51:06That'd be great.

51:07And so I think probably the most valuable thing that the

51:11high-blue paper that I published has done is to spur

51:18thinking in terms of new transportation systems.

51:21So it's not just, oh, let's have a fast train.

51:28OK, that's not even as fast as what Japan did in the 80s.

51:31OK, well, why don't see what the point of that is?

51:35Like, we should really be trying to think of something

51:37that's, particularly in California,

51:40like we should be like saying, hey, what is the best?

51:42Let's invent something new that's way better than anything

51:46else.

51:47Do you want to shoot your mouth off about that?

51:53So I'm not an investor in any of the companies that are working on it.

51:57I've tried to be neutral because I'm trying not to favor one company over another.

52:01But just to encourage anyone that is interested to say that, try to give them moral support.

52:09And I hope they succeed.

52:12The only thing that I am doing at the HAPLU Fund is that we're holding a student competition

52:18and the student competition is really just aimed towards encouraging students to think

52:24about exciting new transport methods.

52:27And it's totally cool if they want to do some architecture that's different from what

52:33I propose in HAPLU.

52:34In fact, the winning team at the student competition that we held earlier this year used a different

52:40suspension mechanism than what I propose, which is I propose

52:44using essentially taking the air that pulls up on the nose

52:50from the compressor and flowing that through air skis.

52:55So you simultaneously remove the drag from the nose and

52:59provide a means of suspending the pod.

53:04And that's also something that works well even at super

53:08supersonic velocities, you can go, it's been demonstrated up to Mach 1.1 in terms of using

53:12air bearings as-

53:14But they use something different.

53:16Yeah, basically, electromagnetic suspension.

53:19And the reason I didn't suggest any kind of magnetic system suspension is that it's very

53:31important that the cost of the tube be minimized.

53:36Like you really want, because the pod is cheap, the tube is expensive.

53:41So if you want to go say 400 miles and tune two directions, so you've got 800 miles of

53:45tube, the critical economic optimization parameter is the cost of the tube.

53:52So you want that tube to be as low cost as possible.

53:55And so if you do anything that requires action on the tube side, it's going to make that

54:05too much more expensive. So if you use air bearings, it doesn't change. Like that's real cheap.

54:13And yeah. So.

54:15You think this is going to happen?

54:17Yeah, I think something like that. I think something will happen in the future. You know,

54:24I think if the companies that are trying to make it happen now, if for whatever reason

54:30that doesn't work out, then you know, I think I'll, you know, I might do something like

54:37in the future.

54:39I don't want to do something, I don't want to sort of front run them, you know, and say

54:43here's this free idea and then meanwhile I go and do it myself, you know, that wouldn't

54:46be nice.

54:49So but if they, if much of the companies don't try and it doesn't work out, then I think,

54:56I think I'll try to just at least do a demonstration system, yeah.

55:01Okay, last question.

55:03Do you think tech has gotten more serious?

55:05How do you look at the tech landscape as someone who's well known?

55:09You probably qualify as a visionary.

55:12The concept, what do you imagine we are right now in the tech space?

55:15And then we'll get to questions from the audience.

55:20I think there's a lot of innovation happening in many different areas.

55:24The advancements in AI are quite astonishing.

55:30The advancements in genetics are amazing.

55:34So I think that there is a lot of innovation going on.

55:43I think that there's probably a few too many talented entrepreneurs

55:48in kind of the internet space and I think their talent actually would be better served

55:54in some other industries.

55:55But I do think that, I mean I don't think we're like facing some sort

56:00of low innovation period or anything like that.

56:03I think there's a lot of innovation going on.

56:04they need to move to other.

56:06I just think that if you had some ideal distribution,

56:09it would probably be fewer.

56:11Like there's just a lot of talent focused on the internet,

56:13and probably some of that talent would be better

56:20to have some of that talent in other industries.

56:23That's about all.

56:24But there's a tremendous amount of innovation

56:27that's happening.

56:29It's something that I think is going to be quite important.

56:37And I don't know if a company that's working on it seriously

56:40is a neural lace.

56:48So going back to the AI situation,

56:53this is quite an important debate.

56:59If you assume any rate of advancement in AI,

57:03we will be left behind by a lot.

57:08And so then we could be in a benign situation.

57:13If you have ultra-intelligent AI,

57:21so far below the main intelligence

57:22that it would be like a pet.

57:25Like a cat.

57:26Like a cat.

57:27Like a cat.

57:28Like a cat.

57:29Like a cat.

57:30Like a cat.

57:31Like a cat.

57:32Yeah, we like the house cat.

57:33Right.

57:34And yeah, that's the end of the world.

57:35You know, sort of.

57:36Well, you've seen the movie.

57:38It could be.

57:39It could be.

57:40Could be.

57:41You know, so that, honestly, that would be the benign scenario.

57:45So house cat is okay.

57:48I mean, I don't love the idea of being a house cat.

57:52Okay.

57:53But what's the solution?

57:55Yeah, so I think it's two, essentially, I think one of the solutions.

58:05The solution that seems maybe the best one is to have an AI layer.

58:13If you think of your limbic system, your cortex, and then a digital layer,

58:19sort of a third layer above the cortex that could work well and symbiotically with you.

58:28I mean, just as your cortex works symbiotically with the olympic system, your sort of a third

58:33digital layer could work symbiotically with the rest.

58:36This is something that's surgically inserted or bred into the species or what?

58:44The fundamental limitation is input output.

58:46So we already have, we're already a cyborg.

58:52It's just that, I mean, you have a digital version

58:55of yourself, a partial version of yourself online

58:59in the form of your emails and your social media

59:01and all the things that you do.

59:05And you have basically superpowers

59:07in that with your computer and your phone

59:09and the applications that are there.

59:11You have more power than the president of the United States

59:14had 20 years ago.

59:15you can add to any question, you can video conference with anyone anywhere, you can send

59:22a message to millions of people instantly, you know, you just do incredible things.

59:29And, but the constraint is input output.

59:36So we're IO bound.

59:39Particularly output bound.

59:40I mean, like your output level is so low,

59:43it's like particularly on a phone,

59:45like you're two thumbs just sort of tapping away.

59:47This is ridiculously slow.

59:51Our input is much better

59:52because we have a high bandwidth visual interface

59:55to the brain.

59:56Like our eyes are taking a lot of data.

1:00:00So there's many orders of magnitude difference

1:00:01between input and output.

1:00:05So, mostly, effectively, merging in a symbiotic way with digital intelligence revolves around

1:00:18eliminating the I.O. constraint.

1:00:23So it's some sort of direct cortical interface.

1:00:26And you called it a neural...

1:00:29Neural-lace, yeah.

1:00:31It's totally not go over glass, right?

1:00:36No. I'm talking about something which...

1:00:38But did you wear it?

1:00:39No. I mean, it would be... I mean, there are a few ways to

1:00:46approach this, but some sort of interface directly with your

1:00:49cortical neurons, particularly.

1:00:51But doesn't that apply surgical insertion?

1:00:53Not necessarily. You could go through the veins and arteries

1:00:56because that provides a complete roadway to all of your neurons.

1:01:01Your neurons are very heavy users of energy, so they need high blood flow.

1:01:06So you automatically with your beans and arteries have a road network to your neurons.

1:01:11Still some kind of surgery, right?

1:01:14Yes, but you could insert something basically into the jugular and have...

1:01:22It gets in the car, but...

1:01:24It sounds really easy and...

1:01:26It doesn't involve, it doesn't involve, you know, like,

1:01:29chopping your skull off or anything like that.

1:01:32Yeah, and plus you're not a house cat anymore.

1:01:34Not a house cat, so.

1:01:36I mean, essentially, if we can figure out

1:01:39how to establish a high band with neural interface.

1:01:43With ourselves.

1:01:45With your digital self, effectively.

1:01:50Then you're no longer a house cat.

1:01:54All right.

1:01:55No, I just one closing thing.

1:01:59I mean, I think that's probably the best outcome, I think.

1:02:03Are you interested in exploring this possibility

1:02:08that you have just laid out?

1:02:09Somebody's got to do it.

1:02:12I'm not saying that I will, but somebody's got to do it.

1:02:15I mean, somebody should do it.

1:02:18And if somebody doesn't do it, then I think I should probably

1:02:20do it.

1:02:22And the goal of this is to prevent

1:02:24there being an external AI, particularly one controlled by a small group of people that could

1:02:32be so much more powerful and intelligent than we are that they have to get.

1:02:36It would be a god-like situation.

1:02:38Situition.

1:02:39This has been really cheerful.

1:02:41Thank you.

1:02:42But if we can establish that.

1:02:44I was worried about asteroids at the beginning of this.

1:02:47I mean asteroids are a low probability existential threat on the time scale that's relevant to

1:02:52us.

1:02:53Okay.

1:02:54Okay, this is different.

1:02:56This requires urgency.

1:02:57So what do you do for fun?

1:02:58Yes, this is much more energy.

1:02:59What do you do for fun, fun?

1:03:01What do you do, anything?

1:03:03I play video games with my kids.

1:03:04All right, that sounds good.

1:03:05Let's get some questions.

1:03:06Elon, Elon the house cat.

1:03:09Much movies, you know, I think.

1:03:10Almost like that.

1:03:11Why don't we start over here?

1:03:13Yeah, hi.

1:03:13I think this last question just did that.

1:03:18I want to know how do you live through this stress

1:03:21that kind of conversations we just heard

1:03:23that you went through and kind of ambitions that you carry

1:03:27and then how do you adjust to the everyday work,

1:03:31life, balance, et cetera, things that in your life.

1:03:33So a little bit of your personal side, actually.

1:03:35You're very busy.

1:03:37How does that work?

1:03:38Yeah, I mean, I am sort of in kind of work triage mode

1:03:42a lot of the time.

1:03:43So I know it seems to be, as long as there's not

1:03:47like a crisis simultaneously at SpaceX and Tesla,

1:03:50it's OK.

1:03:52But companies are, I mean, the situation in any given company,

1:03:56particularly if it's sort of growing fast and sort of

1:03:59quasi-startup, it's somewhat sinusoidal.

1:04:01So I mean, it's okay if the waves don't crest together.

1:04:08When that does happen, then that's a huge strain.

1:04:12Right now things are like, you know,

1:04:14motoring along okay.

1:04:16And I have like the context loaded for both companies,

1:04:20and I can look sort of see a path to a good outcome.

1:04:23So I feel pretty good right now.

1:04:26But they've been super stressful times in the past.

1:04:33And then I always try to reserve time for my kids

1:04:36because I love hanging out with them.

1:04:37Like, I mean, kids are really great.

1:04:39I mean, they're like the 99% of the time

1:04:44they make you happier.

1:04:46Their kids are awesome.

1:04:47Yeah.

1:04:49Then there's that 1%.

1:04:501%, you know, like, yeah, 1%.

1:04:52But like, it's of anything in my life,

1:04:55I would say kids by far make me the happiest.

1:04:58I don't know.

1:04:59I agree.

1:05:00Yeah, that's great.

1:05:01I agree.

1:05:02Yeah.

1:05:03That's true.

1:05:04So hang out with them.

1:05:05So like, no, like things like, a lot of times kids

1:05:07are kind of in their own world, so you don't need to like,

1:05:10they don't want to like talk to their dad for hours

1:05:11and end, generally.

1:05:12I've noticed that.

1:05:13Yeah.

1:05:15So like, so I can be in the same room with them.

1:05:17They can talk to me from time to time.

1:05:18But I can get some emails done, get some work done,

1:05:22and then whenever they want to talk to me, they can.

1:05:25And then we try to do things like travel places.

1:05:32And actually, we play video games together.

1:05:36Or actually, on Monday, we went to the new Harry Potter

1:05:40land at Universal, which was quite fun.

1:05:45So I think that that-

1:05:46Nobody from the university just clap.

1:05:48Yeah.

1:05:49Whoever was in charge of Harry Puddland did a great job.

1:05:51It's really good.

1:05:52Yeah, it is good.

1:05:53I highly recommend it.

1:05:54Yeah.

1:05:55The Butterbeer is amazing.

1:05:56Yeah, I was just saying the Butterbeer is amazing.

1:05:58Yeah.

1:05:59Okay.

1:06:00Right here.

1:06:01Over here.

1:06:02Hi, my name is Evan Burns and the founder of Odyssey.

1:06:03And I hope into the future to be in something in the space industry.

1:06:07And my curiosity is you've talked about SpaceX getting in many different businesses.

1:06:11For example, Global Wi-Fi, through launching many satellites.

1:06:15Do you hope SpaceX becomes a platform for others to launch businesses, or do you see SpaceX

1:06:19being a business that launches many business lines?

1:06:22Well, I mean the general strategy of SpaceX is to, like we clearly need a lot of money

1:06:30in order to develop the transport system to establish a city on Mars.

1:06:35So we're like kind of gathering revenue, like Earth-based revenue that's, we're trying

1:06:47to maximize Earth.

1:06:48Is there some other revenues?

1:06:49Well, right now it's only Earth.

1:06:51So we've got to maximize Earth revenue as it relates to space, as it relates to rockets

1:06:55and spacecraft.

1:06:58But I think what, assuming SpaceX is able to transport large numbers of people and goods

1:07:03to Mars.

1:07:04an enormous enabler for entrepreneurial activity on Mars because there's going to be so much

1:07:10to do. Everything from creating the first iron ore refinery to the first pizza joint to something

1:07:18that doesn't even exist on Earth. It's kind of like when the Union Pacific crossed, everybody

1:07:24thought, what a stupid idea. There's nobody living in California. Well, okay, now there's

1:07:32quite a lot of people living in California. So just having, you need the transport link.

1:07:40And so what SpaceX is trying to do is establish a transport link and then try to create a fertile

1:07:45environment for entrepreneurs on Mars to flourish. And I think that will be an amazing

1:07:53expansion of entrepreneurial life.

1:07:54How long would it take to deliver a pizza from Mars?

1:07:57Well, it's going to be a little cold, I think.

1:08:00But, I mean, we can certainly see a way to get to Mars in under three months.

1:08:07And I think ultimately you'll be able to get to Mars in under a month.

1:08:10It does get exponentially difficult as you reduce the time.

1:08:14But three months is a way to think of it.

1:08:18And I think that's probably, you know, that's really where SpaceX will, I think, create a

1:08:23a great environment for entrepreneurial potential.

1:08:27Thanks.

1:08:29Nila?

1:08:31I hope Domino's does not get to Mars.

1:08:34Please don't let it.

1:08:34But I was going to fill up special Mars.

1:08:36I said what it would be.

1:08:37I'm assuming you're the king of Mars.

1:08:38Nila.

1:08:41So you're obviously very ambitious.

1:08:43That's led to some really ambitious deadlines that have been missed.

1:08:47So Falcon Heavy was originally 2012.

1:08:50The Model X was a little bit delayed.

1:08:52the Model 3, the Model X was delayed, the Model 3 seems to be stretching, but the Model

1:08:563 in particular is a consumer product.

1:08:59You're taking money from people against a really aggressive production schedule and

1:09:05a huge amount of orders.

1:09:06What are you going to do to hit your deadlines on that because it's real consumers this time

1:09:12in a big class of people?

1:09:15I think the biggest thing is just designing the car for manufacturing.

1:09:24So in the case of Model S, like the Model S was the first time we'd really built a car,

1:09:29a whole car.

1:09:30Like with the roads to Lotus, to the body and chassis, we did the powertrain, then we did

1:09:34the sort of final installation of the powertrain to the chassis.

1:09:36But the Model S was the first time we made a car.

1:09:39So we were just trying to make a great car, but we had no idea what it meant to design

1:09:44something to be manufacturable. So the Model S is super hard to make and then the Model X

1:09:49is both of the Model S platform except it's got a bunch of other WIS bank technologies

1:09:53that make it even harder to build. So, you know, so, like, I mean, definitely we want

1:10:03to do the opposite of what we do with the X, which is make something that is going to

1:10:08be a lot simpler but still a car that people will love and where every design decision

1:10:16is factoring in the manufacturability and making sure that when we design something that you

1:10:23can manufacture at volume at an affordable price on the schedule that we're targeting.

1:10:33One of the things that makes a car very difficult particularly if it's a new car is that it's

1:10:37an integrated product with several thousand unique components. So we are somewhat at the

1:10:43mercy of whatever the slowest component is, whatever basically, if you say go to tier

1:10:49two and three suppliers, they end up being several thousand suppliers. So things move

1:10:53as fast as the least lucky and least competent supplier. But just, and you can think of

1:11:01like any natural disaster you care to name, all of those things have happened to our suppliers.

1:11:06If factory has burnt down, there's been an earthquake,

1:11:08there's been a tsunami, there's been massive hail,

1:11:14there's been a tornado, the ship sank,

1:11:18there was a shoot out of the Mexican border,

1:11:21no kidding, that's laid trunk carpet at one point.

1:11:27And we couldn't get in like the border patrol

1:11:30wouldn't give us the truck,

1:11:32because it had a bullet holes in it.

1:11:33We just wanted our trunk carpet.

1:11:38Like it's pretty safe.

1:11:39It's like no cocaine or anything.

1:11:41It's good.

1:11:42But you know, that's shut down the brush line as an example for several days.

1:11:48So that's the biggest issue.

1:11:50Like the supply chain stuff is really tricky.

1:11:53We're trying to anticipate as much of that as possible, increase our optionality so that

1:12:00there's more internal capability at Tesla.

1:12:02that we want to do things internally, but if a supplier is unwilling to deliver the part,

1:12:10we can quickly make that internally.

1:12:14So I think the whole company is geared for that.

1:12:20And I mean, right now it looks like we should be able to do that.

1:12:26We expect to, I mean, almost all of the Model 3 design is done, and we're aiming for pencils

1:12:34down basically in about six weeks, complete pencils down.

1:12:42And we're tabling all, you know, like if there are ideas for future cool things, we'll have

1:12:48it in version 2, version 3, in our future years type of thing.

1:12:50So overall, I feel pretty good about it.

1:12:55And I particularly am major.

1:12:58So a lot of partners have been very supportive and are on board.

1:13:08I mean, I should say, when I sort of cite a schedule,

1:13:14it is actually the schedule I think is true.

1:13:17It's not some fake schedule that I don't think is true.

1:13:20So, I mean, you know, it's never, you know, it's maybe delusional.

1:13:30That is entirely, you know, possible news happen from time to time.

1:13:36But it's never, you know, some knowingly fake deadline ever.

1:13:42So, is there an event in six weeks where you're going to announce autonomous driving is included

1:13:47in the pencils down plan for the Model 3?

1:13:55not expecting any event in six weeks.

1:13:59Josh.

1:14:01Hi.

1:14:03This is kind of a weird question.

1:14:04I feel like you would be the guy with the right answer for it.

1:14:07There's sort of a philosophic concept

1:14:11that a sufficiently advanced civilization

1:14:13will be able to create a simulation.

1:14:17Yeah, maybe you've answered this before?

1:14:19A simulation.

1:14:20I've had so many simulation discussions.

1:14:21It's crazy.

1:14:22OK.

1:14:23Because in fact, it got to the point

1:14:25where basically every conversation was the AI slash simulation

1:14:29conversation.

1:14:30And my brother and I finally agreed

1:14:32that we would ban such conversations

1:14:34if we were ever in a hot tub.

1:14:36That was like, because that really kills the matter.

1:14:39Or not a tub.

1:14:41So the idea is, right, any sufficiently advanced

1:14:44civilization could create a simulation that's

1:14:47like our existence.

1:14:49And so the theory follows that maybe we're in the simulation.

1:14:54Have you thought about this and a lot?

1:14:56Are we?

1:14:57Are we?

1:14:58Even in hot tub.

1:15:00So much so it had to be banned from a hot tub.

1:15:04Okay.

1:15:05It's not the sexiest conversation.

1:15:08Are we in?

1:15:09Are we in?

1:15:10I think here's, in my mind, the strongest argument for us being in a simulation, probably

1:15:18being in a simulation, I think is the following.

1:15:22But 40 years ago, we had Pong, like two rectangles and a dot.

1:15:28That was what games were.

1:15:31Now 40 years later, we have photo-realistic 3D simulations with millions of people playing

1:15:36simultaneously and it's getting better every year.

1:15:39And soon we'll have virtual reality, augmented reality.

1:15:45If you assume any rate of improvement at all, then the games will become indistinguishable

1:15:52from reality. Just instinctable. Even if that rate of

1:15:56advancement drops by a thousand from what it is right now, then you just say, okay,

1:16:02well, let's imagine it's a ten thousand years in the future, which is

1:16:06nothing in the evolutionary scale. So, given that we're clearly on our

1:16:14trajectory to have games that are indistinguishable from reality and those

1:16:18games could be played on any set-top box or on a PC or whatever and they would probably

1:16:23be billions of such computers or set-top boxes, it would seem to follow that the odds that we're

1:16:36in base reality is one in billions.

1:16:41Tell me what's wrong with that argument.

1:16:42Is the answer yes.

1:16:43The argument is probably.

1:16:45I mean, I just like, is there a flaw in that argument?

1:16:48I mean, but someone, I'm not sure what the error is.

1:16:50No, no, the argument makes sense.

1:16:52So the assumption then is that somebody beat us to it.

1:16:56And this is a game.

1:16:58No, no, there's a one in billions chance

1:17:00that this is based reality.

1:17:01Oh, OK.

1:17:03What do you think?

1:17:05Well, I think it's one in billions.

1:17:06OK.

1:17:08That seems to be clearly what it suggests.

1:17:15And actually, arguably we should hope that that's true, because otherwise if civilization

1:17:21stops advancing, then that may be due to some calamities event that erase the civilization.

1:17:28So maybe we should be hopeful that this is a simulation, because otherwise...

1:17:32Because they could reboot it.

1:17:35Otherwise, either we're going to create simulations that are

1:17:39indistinguishable from reality or civilizational cease to

1:17:42exist.

1:17:42Those are the two options.

1:17:44I like those odds.

1:17:47Okay.

1:17:49It's unlikely to go into some multimillionaire stasis.

1:17:53So it's either going to increase or decrease.

1:17:56Hi, I'm G.T.Patel from Box.

1:17:58Two part question for you.

1:17:59One is, if you think about fully autonomous vehicles,

1:18:04of pass through regulatory approvals, have passed through in-city driving and traffic

1:18:10conditions, how far do you think from a time frame perspective we are for that becoming reality?

1:18:16And number two would be the second part of that question is how far before, how long before

1:18:22you think it's either illegal or extremely prohibitively expensive for humans to drive

1:18:27on the road?

1:18:31Well, I mean, I think I really would consider autonomous driving to be basically a solved problem.

1:18:41Even in cities like Beijing and...

1:18:43Yeah, yeah, actually this...

1:18:46There's really only one area where it's like a little dodgy,

1:18:51and that's basically if you're at roughly 30 to 40 miles an hour

1:18:56in open environments, which is that's difficult to achieve in

1:19:00Beijing.

1:19:02It's like heavy traffic.

1:19:03In dense traffic situations, autonomy is really easy,

1:19:08because you can just maintain a set distance from various

1:19:10cars.

1:19:11It's actually quite, quite easy.

1:19:13You aren't very unlikely to run anyone over.

1:19:15You're just not moving fast enough, and you can break in

1:19:17time.

1:19:18On highways, particularly highways that are barriers, so

1:19:24that you don't have pedestrians.

1:19:25That's also relatively easy, and like a Model S and Model X,

1:19:29at this point, can drive autonomously

1:19:31with greater safety than a person right now.

1:19:34That's a question.

1:19:35My point is when does it get to be where you don't need

1:19:38to be sitting behind a vehicle, and actually the way

1:19:41that society starts expecting this is?

1:19:43I can have my 75-year-old mother, who

1:19:46doesn't speak any English or doesn't drive,

1:19:48be able to be transported from point A to point B

1:19:51by just sitting in a car by herself and being taken.

1:19:53I know it's technically possible,

1:19:55But how far do you think the regulatory approvals are

1:19:57for that happening?

1:19:59I think we're basically less than two years away

1:20:06from a complete autonomy.

1:20:08Well, that.

1:20:09Complete.

1:20:09Safely than a human.

1:20:11However, regulators will take, I think, at least another year,

1:20:15at least another year.

1:20:17And to pick, is it going to depend on which part of the world

1:20:20you're in?

1:20:22Because they will want to see billions of miles

1:20:24of data to show that it is statistically true

1:20:28that there is a substantial improvement in safety

1:20:31if something's autonomous versus not autonomous.

1:20:33I don't think that regulators will accept something

1:20:35that's close to, that's sort of approximately

1:20:39as good as a person.

1:20:40I think they'll have to be at least twice

1:20:42as good as a person, maybe five or 10 times,

1:20:47you know, better in terms of safety.

1:20:51And that will have to be a statistically

1:20:54relevant data set. So like billions of miles over widely differing roads and

1:21:00situations. So yeah, you know, I think it's like probably three years before

1:21:10it's from a regulatory standpoint but less than two before it is technically

1:21:15possible. And do you think there's a day when it's illegal to drive for humans or, you know?

1:21:20Well I mean we live in a democracy so presumably that would be a function of

1:21:24of the population deciding.

1:21:29I mean, I'm not in favor of banning people

1:21:32from driving cars, like I'm in favor of freedom

1:21:37and not restricting what people do.

1:21:43Yeah, but maybe the requirements for a license

1:21:46will get more stringent.

1:21:47I think that seems like maybe a good move,

1:21:50so you have to demonstrate a higher level of skill to drive

1:21:53in order to be allowed to manually drive.

1:21:56Very last question.

1:21:57This is the last question.

1:21:58I'll make it a good one.

1:21:59But Elon has to go.

1:22:00I'm going to make it a great question.

1:22:03Thinking about life on Mars again, how do you think about cultural unification, systems

1:22:10of government, rules of law, establishing those very early on?

1:22:15Well, I think I just declared King of Ma's a moment ago.

1:22:19Yeah.

1:22:20I like that.

1:22:21Yeah.

1:22:22Take it.

1:22:23Yeah, thank you.

1:22:24Thank you.

1:22:25So I think most likely the form of government en masse

1:22:31would be a direct democracy, not representative.

1:22:35So it would be people voting directly on issues.

1:22:40And I think that's probably better because the potential

1:22:44for corruption is substantially diminished

1:22:46in a direct versus a representative of democracy.

1:22:51So I think that's probably what will occur.

1:22:55I think that some, I think, I would recommend like some adjustment for the inertia of laws

1:23:03is what would be wise in that it should probably be easier to remove a law than create one.

1:23:11I think, you know, this is, I would just be like, let's just, I mean, I think that's probably

1:23:17good because just laws have infinite life unless they're taken away.

1:23:22So I think my recommendation would be like something like, let's say, 60% of people need

1:23:28to vote in a law, but at any point greater than 40% of people can remove it.

1:23:34And any law should come with a built-in sunset provision.

1:23:37If it's not good enough to be voted back in, maybe it shouldn't be there.

1:23:42So that's the framework for the government on Mars.

1:23:46I mean, those would be my recommendations.

1:23:48is a direct democracy where it's slightly harder

1:23:52to put laws in place than to take them away

1:23:57and where laws don't just automatically live forever.

1:24:00You'll be a good king.

1:24:02Thank you.

1:24:03Thank you.

1:24:03Thank you.

1:24:04Thank you.

1:24:06Thank you.