Elon Musk: SpaceX, Mars, Tesla Autopilot, Self-Driving, Robotics, and AI | Lex Fridman Podcast #252

thumbnail
0:00-0:73

the following is a conversation with elon musk his third time on this the lex friedman podcast yeah make yourself comfortable oh no wow okay no no you don't do the headphone thing no okay i mean how close do i get need to get this thing the closer you are the sexier you sound hey babe yeah can't get enough of the what y'all that baby i'm gonna clip that out anytime somebody messages me about it and you think i'm sexy come right out and tell me so so good okay serious mode activate all right mode come on you're russian you can be

0:73-1:28

serious everyone's serious all the time in russia yeah yeah we'll get there we'll get there yeah it's gotten soft allow me to say that the spacex launch of human beings to orbit on may 30th 2020 was seen by many as the first step in a new era of human space exploration these human space flight missions were a beacon of hope to me and to millions over the past two years is our world has been going through one of the most difficult periods in recent human history

1:28-1:87

we saw we see the rise of division fear cynicism and the loss of common humanity right when it is needed most so first elon let me say thank you for giving the world hope and reason to be excited about the future oh it's kind of you to say it i do want to do that humanity has uh obviously a lot of issues and and uh you know people at times do do bad things but you know despite all that um you know i love humanity and i think we should uh make sure we do everything we can to have a good future and an exciting

1:87-2:46

future and one where that maximizes the happiness of the people let me ask about uh crew dragon demo two so that that first flight with humans on board how did you feel leading up to that launch were you scared are you excited what's going through your mind so much was at stake yeah no that was extremely stressful no question we obviously could not um let them down in any way um so extremely stressful i'd say uh to say the least but we did

2:46-3:06

i was confident that at the time that we launched that no one could think of anything at all to do that would improve the probability of success um and we we racked our brains to think of any possible way to improve the probability of success we could not think of anything more and and nor could nasa and so then that that's just the best that we could do so then we we had we went ahead and launched now i'm not a religious person um but i nonetheless got on my knees and prayed for that mission you able to sleep

3:06-3:64

no how did it feel when it was a success first when the launch was a success and when they returned back home or back to earth it was a great relief yeah it's for high stress situations i find it's not so much elation as relief um and um you know i think once as as we we got more comfortable and improved out the systems because you know we really um you know you got to make sure everything

3:64-4:21

works um i was it was definitely a lot more enjoyable with the subsequent uh astronaut uh missions and i thought the the inspiration mission was was actually very inspiring um the inspiration for mission um i'd encourage people to watch the inspiration documentary on netflix it's actually really good um and it really isn't so i i was actually inspired by that um and i i i so that one i felt i i was kind of able to enjoy the the actual mission and not just be super stressed all the time so for people that somehow don't know it's the all civilian first

4:21-4:83

time all civilian out to space out to orbit yeah it was the high i think the highest orbit that uh in like under 30 or 40 years or something the only one that was higher was the one shuttle sorry a hubble uh servicing mission um and then before that it would have been um apollo in 72. it's pretty wild so it's cool it's good you know i think uh as you know as a species like we want to be you know continuing to do better and and reach higher ground and and like i think it would be tragic

4:83-5:40

extremely tragic if um apollo was the high water mark for humanity you know and that then that's as far as we ever got and it's um it's concerning that here we are um 49 years after the last mission to the moon it's almost half a century uh and we've not been back um and that's that's worrying it's like is that does that mean we've peaked as a civilization or what so like i think we got to get back to the moon and build a base there you know a

5:40-5:97

science base i think we could learn a lot about the nature of the universe if we have a proper science base on the moon um you know like we have a science based in antarctica and you know many other parts of the world and um so that that that's like i think the next big thing we've got to have like a serious like moon base um and then get people to mars and you know get get out there and be a space bearing civilization i'll ask you about some of those details but since you're so busy with the hard engineering challenges of everything that's involved

5:97-6:53

are you still able to marvel at the magic of it all of space travel of every time the rocket goes up especially when it's a crude mission or you're just so overwhelmed with the all the challenges that you have to solve and actually sort of to add to that the reason i want to ask this question of may 30th it's it's been some time so you can look back and think about the impact already it's already at the time it was an engineering problem maybe now it's becoming a historic moment like it's a moment that how many moments would be remembered about the 21st century to me

6:53-7:08

that or something like that maybe inspiration for one of those would be remembered as the early steps of a new age of uh space exploration yeah i mean during the launches itself so i mean i think i think maybe some people know but a lot of people don't know it's like i'm actually the chief engineer of spacex so um the you know i've signed off on pretty much all the design decisions um and you know so if there's something that goes wrong with that vehicle it's it's fundamentally my fault

7:08-7:65

you know so um so i'm really just thinking about all the things that like so so when i see the rocket i see all the things that could go wrong and the things that could be better and the same with the dragon spacecraft it's like other people say oh this is a spacecraft or a rocket and this looks really cool i'm like i've like a readout of like this is the these are these are the risks these are the pro the problems that's what i see so it's not what other people see when they see the product you know so let me uh ask

7:65-8:27

you then to analyze starship in that same way i know you have you'll talk about in more detail about starship in the near future perhaps yeah talk about it now if you want um but just in that same way like you said you see when you see up when you see a rocket you see a sort of a list of risks in that same way you said that starship is a really hard problem so there's many ways i can ask this but if you magically could solve one problem perfectly one engineering problem perfectly which one would it be on sasha on sorry on starship so is it maybe related to the efficiency the

8:27-9:00

uh the engine the weight of the different components the complexity of various things maybe the controls of the the crazy thing has to do to land no it's actually the by far the the biggest thing absorbing my time is uh uh engine production not not the design of the engine but i've often said prototypes are are easy production is hard so we have the most advanced rocket engine that's ever been designed um the because i say currently the the best rocket engine ever is probably the rd 180 or rd-170 the um that that's the russian engine basically

9:00-9:51

um and um and still i think an engine should only count if it's gotten something to orbit um so our engine has not gotten anything to orbit yet um but it is it's the first engine that's actually better than than the the russian rd engines which are amazing design so you're talking about raptor engine what makes it amazing what what are the different aspects of it that make it like what are you the most excited about uh if the whole thing

9:51-10:02

works in terms of efficiency all those kinds of things well it's the raptor is a a full flow uh staged combustion um engine and it's operating at a very high chamber pressure so one of the key figures america perhaps the key figure of merit is um what is the chamber pressure at which the rocket engine can operate that's the combustion chamber pressure

10:02-10:57

so a raptor is designed to operate at 300 bar possibly maybe higher that's 300 atmospheres so um the record right now for operational engine is the rd engine that i mentioned the russian rd which is i believe around 267 bar and the the the difficulty of the chamber pressure is increases on a non-linear basis so 10 more pressure is more like uh 50 more difficult

10:57-11:21

um but that that chain of pressure is that that is what allows you to get a very high uh power density for the engine um so uh enabling um a very high thrust to weight ratio and a very high specific impulse so specific impulse is like a measure of the efficiency of a rocket engine or um it's really the the the uh exhaust the effective exhaust velocity of of the gas coming out of the engine um so uh

11:21-11:78

with a very high chamber pressure you can have um a a compact engine that nonetheless has a high expansion ratio which is the ratio between the uh exit nozzle uh and the uh throat so you know engine's got like you see a rocket just got like sort of like about like a hourglass shape it's like a chamber and then it next down and there's a nozzle and the ratio of the the exit diameter to the throat expansion ratio so why is it such a hard engine to manufacture

11:78-12:42

at scale uh it's very complex so a lot of what does complexity mean here's a lot of components involved there's a lot of a lot of components and a lot of uh unique materials that so we had to invent a several alloys that don't exist in order to make this engine work um materials problem too it's a materials problem and um it is in a staged combustion a full flow stage combustion there there are many uh feedback loops in the system so you uh

12:42-12:94

basically you've got uh propellant and and and uh hot gas flowing um simultaneously to so many different places on the engine and they all have a recursive effect on each other so you change one thing here it has a recursive effect here it changes something over there and and it's it's it's quite hard to control um like there's a reason no one's made this before um but

12:94-13:47

um and the reason we're doing um a stage combustion uh full flow is because it it has the highest uh the highest uh theoretical possible uh efficiency um so in in in order to make a fully reusable rocket um which that's the really the holy grail of orbital rocketry um you have to have everything's got to be the best uh it's got to be the best engine the best

13:47-14:09

airframe the best heat shield um extremely light uh avionics um you know very clever control mechanisms um you've got to shed mass in in any possible way that you can um for example instead of putting landing legs on the booster and chip we are going to catch them with a tower to save the weight of the landing lens legs so that's like i mean we're talking about catching the largest flying object ever made uh with on a giant tower with with chopstick arms

14:09-14:75

it's like karate kid with the fly but much bigger i mean pulling this probably won't work the first time uh and anyway so this is bananas this is banana stuff so you mentioned that you doubt well not you doubt but there there's days or moments when you doubt that this is even possible it's so difficult the possible part is well at this point we'll i think we will get starship to work um um there's a question of timing how long will it take us to do this how long will it take us to actually

14:75-15:47

achieve a full and rapid reusability because it will take probably many launches before we are able to have full and rapid reusability but i can't say that that the physics pencils out like the like we're not uh like at this point i'd say we're confident that that like let's say i'm very confident success is in the set of all possible outcomes for a while there i was not convinced that success was in the set of possible outcomes which is very important actually but so um saying there's a chance i'm saying

15:47-16:00

there's a chance exactly um uh just not sure how how how long it will take uh we're very very talented team they're working night and day to make it happen um and uh and like like i said the the the critical thing to achieve for the revolution in space flight and for humanity to be a space frank civilization is to have a fully and rapidly reusable rocket oval rocket there's not even been any orbital rocket that's been fully reusable ever and this has always been the

16:00-16:60

the the holy grail of rocketry and many smart people very smart people have tried to do this before and have not succeeded so um because it's such a hard problem what's your source of belief in situations like this when the engineering problem is so difficult there's a lot of experts many of whom you admire who have failed in the past yes and um a lot of people you know a lot of experts maybe journalists all

16:60-17:37

the kind of you know the public in general have a lot of doubt about whether it's possible and you yourself know that even if it's a non-null set not empty set of success it's still unlikely or very difficult like where do you go to both personally um intellectually as an engineer as a team like for source of strength needed to sort of persevere through this and to keep going with the project take it to completion a source of strength hmm i i just really not how i think about things um

17:37-17:94

i mean for me it's simply this this is something that is important to get done um and we we should just keep doing it um or die trying and i i don't need a source of strength so quitting is not even like um that's not it's not my nature okay and i i don't care about optimism or pessimism [ __ ] that we're gonna get it done gonna get it done can you uh then zoom back in to specific problems with starship or any engineering problems you work on

17:94-18:53

can you try to introspect your particular biological neural network your thinking process and describe how you think through problems the different engineering and design problems is there like a systematic process you've spoken about first principles thinking but is there kind of a process to it well um you know like saying like like physics is law and everything else is a recommendation um like i've met a lot of people who can break the law but i haven't met anyone who could break physics so uh so first for you know any kind of technology problem you have to sort of

18:53-19:06

just make sure you're not violating physics um and you know uh first principles analysis i think is something that can be applied to really any walk of life uh any anything really it's just it's it's really just saying um you know let's let's well something down to the most fundamental uh principles the things that we are most confident are true at a foundational level and that sets you at your sets your axiomatic base and then you reason up

19:06-19:58

from there and then you cross-check your conclusion against the the axiomatic truths um so um you know some basics in physics would be like are you violating conservation of energy or momentum or something like that you know then it's not gonna work um so uh that's you know so that's just to establish is it is it possible and then another good physics tool is thinking about things in the limit if

19:58-20:11

you if you take a particular thing and you uh scale it to a very large number or to a very small number how does how do things change um well it's like tempo like in number of things you manufacture something like that and then in time yeah like let's say say the example of like um like manufacturing which i think is just a very underrated problem um and and uh likes it it's it's much harder to take

20:11-20:57

an advanced technology product and bring it into volume manufacturing than it is to design it in the first place my orders magnitude so um so let's say you're trying to figure out is like why is this this uh part or product expensive is it um because of something fundamentally foolish that we're doing or is it because our volume is too low and so then you say okay well what if our volume was a million units a year is it still expensive that's what i'm

20:57-21:08

radically thinking about things to the limit if it's still expensive at a million units a year then volume is not the reason why your thing is expensive there's something fundamental about design and then you then can focus on the reducing complexity or something like that and change the design to change changes apart to be something that is uh uh not fundamentally expensive but like that's a common thing in rocketry because the the unit volume is relatively low and so a common excuse would be well it's expensive because our unit volume is low um and if we were in

21:08-21:55

like automotive or something like that or consumer electronics then our costs would be lower and like i'm like okay so let's say we skip now you're making a million units a year is it still expensive if the answer is yes then uh economies of scale are not the issue do you throw into manufacturing do you throw like supply chain you talk about resources and materials and stuff like that do you throw that into the calculation of trying to reason from first principles like how we're going to make the supply chain work here

21:55-22:21

yeah yeah and then the cost of materials things like that or is that too much exactly so um like another like a good example of thinking about things uh in the limit is um if you take any uh you know any any product any machine or whatever um like take a rocket or whatever and say uh if you've got if you look at the room raw materials in the rocket um so you're gonna have like uh aluminum steel titanium inconel especially specialty alloys um copper and and you say what are the

22:21-22:72

how what what's the weight of the constituent elements of each of these elements and what is their raw material value and that sets the asymptotic limit for how low the cost of the vehicle can be unless you change the materials so and then when you do that i call it like maybe the magic wand number or something like that so that would be like if you had the you know a like just a pile of these raw materials here and you could wave magic wand and rearrange the atoms into the

22:72-23:23

final shape um that would be the lowest possible cost that you could make this thing for unless you change the materials so then and that is always a you're almost always a very low number um so then the what's actually causing these to be expensive is how you put the atoms into the desired shape yeah actually if you don't mind me taking a tiny tangent i had uh i often talked to jim keller who was somebody that worked with you oh yeah that's a fantastic job jim was

23:23-23:80

yeah did great work at tesla so um i suppose he carries the flame of the same kind of thinking that you're you're talking about now um and i guess i see that same thing at tesla and spacex folks who work there they kind of learn this way of thinking and it kind of becomes obvious almost but anyway i had um argument not argument uh he educated me about how cheap it might be to manufacture teslabot we just we had an argument what is how can you reduce the cost of scale

23:80-24:40

of producing a robot because i've gotten a chance to interact quite a bit um obviously in in the academic circles with humanoid robots and then boston dynamics and stuff like that and they're very expensive to to build and then uh jim kind of schooled me on saying like okay like this kind of first principle is thinking of how can we get the cost of manufacturing down um i suppose you do that you have done that kind of thinking for teslabot and for all kinds of all kinds of complex systems that are traditionally seen as complex and you say okay how can we simplify everything

24:40-24:95

down yeah i mean i think if you are really good at manufacturing you can basically make at high volume you can basically make anything for a cost that asymptotically approaches the real raw material value of the constituents plus any intellectual property that you need to license anything right but it's hard it's not like that's a very hard thing to do but but it is possible for anything anything in volume can be made like i said for a class

24:95-25:56

that asymptotically approaches raw material uh constituents plus intellectual property license rights so what will often happen in trying to design a product is people will start with the tools and and parts and methods that they are familiar with and then and try to create a product using their existing tools and methods the other way to think about it is actually imagine the try to imagine the platonic ideal of the perfect product or technology whatever it might be and so what is this what is the perfect arrangement of atoms

25:56-26:13

that would be the the best possible product and now let us try to figure out how to get the atoms in that shape i mean it's it sounds um it's almost like a rick and morty absurd until you start to really think about it and you really should think about it in this way because everything else is kind of uh if you if you think uh you you might fall victim to the momentum of the way things are done in the past unless you think in this way well just as a function of inertia people will

26:13-26:60

want to use the same tools and methods that they are familiar with um they just that's what they'll do by default yeah and then that will lead to an outcome of things that can be made with those tools and methods but it is unlikely to be the platonic idea of the perfect product um so then so that's why it's good to think of things in both directions they're like what can we build with the tools that we have but then but but also what is the what is the perfect the theoretical perfect product look like and that that

26:60-27:25

theoretical perfect part is going to be a moving target because as you learn more the definition of or or for that perfect product will change because you don't actually know what the perfect product is but you can successfully approximate a more perfect product so the thing about it like that and then saying okay now what tools methods materials whatever do we need to create in order to get the atoms in that shape but for people rarely think about it that way but it's a powerful tool i should mention that the brilliant siobhan zillis is hanging on

27:25-28:14

hanging out with us in case you hear a voice of uh wisdom from uh from from outside from up above okay so let me ask you about mars you mentioned it would be great for science to put um a base on the moon to do some research but the truly big leap again in this category of seemingly impossible is to put a human being on mars when do you think spacex will land a human being on mars hmm best case is about five years

28:14-28:88

worst case tenures what are the determining factors would you say from an engineering perspective or is that that not the bottlenecks uh you know it's fundamentally um you know engineering the the vehicle um i mean starship is the most complex and advanced rocket that's ever been made by i don't know order of magnitude or something like that it's a lot it's really next level so um and the fundamental optimization of starship is minimizing cost per ton to orbit and ultimately cost per ton to the surface of mars

28:88-29:45

um this may seem like a mercantile objective but it is actually the thing that needs to be optimized um like there is a certain cost per tonne to the surface of mars where we can afford to establish a self-sustaining uh city um and and then above that we cannot afford to do it um so right right now you couldn't fly to mars for a trillion dollars doesn't no amount of money could get your ticket to mars so we need to get that above uh you know to get that like something that is actually possible at all um um

29:45-29:97

but but then but that's that's we don't just want to have you know with mars flags and footprints and then not come back for a half century like we did with the moon uh in order to pass a very important great filter i think we need to be a multi-planet species um this may sound somewhat esoteric to to a lot of people but uh like eventually given enough time uh that's something the earth is likely to experience

29:97-30:56

some calamity um that could be uh something that humans do to themselves or an external event like happen to the dinosaurs um and um but a bit of you know eventually and if nothing if none of that happens and somehow magically we keep going then the sun will the sun is gradually expanding um and will engulf the earth um and probably earth gets too hot for uh life in uh about 500 million years it's a long time

30:56-31:11

but that's only 10 longer than earth has been around and so if you think about like the because the current situation is really remarkable um and kind of hard to believe but uh it's been around four and a half billion years and this is the first time if one half billion years that has been possible to extend life beyond earth and that window opportunity may be open for a long time and i hope it is but it also may be open for a short time and we should uh i think it was wise for us to uh

31:11-31:68

act quickly while the window is open just in case it closes yeah the existence of nuclear weapons pandemics all kinds of threats yeah should uh should kind of um give us some motivation i mean civilization could get um could die with a bang or a whimper you know if it's uh if it dies a demographic collapse then it's more of a whimper obviously but if it's world war three it's more of a bang but but these are all risks

31:68-32:17

um i mean it's important to think of these things and just you know think of things as like probabilities not certainties um there's a certain probability that something bad will happen on earth like i think most likely the future will be good um but there's like let's say for argument's sake um a one percent chance per century of of a civilization ending event like that was stephen hawking's estimate um i think he's he might be right about that uh so

32:17-32:77

then uh you know we should basically think of this like being a multi-planet species just like taking out insurance for life itself like life insurance for life [Laughter] so it's turned into an infomercial real quick life insurance for life yes um and you know we we can bring the the the creatures from uh you know plants and animals from earth to mars and breathe life into the planet um and and have a second planet with with life um that would be great um they can't bring themselves there you know so if we don't

32:77-33:45

bring them to mars then they will just for sure all die when the sun expands anyway and then that'll be it what do you think is the most difficult aspect of building a civilization on mars terraforming mars like from engineering perspective from a financial perspective human perspective to get get a large number of folks there who will never return back to earth uh no they could certainly return some will return back to earth they will choose to stay there for the rest of their lives yeah many will um but uh we you know we we need the spaceships back like the

33:45-33:91

ones that go to mars read them back so you can hop on if you want you know it's like but we can't just not have the spaceships come back those things are expensive we need them i'd like to come back and do another trip i mean do you think about the terraforming aspect like actually building are you so focused right now on the spaceships part that's so critical yeah tomorrow we absolutely if you can't get there nothing else matters so and like i said we can't get there with at some extraordinarily high cost i mean the current cost of

33:91-34:46

um let's say one ton to the surface of mars is on the order of a billion dollars so because you don't just need the rocket and the launch and everything you need like heat shield you need you know guidance system you need deep space communications you need some kind of landing system so like rough approximation would be a billion dollars per ton to the surface of mars right now um this is obviously um way too expensive to create a self-sustaining civilization

34:46-35:07

um so we need to improve that by at least a factor of a thousand a million per ton yes ideally less than much less than a million ton but if it's not like it's got to be say if i say like what well how much can society affords to spend or want to just want to spend on a self-sustaining city on mars the self-sustaining part is important like it's just the key threshold um the grateful to will will have been passed when the city on mars it can survive even if the spaceships

35:07-35:62

from earth stop coming for any reason doesn't matter what the reason is but if they stop coming for any reason will it die out will it not and if there's even one critical ingredient missing then it still doesn't count it's like you know if you're on a long sea voyage and you've got everything except vitamin c and it's only a matter of time you know you're going to die so so we're going to get mars city to the point where it's self-sustaining um i'm not sure this will really happen in my lifetime but i i hope to see it at least have a lot of momentum and and then you could say okay what is

35:62-36:18

the minimum tonnage necessary to have a self-sustaining city um and there's a lot of uncertainty about this you could say like i don't know it's probably at least a million tons um because you have to set up a lot of infrastructure on on mars like i said you can't be missing any anything that in order to be self-sustaining you can't be missed like you need you know semiconductor fabs you need iron ore refineries like you need lots of things you know so um and mars is not super hospitable it's

36:18-36:73

it's the least inhospitable planet but it's definitely a fixer-upper of a planet outside of earth yeah the earth is pretty earth is like easy yeah and also we should clarify in the solar system yes in the solar system there might be nice like vacation spots there might be some great planets out there but it's hard to get there yeah way way way way too hard to say at least let me push back on that not really a pushback but a quick curveball of a question so you did mention physics as the the first starting point so

36:73-37:33

um general relativity allows for wormholes uh they technically can exist do you think um those can ever be leveraged by humans to travel faster than the speed of light well are you saying the whole thing is debatable uh the that we currently do not know of any means of going faster than the speed of light uh there is like like there are some ideas about having space like so so

37:33-38:01

you can only move at the speed of light through through space but if you can make space itself move that that's like that that's warping space um space is is capable of moving faster than the speed of light right uh like the universe in the big bang the universe the universe expanded at much much more than the speed of light by a lot yeah um so um but the if this is possible the the amount of energy required towards space is so gigantic it's boggles the mind

38:01-38:57

so all the work you've done with propulsion how much innovation is possible with rocket propulsion is this um i mean you've seen it all and you're constantly innovating in every aspect how much is possible like how much can you get 10x somehow is there something in there in physics that you can get significant improvement in terms of efficiency of engines and all those kinds of things well as i'm saying like the really the holy grail is a fully and rapidly reusable orbital system um so uh right now uh

38:57-39:08

the falcon 9 is the only reusable rocket out there that but it but the the booster comes back and lands and you've seen the videos and we get the nose cone or faring back but we do not get the upper stage back so uh that means that we have a minimum cost of building up for stage um you can think of like a two-stage rocket of sort of like two airplanes like a big airplane and a smaller airplane um and we get big airplane back but not the smaller airplane and so it still

39:08-39:59

costs a lot you know so that upper stage is you know at least 10 million dollars um and then the degree of the the booster is not as reused it's not as rapidly and completely reusable as we'd like in order of the fairings so you know our kind of minimum marginal cost not counting overhead for per flight is on the order of 15 to 20 million dollars maybe um so uh that's that's extremely good for

39:59-40:19

it's by far better than any rocket ever in history um but uh with full and rapid reusability we can reduce the cost per ton to orbit by uh a factor of a hundred but just think of it like um like imagine if you had an aircraft or something or a car oh yeah and if you had to buy a new car every time you went for a drive it would be very expensive it'll be silly frankly but um but you in fact you just refuel the car or recharge the car

40:19-40:76

and that's uh makes your trip like i don't know a thousand times cheaper so it's the same for rockets uh if you it's very difficult to make this complex machine that can go to orbit and so if you cannot reuse it and if you have to throw even any part of any significant part of it away that massively increases the cost so you know starship in theory could do a cost per launch of like a million maybe two million dollars or something like that

40:76-41:25

um and uh and put over 100 tons in orbit which is crazy yeah so that's incredible so you're saying like it's uh by far the biggest bang for the buck is to make it fully reusable versus like some kind of brilliant breakthrough in theoretical physics no no there's no there's no brilliant break no there's no it just me you're gonna make the rocket reusable this is an extremely difficult insuring problem got it uh but no no new physics is required

41:25-41:86

just brilliant engineering let me ask a slightly philosophical fun question gotta ask i know you're focused on getting to mars but once we're there on mars what do you what form of government economic system political system do you think would work best for an early civilization of humans is the i mean the the interesting reason to talk about this stuff it also make helps people dream about the future i know you're really focused about the short-term engineering dream but it's like i don't know there's something about imagining an actual civilization on mars that gives people

41:86-42:52

it really gives people hope well it would be a new frontier and an opportunity to rethink the whole nature of government just as was done in the creation of the united states so i mean i would suggest having a direct democracy like people vote directly on things as opposed to representative democracy so uh representative democracy i think is too uh subject to a special interest and you know a coercion of the politicians and that kind of thing um so i i'd recommend

42:52-43:07

uh that that there's just um direct democracy people vote on laws the population votes on laws themselves and then the laws must be short enough that people can understand them yeah and then like keeping a well-informed populist like really being transparent about all the information about what they're voting for absolute transparency yeah and not make it as annoying as those cookies we have to accept cookies like always like you know there's like always like a slight amount of trepidation when you click accept cookies like i feel as

43:07-43:67

though there's like perhaps like a like a very tiny chance that'll open a portal to hell or something like that that's exactly how i feel why why do they why do they keep wanting to accept that what do they want with this cookie like somebody got upset with accepting cookies or something somewhere who cares like so annoying to keep accepting all these cookies to me this is just a great exception yes you can have my damn cookie i don't care whatever you heard it from me on first he accepts all your damn cookies yeah and start asking me annoying

43:67-44:17

yeah it's uh it's one example of um implementation of a good idea done really horribly yeah it's somebody was like there's some good intentions of like privacy or whatever but now everyone's just has to take accept cookies and it's not you know you have billions of people who have to keep clicking except cookie it's super annoying then we just accept the damn cookie it's fine there is like um i think a fundamental problem that we're because we've not really had a a major

44:17-44:70

uh like a world war or something like that in a while and obviously we would like to not have world wars um the there's not been a cleansing function for rules and regulations um so wars did have uh you know some sort of lining in that there would be a a reset on rules and regulations uh after a war um so world wars one and two there were huge resets on rules and regulations now as for if the society society does not have a war and there's no cleansing function or garbage collection for rules and regulations then rules and regulations

44:70-45:23

will accumulate every year because they're immortal there's no actual humans die but the laws don't so we need a garbage collection function for rules and regulations they should not just be immortal um because some of the rules and regulations that are put in place will be counterproductive uh done with good intentions but counterproductive sometimes not done with good intentions so um if you just if rules and regulations just accumulate every year and you get more and more of them then eventually you won't be able to do anything you're just like gulliver with

45:23-45:77

you know tied down by thousands of little strings and we had we see that in um you know us and like basically all economies that uh have been around for for a while uh and and regulators and legislators create new rules and regulations every year but they don't put effort into removing them and i think that's very important that we put effort into removing rules and regulations um but it gets tough because you get special interests that then are dependent on like they they have a

45:77-46:27

you know a uh vested interest in that whatever rule and regulation and that they then they fight to not get it removed um yeah so it i mean i guess the problem with the constitution is it's it's kind of like c versus java because it doesn't have any garbage collection built in i think there should be i when you first said the the the metaphor of garbage collection from a coding standpoint from the colony standpoint yeah yeah i it would be interesting interesting if the laws themselves kind of had a built-in

46:27-46:84

thing where they kind of die after a while unless somebody explicitly publicly defends them yeah so that that's sort of it's not like somebody has to kill them they kind of die themselves they disappear yeah um not to defend java or anything but you know the c plus plus you know you could also have a great garbage collection in python and so on yeah so yeah something's good something needs to happen or or just the the civilization's arteries arteries just harden over time and and you can just get less and less done because there's just a rule against

46:84-47:35

everything so yeah i think like i don't know for mars order i'd say or even for you obviously for earth as well like i think there should be an active process for removing rules and regulations and questioning their existence just um like if we've got a function for creating rules and regulations because rules and regulations can also think of as like they're like software or lines of code for operating uh civilization that's the rules and regulations um so it's like we shouldn't have rules and

47:35-47:86

regulations but you have code accumulation but no code removal um and so it just gets to become basically archaic bloatware after a while um and and it's just it makes it hard for things to progress so i don't know maybe mars you'd have like an uh you know any given law must have a sunset you know and and and uh and require active voting to keep restoring to keep it up there you know um and i actually also say like

47:86-48:51

and these just i don't know recommendations or thoughts and ultimately we'll be up to the people on mars to decide but i think um it should be easier to remove a law than to add one because of the just to overcome the inertia of laws so maybe it's like uh for argument's sake you need like say 60 percent vote to have a law take effect but only a 40 vote to remove it so let me be the guy you posted meme on twitter recently where there's there's like a row of urinals a guy just walks all the way across

48:51-49:10

and he tells you about crypto listen i mean that's happened to me so many times i think maybe even literally uh yeah do you think technologically speaking there's any room for ideas of smart contracts or so on because you mentioned laws um that's an interesting implement use of things like smart contracts to implement the laws by which governments function like something built on ethereum or maybe a dog coin that enables smart contracts somehow i never i don't quite understand this whole smart contract thing um you

49:10-49:64

know i mean so it's two dumb times to have small contracts it's a good line i mean my general approach to any kind of like deal or whatever is just make sure there's clarity of understanding that's the most important thing um and and just keep any kind of deal very very short and simple plain language and just make sure everyone understands this is the deal does everyone is it clear and uh and and what are the consequences if

49:64-50:23

various things don't happen um but usually deals are um you know business deals or whatever are way too long and complex and overly lawyered and pointlessly you mentioned that uh doge is the people's coin yeah and you said that you were literally going spacex may consider literally putting uh a dosh coin on the moon is this something you're still considering uh mars perhaps do you think there's some chance we've talked about political systems on mars

50:23-50:83

that uh dogecoin is the official currency of mars that's coming in the future well i think mars itself will need to have a different currency because you can't synchronize due to speed of light or not easily um so it must be completely standalone from earth well yeah because the but mars is at closest approach it's four light minutes away roughly and then at furthest approach uh it's roughly 20 light minutes away uh maybe a little more um so you can't really have something

50:83-51:34

synchronizing you know if you could if if you've got a 20 minute speed light issue if it's got a one minute blockchain uh it's not going to synchronize properly um so mars really would i don't know if mars would have a cryptocurrency as a thing but probably seems likely but it would be some kind of localized thing on mars and you let the people decide yeah absolutely the future of mars should be up to the martians yeah so um

51:34-52:00

i mean i think the cryptocurrency thing is an interesting approach to reducing the um error in the the database that is called money um you know i think i have a pretty deep understanding of the of what money actually is on a practical day-to-day basis because of paypal um you know i really got in deep there um and right now the money system actually for practical purposes is is really a bunch of uh heterogeneous uh

52:00-52:67

mainframes running uh old cobalt okay you mean literally that's literally that is literally what's happening in batch mode okay in patch mode yeah uh pretty the poor fastest who have to maintain that code okay that's a that's a pain that's pain not even for trans cobalt yep it's cobalt it's like and they still banks are still buying mainframes in 2021 and running ancient global code uh and uh you know the the federal reserve is like probably even older than the what the banks have and they have an old kobold mainframe

52:67-53:25

and so now and and so the the government effectively has editing privileges on the on the money database um and they use those editing privileges to um make more money whenever they want and this increases the error in the database that is money so i think money should really be viewed through the lens of information theory and uh and so it's uh you're kind of like uh like an internet connection like what's the bandwidth uh you know total bit rate uh what is the latency judder

53:25-53:77

uh packet drop uh you know errors in errors in the network uh communication using money like that basically um i think that's probably why i really think of it and and then say what what system uh from an information theory standpoint allows an economy to function the best uh and you know um crypto is an attempt to reduce the the error uh in uh in money that is contributed by

53:77-54:48

uh governments uh diluting the money supply as basically a pernicious pernicious form of taxation so both policy in terms of with inflation and actual like technological cobalt like cryptocurrency takes us into the 21st century in terms of the actual systems that allow you to do the transaction to store wealth all those kinds of things like i said just think of money as information people um often will think of money as having power in and of itself um it does not money is uh is information and it it does not have power in and of itself

54:48-54:94

uh the like the you know again applying the physics tools of thinking about things in the limit is helpful if you are stranded on a tropical island um and uh you have a trillion dollars it's useless because there's no there's no resource allocation money is a database for resource allocation but there's no resources to allocate except yourself so money is useless um

54:94-55:57

uh if you're surrounded on desert island with no food you'd uh all the bitcoin in the world will not stop you from starving yeah so um so like just think of money as as a database for resource allocation um across time and space and um and then what what what system uh it is what what in what form should that that database or data system

55:57-56:15

what what what would be most effective now there's a there is a fundamental issue with um say bitcoin in its current form uh in that it's the transaction volume is very limited um and uh the latency this is the latency for for a properly confirmed transaction is to is too long much longer than you'd like so it's not it's actually not great from um a transaction volume standpoint or a latency standpoint um uh so it is perhaps useful as

56:15-56:76

as to search to solve an aspect of the money database problem uh which is the sort of store of wealth or an accounting of relative obligations i suppose but it is not useful as a currency as a day-to-day currency but people have proposed different technological solutions yeah lightning network and the layer two technologies on top of that i mean it's it's all it seems to be all kind of a trade-off but the point is it's kind of brilliant to say that just think about information think about what kind of database what kind of infrastructure

56:76-57:24

enables yeah like you're operating an economy um and you need to have some thing that allows for the efficient to have efficient uh value ratios between products and services so you've got this massive number of products and services and you need to you can't just bar barter it's like that would be extremely unwieldy so you need something that gives you the the an a

57:24-57:76

a ratio of exchange between goods and services um and and then something that allows you to uh shift obligations across time like debt debt and equity shift obligations across time then what does what does the best job of that um part reason why i think there's some um merit to dogecoin even though it was obviously created as a joke um is that it it actually does have a much higher uh transaction volume capability than

57:76-58:35

bitcoin um and the you know the cut like the costs of doing a transaction the the dogecoin fee is is very low like right now if you want to do a bitcoin transaction the price of doing that transaction is very high so you could not use it effectively for most things um and nor could it even scale to a high volume um uh and when bitcoin was you know started i guess when around 2008 or something like that um the internet connections were much worse than they are today like order of magnitude cup

58:35-58:94

i mean there's the way way worse you know in 2008 so so like having us you know a small uh block size or whatever is you know and a long synchronization time is made sense in 2008 but did you know 2021 or fast forward 10 years it's like it's it's like comically low you know it's uh so um and i think there's some value to having a linear increase in the amount of currency that uh is generated um

58:94-59:54

so because some amount of the currency like like if a currency is too deflationary or like uh or should say if if if if a currency is expected to increase in value over time there's reluctance to spend it because they're like oh i if i i'll just hold it and not spend it because it's scarcity is increasing with time so if i spend it now then i will regret spending it so i will just you know total it but if there's some dilution of the currency occurring over time that's that's more of an incentive to use that as a currency so

59:54-60:08

um those coins somewhat randomly has uh a um just a fixed a number of of sort of coins or hash strings that uh are generated every year so there's there's some inflation but it's not a percentage base it's a it's so the it's a fixed number so the percentage of inflation will necessarily decline over time um so it just i'm not saying that it's like the ideal system for a currency but i

60:08-61:05

think it actually is uh just fundamentally better than anything else i've seen just by accident um so i like how you said um around 2008 so you're not uh you know some people suggested you might be satoshi nakamoto you previously said you're not i'm not would you tell us if you were yes okay uh do you think it's a feature of bug that he's anonymous or she or they it's an interesting kind of quirk of human history that there is a particular technology that is a completely anonymous inventor or creator well i mean you can you can look at the

61:05-61:69

um evolution of ideas um before the launch of bitcoin and see who wrote you know about those ideas um and then i like i don't know exactly obviously i don't know who created bitcoin for practical purposes but the evolution of ideas is is pretty clear before that and like it seems as though like nick zabo uh is probably more than anyone else uh responsible for the evolution of those ideas so he claims not to be nakamoto but

61:69-62:29

i'm not sure that's that's neither here nor there uh but he he seems to be the one more responsible for the ideas behind bitcoin than anyone else so it's not perhaps like singular figures aren't even as important as the the figures involved in the evolution of ideas that led to a thing so yeah yeah it's you know and most perhaps it's sad to think about history but maybe most names will be forgotten anyway what is the name anyway it's a name a name attached to an idea what does it even mean really i think shakespeare had a thing about roses and stuff whatever he said arose by any other name it smells sweet

62:29-62:94

i gotta underquote shakespeare i feel i feel like i accomplished something today shall i compare it to a summer's day [Laughter] i'm gonna clip that uh out more tempered animal fare [Laughter] autopilot tesla [Laughter] tesla autopilot has been through an incredible journey over the past six years um or perhaps even longer in the minds of in your mind in the minds of many involved uh i think that's where we first like

62:94-63:49

connected really with the autopilot stuff autonomy and the whole journey was incredible to me to watch i was um because i knew well part of as i was at mit and i i knew the difficulty of computer vision yeah and i knew the whole i had a lot of colleagues and friends about the darpa challenge i knew how difficult it is and so there was a natural skepticism when i first drove a tesla with uh the initial system based on mobile eye yeah i thought there's no way the first one i got in i thought there's no way this car could maintain

63:49-64:15

um like stay in the lane and create a comfortable experience so my intuition initially was that the lane keeping problem is way too difficult to solve oh thank you yeah that's relatively easy well yeah but like uh but not this but solve in the way that we just we talked about previous is prototype versus a thing that actually creates a pleasant experience over hundreds of thousands of miles or millions a lot of code around the mobile eye thing it doesn't just work by itself yes i mean there's part that's part of the story of how you approach things sometimes sometimes you do things from scratch sometimes at first you kind of

64:15-64:67

see what's out there and then you decide to from scratch that was one of the boldest decisions i've seen is both on the hardware and the software to decide to eventually go from scratch i thought again i was skeptical whether that's going to be able to work out because it's such a such a difficult problem and so it was an incredible journey what i see now with um everything the hardware the compute the sensors the the things i maybe care and love about most is the the stuff that andre karpathy is leading with the data set selection the whole

64:67-65:23

data engine process the neural network architectures the the way that's in the real world that network is tested validated all the different test sets uh you know versus the image net model of computer vision like what's in academia is like real world artificial intelligence so um andrei's awesome and obviously plays an important role but we have a lot of really talented people driving things so um and uh ashok is actually the head of autopilot engineering um uh

65:23-65:82

andre is the director of ai ai stuff yeah yeah so yeah there's i'm aware that there's an incredible team of just a lot going on yeah just uh you know as people people will give off will give me too much credit and they'll give andre too much credit so and people should realize how much is going on under the yeah it's just a lot of really talented people um the tesla autopilot ai team is extremely talented it's like some of the smartest people in the world so yeah we're getting it done what are some insights you've gained over those five six years of autopilot about the

65:82-66:36

problem of autonomous driving so you leaped in having some sort of first principles kinds of intuitions but nobody knows how difficult the problem yeah like i thought the self-driving problem would be hard but it's it was harder than i thought it's not like i thought it'd be easy i thought it'd be very hard but it was actually way harder than than even that so i mean what it comes down to at the end of the day is to solve self-driving uh you have to solve uh

66:36-66:87

you you basically need to recreate um what you what humans do to drive which is humans drive with optical sensors eyes and biological neural nets um and so in order to that that's how the entire road system is designed to work with with uh basically passive optical and neural nets um it biologically um and now that we need to it so for actually for full-size driving to work we have to recreate that in digital form um so we have to

66:87-67:43

um that that means cameras with uh advanced uh neural nets in silicon form uh and and then you it will obviously solve for full self-driving that's the only way i don't think there's any other way but the question is what aspects of human nature do you have to encode into the machine right so you have to solve the perception problem like detect and then you first well realize what is the perception problem for driving like all the kinds of things you have to be able to see like what what do we even look at when we drive there's uh

67:43-68:00

i just recently heard andre talked about at mit about like car doors i think it was the world's greatest talk of all time about car doors yeah um the the you know the fine details of car doors like what what is even an open car door man so like the the ontology of that that's the perception problem we humans solve that perception problem and tesla has to solve that problem and then there's the control and the planning coupled with the perception you have to figure out like what's involved in driving like especially in all the different edge cases

68:00-68:64

and and then i mean maybe you can comment on this how much game theoretic kind of stuff needs to be involved you know at a four-way stop sign you know our as humans when we drive our actions affect the world like sure it changes how others behave most autonomous driving if you you're usually just responding um to the scene as opposed to like really um asserting yourself in the scene do you think i think these actually i think i think these could these sort of control control logic conundrums are not are not the hard part

68:64-69:18

um the you know let's see um what do you think is the hard part in this whole um beautiful complex problem so it's a lot of freaking software man a lot of smart lines of code um for sure in order to have um create an accurate vector space so like you you're coming from image space which is like this this flow of um photons you're going to camera

69:18-69:83

cameras and and then uh so you have this massive bitstream um in image space and then you have to effectively compress uh the a massive bit stream uh corresponding to photons that knocked off an electron in a camera sensor uh and and turn that put stream into into vector space um i by by vector space i mean like uh

69:83-70:45

you know you've got cars and and humans and uh lane lines and curves and uh traffic lights and that kind of thing um once you uh have an accurate vector space um the control problem is similar to that of a video game like a grand theft auto cyberpunk um if you have accurate best vector space it's the control problem is it's i wouldn't say it's it's trouble it's not trivial but it's um like it's it's it's it's a it's not like some insurmountable thing it's a it's but

70:45-71:02

having accurate vector space is very difficult yeah i think we humans uh don't give enough respect to how incredible the human perception system is to mapping the raw photons to the vector space representation in our heads your brain is doing an incredible amount of processing um and and giving you an image that is a very cleaned up image like when we look around here we see like you see color in the corners of your eyes but actually your eyes have very few uh cones like the cone receptors in the peripheral vision your your eyes are

71:02-71:54

painting color in the peripheral vision you don't realize it but their eyes are actually painting color and your eyes also have like this blood vessels and all sorts of gnarly things and there's a blind spot but do you see your blind spot no your your brain is painting in the missing the blind spot you're gonna do these like see these things online where you look look here and look at this point and and then look at this point and it's if it's in your blind spot it your brain will just fill in the the missing version

71:54-72:03

the peripheral vision is so cool yes you realize all the illusions for vision sciences so makes you realize just how incredible the brain is the brain is doing crazy amount of post-processing on the vision signals from your eyes um it's insane so um and then and then even once you get all those vision signals uh your your brain is constantly trying to fig to forget as much as possible so human memory is perhaps the weakest thing about the brain is memory so because memory is

72:03-72:58

so expensive to a brain and so limited your brain is trying to forget as much as possible and distill the things that you see into uh the smallest smallest amounts of information possible so your brain is trying to not just get to a vector space but get to a vector space that is the smallest possible vector space of only relevant objects um and i think like you can sort of look inside your brain or at least i can like when you drive down the road and and try to think about what your brain is actually

72:58-73:03

doing consciously and it's conscious it's it's it's it's like you'll see a car that's you could because you're you don't have cameras you i don't have eyes in the back your head on the side you know so you say like but you you basically your your head is like a you know you basically have like two cameras on a slow gimbal um and and what's your and i say it's not that great okay you and i is uh you know like um and people

73:03-73:60

are constantly distracted and thinking about things and texting and doing all sorts of things they shouldn't do in a car changing radio station so having arguments you know is like um so so then like say like like uh like when's the last time you look right and left and you know or and rearward um or even diagonally you know forward to actually refresh your vector space so you're glancing around and what your mind is doing is is is trying to distill um

73:60-74:15

the relevant vectors basically objects with a position and motion uh and and and and then uh editing that down to the least amount that's necessary for you to drive it does seem to be able to uh edit it down or compress it even further into things like concepts so it's not it's like it goes beyond the human mind seems to go sometimes beyond vector space to sort of space of concepts to where you'll see a thing it's no longer represented spatially somehow it's almost like a concept that you should be aware of like if this is a school zone

74:15-74:77

you'll remember that as a concept which is a weird thing to represent but perhaps for driving you don't need to fully represent those things or maybe you get those kind of um well you indirectly you need to like establish vector space and then actually have predictions for uh that those vector spaces so like um you know like if uh you know like you drive past say say a a a a bus and and you see that there's there's people

74:77-75:29

before you drove past the bus you saw people crossing the interstate like or some just imagine there's like a large truck or something blocking site um but you before you came out to the truck you saw that there were some kids about to cross the road in front of the truck now you can no longer see the kids but you you need to be able but you would now know okay those kids are probably gonna pass by the truck and cross the road even though you cannot see them so you have to have um memory uh you need to remember that there were

75:29-75:81

kids there and you need to have some forward prediction of what their position will be it's a really hard time of relevance so with occlusions and computer vision when you can't see an object anymore even when it just walks behind a tree and reappears that's a really really i mean at least in academic literature it's tracking through occlusions it's very difficult yeah we're doing it i understand this yeah so some of it it's like object permanent it's like same thing happens with the humans with neural nets like when like a toddler grows up like there's a there's

75:81-76:30

a point in time where uh they develop they have a sense of object permanence so before a certain age if you have a ball uh or a toy or whatever and you put it behind your back and you pop it out if they don't before they have object permanence it's like a new thing every time it's like whoa this toy went poof just faired and now it's back again and they can't believe it and that they can play peekaboo all day long because this peekaboo is fresh every time [Laughter] but then we figure out object permanence then they realize oh no the object is

76:30-76:86

not gone it's just behind your back um sometimes i wish we never did figure out objective permanence yeah so that's uh that's an important problem to solve yes so so like an important evolution of the neural nets in the car is uh um memory card memory across both time and space um so now you can't remember like you have to say like how long do you want to remember things for and and it's it there's there's a cost to remembering things for a long time so you could you

76:86-77:43

know like run out of memory to if you try to remember too much for too long um and and then you also have things that are stale if if they're remember them for too long and then you also need things that are remembered over time so even if you like say have like fragrance like five seconds of memory uh on a time basis but like let's say you you're parked at light and you and you saw you use a pedestrian example that people were waiting to cross the across the road and you can't you can't quite see them because of an occlusion

77:43-77:98

uh but they might wait for a minute before the light changes for them to cross the road you still need to remember that that that's where they were um and that they're probably going to cross the road type of thing um so even if that exceeds your your your time-based memory should not exceed your space memory and i just think the data engine side of that so getting the data to learn all the concepts that you're saying now is an incredible process it's this iterative process of just it's this this hydrogen at many parts

77:98-78:62

we're changing the name to something else okay i'm sure it would be equally as yeah rick and morty like there's a lot of we've re-architected the neural net uh the neural nets in the cars so many times it's crazy oh so every time there's a new major version you'll rename it to something more ridiculous or uh or memorable and beautiful sorry not ridiculous of course if you see the full the full like uh array of neural nets that that that are operating in the car it kind of boggles the mind there's so there's so many layers it's crazy um

78:62-79:13

so yeah um but and we we started off with uh simple neural nets that were uh basically image recognition on a single frame from a single camera uh and then uh trying to knit those together with it you know it with the c i should say we we're really primarily running c here because c plus plus is

79:13-79:62

too much overhead and we have our own c compiler so to get maximum performance we actually wrote rotor and c compiler and are continuing to optimize our c compiler uh for maximum efficiency in fact we've just recently uh done a new rev on a c compiler that will compile directly to our autopilot hardware so you want to compile the whole thing down and with your own compiler yeah like so efficiency here because there's all kinds of compute there's cpu gpu there's like basic types of things and you have to somehow figure out the

79:62-80:17

scheduling across all those things and so you're compiling the code down yeah it does all okay this is so that's why there's a lot of people involved there's a lot of hardcore software engineering at a very sort of bare metal level uh because we're trying to do a lot of compute that's constrained to the you know our full self-driving computer so and we want to try to have the highest frames per second um possible um within a sort of very finite amount of

80:17-80:75

compute um and power so um we really put a lot of effort into the efficiency of our compute um and and uh so there's actually a lot of work done by some very talented software engineers at tesla that uh at a very foundational level to improve the efficiency of compute and how we use the the trip accelerators uh which are basically um dot you know uh doing matrix math dot products like a brazilian dot products

80:75-81:34

and it's like what what are neural nets it's like compute wise like 99 dot products so you know and you want to achieve as many high frame rates like video game you want yeah full resolution high frame rate high frame rate low latency um low jitter uh so um i think one of the things we're um moving towards now is no post processing of the

81:34-81:93

image through the um the image signal processor so um like for what happens for cameras is that almost all cameras is they um there's a lot of post processing done in order to make pictures look pretty and so we don't care about pictures looking pretty um we we just want the data we so we're removing just raw photon counts so the system will like the image that that the computer sees is actually much more than what

81:93-82:46

you see if you're represented on a camera it's got much more data and even in very low light conditions you can see that there's a small photon count difference between you know this spot here and that's about there which means that so it can see in the dark incredibly well um because it can detect these tiny differences in photon counts like much better than you'd possibly imagine um so and and then we also save uh 13 milliseconds on a latency uh so

82:46-83:02

um from removing the post processing in the image yes yeah it's like um because we've got you know eight cameras and and then there's uh roughly i don't know one and a half milliseconds or so maybe 1.6 milliseconds of latency um for each camera and so like uh um going to just uh basically bypassing the image processor uh gets us back 13 milliseconds of latency which is important um and we track latency all the way from

83:02-83:48

you know photon hits the the camera to you know all the steps that it's got to go through to get you know go through the um the various neural nets and the the c code and there's a little bit of c plus plus there as well um well i can maybe a lot but it the the core stuff is the heavy duty computers will see um and uh and so so we track that latency all the way to an output command to the um

83:48-84:01

drive unit to accelerate the brakes just to slow down the steering you know turn left or right um so because you got to output a command that's going to go to a controller and like some of these controllers have an update frequency that's maybe 10 hertz or something like that which is slow that's like now you lose 100 milliseconds potentially so um so then we want to update the the drivers on the like steering and braking control to have um more like uh 100 hertz instead of 10

84:01-84:52

hertz and you could have 10 milliseconds latency instead of 100 milliseconds worst case latency and actually jitter is more of a challenge than than latency because latency is like you can you can you can anticipate and predict but if you're but if you've got a stack up of things going from the camera to the to the computer through then a series of other computers and finally to an actuator on the car if you have a stack up of uh of tolerances of timing tolerances then you can have quite a variable latency which is called jetter and and that makes it

84:52-85:11

hard to to anticipate exactly what how you should turn the car or accelerate because you know if you got maybe 150 200 milliseconds of jitter then you could be off by you know after 0.2 seconds and this can make this could make a big difference so you have to interpolate somehow to to to uh deal with the effects of jitter so that you can make like robust controlled decisions the again you have to uh so the jitters and the sensor information or the jitter can occur at any stage in the pipeline you can if you have just if you have fixed latency you can anticipate um

85:11-85:74

and and uh like say okay we know that uh our information is for argument's sake 150 milliseconds stale like so for um 145 milliseconds from photons taking camera to um where you can measure a change in the acceleration of the vehicle um so then uh then you can say okay well we're gonna and we know it's 150 milliseconds so we're going to take that into account and uh and compensate for that latency however if you've got then 150 milliseconds of

85:74-86:24

latency plus 100 milliseconds of jitter that's which could be anywhere from zero to zero to 100 milliseconds on top so then your latency could be from 150 to 150 milliseconds now you got 100 milliseconds that you don't know what to do with and and that's basically random so getting rid of jitter is extremely important and that affects your control decisions and all those kinds of things okay um yeah the car is just going to fundamentally maneuver better with lower jitter um the cars will maneuver with superhuman

86:24-86:85

ability and reaction time much faster than a human i mean i think over time the autopilot full self driving will be capable of maneuvers that um you know uh you know are far more than what like james bond could do in like the best movie type of thing that's exactly what i was imagining in my mind as you said um it's like an impossible maneuvers that a human couldn't do you know so well let me ask sort of uh looking back the six years looking out into the future based on your current understanding how hard do you think this

86:85-87:40

is this full self-driving problem when do you think tesla will solve level four fsd i mean it's looking quite likely that it will be next year and what does the solution look like is it the current pool of fsd beta candidates they start getting greater and greater as they have been degrees of autonomy and then there's a certain level beyond which they can they can do their own they can read a book yeah so uh

87:40-87:96

i mean you can see anybody who's been following the fossil driving beta closely will see that the um the rate of disengagements has been dropping rapidly so like a disengagement b where where the driver intervenes to prevent the car from doing something right uh dangerous potentially so um so the interventions you know per million miles has been dropping uh dramatically at some point the and that

87:96-88:57

trend looks like it happens next year is the the the probability of an accident on fsd uh is uh less than that of the average human and then and then significantly less than that of the average human um so it certainly appears like we will get there next year um then of course that then there's going to be a case of okay we now have to prove this to regulators and prove it to you know and and we we want a standard that is not just equivalent to a human but uh much better than the average human i think it's got to be at least two or

88:57-89:14

three times uh higher safety than a human so two or three times lower probability of injury than a human before before we would actually say like okay it's okay to go it's not going to be equivalent it's going to be much better so if you look uh 10 point fsd 10.6 just came out recently 10.7 is on the way maybe 11 is on the way somewhere in the future yeah um we were hoping to get 11 out this year but it's uh 11 actually has a whole bunch of uh fundamental rewrites on the neural neural net architecture

89:14-89:72

and some fundamental improvements in creating vector space uh so there is a some fundamental like leap that really deserves the 11. i mean that's a pretty cool number yeah yeah uh 11 would be uh a single stack for all you know one stack to rule them all um and uh but but there's just some really fundamental uh neural net architecture changes that are that that will allow for

89:72-90:33

uh much more capability but but you know at first they're gonna have issues so like we have this working on like sort of alpha software and it's it's good but it's uh it's it it's basically taking a whole bunch of c c plus code and and deleting a massive amount of c plus plus go and replacing it with the neural net and you know andre um makes this point a lot which is like neural nets like kind of eating software you know over time there's like less and less conventional software more and more neural net we're just still software but it's you know still comes out the lines of software

90:33-91:03

but uh let's just more more neural net stuff uh and less uh you know heuristics basically um if you're more more more uh matrix based stuff and less uh heuristics based stuff um and um you know like like like one of the big changes will be um like right now the neural nets uh will um deliver a giant bag of points to the c plus plus or c and c plus plus

91:03-91:57

code yeah um we call the giant bag of points yeah uh and it's like so you go to pixel and and and something associated with that pixel like this pixel is probably car the pixel is probably lane line then you've got to assemble this giant bag of points in the c code and turn it into vectors and it does a pretty good job of it but it's it's a it's we want to just we need another layer of neural nets on top of that to take the

91:57-92:18

giant bag of points and distill that down to uh vector space in the neural net part of the software as opposed to the heuristics part of the software this is a big improvement um you know that's all the way down it's what you want it's not even your only realness but it's it's it's uh this will be just a ga this is a game changer to not have the bag of points the giant bag of points that has to be assembled with um many lines of cfc plus plus uh and and have the and have a neural net just assemble those into a vector so so that the

92:18-92:78

neural net is outputting um much much less data it's it's it's outputting this this is a lane line this is a curve this is drivable space this is a card this is a you know a pedestrian or a cyclist or something like that it's outputting um it's really outputting um profit vectors to the the cc plus less control control code as opposed to this sort of constructing the the vectors uh in c

92:78-93:31

um which we've done i think quite a good job of but it's it's a it you're kind of hitting a local maximum on the how well the c can do this um so this is this is really this is really a big deal and and just all of the networks in the car need need to move to surround video there's still some legacy networks that are not surround video and all of the training needs to move to surround video and the efficiency of the training uh it needs to get better and it is and then we need to move everything to

93:31-93:89

uh raw uh photon counts as opposed to um processed images okay so which is just quite a big reset on the training because the system's trained on post processed image images so we need to redo all the training uh to train against the the raw photon accounts instead of the post-processed image so ultimately it's kind of reducing the complexity of the whole thing so uh reducing reducing lines of code will actually go lower yeah that's fascinating um so you're doing fusion of all the

93:89-94:44

sensors so reducing the complexity of having to deal with these cameras cameras really right yes um same with humans yeah well i guess we got years too okay yeah well we'll actually need to incorporate um sound as well um because you know you need to like listen for ambulance sirens or you know fire trucks you know uh somebody like you know yelling at you or something i don't know just that there's there's a little bit of audio that needs to be incorporated as well do you need a cookie bath break yeah we listen to the trolls take a break okay

94:44-95:02

honestly frankly like the ideas are the easy thing and the implementation is the hard thing like the idea of going to the moon is is the easy part but going to the moon is the hard part it's the hard part um and there's a lot of like hardcore engineering that's got to get done at the hardware and software level uh likes it optimizing the c compiler and just you know uh cutting out latency everywhere like this is if we don't do this the system will not work properly um so the work of the engineers doing this

95:02-95:55

they are like the unsung heroes you know but they are critical to the success of the situation i think you made it clear i mean at least to me it's super exciting everything that's going on outside of what andre is doing yeah just the whole infrastructure of the software i mean everything is going on with data engine uh whatever whatever it's called the whole process is just the scale of it is boggle's mind like the training the amount of work done with like we've written all this custom software for training and labeling um and to do order labeling auto

95:55-96:06

leveling is essential um because especially when you've got like surround video it's very difficult to like label surround video from scratch is extremely difficult um like take a human such a long time to even label one video clip like several hours or the order label it basically would just apply like heavy duty uh like a lot of compute to the to the video clips um to pre-assign

96:06-96:66

and guess what all the things are that are going on in the surround video and then there's like correcting it yeah and then all the human has to do is like tweet like say this you know chan adjust what is incorrect this this is like increasing increases productivity by effect 100 or more yeah uh so you've presented teslabot as primarily useful in the factory first of all i think humanoid robots are incredible from a fan of robotics i think uh the elegance of movement that cuba um the humanoid robots the bipedal robots show are just so cool so it's uh really interesting that you're working on this

96:66-97:20

and also talking about applying the same kind of all the ideas of some of which we've talked about with data engine all the things that we're talking about with tesla autopilot just uh transferring that over to the just yet another robotics problem i have to ask since i care about human robot interaction so the human side of that so you've talked about mostly in the factory do you see it uh also do you see part of this problem that tesla bot has to solve is interacting with humans and potentially having a place like in the home so interacting not just sure not replacing labor but also like

97:20-97:85

i don't know being a friend or an assistant yeah i think the possibilities are you know endless yeah i mean it's it's obviously like a it's not quite in tesla's primary mission direction of accelerating sustainable energy but it is a an extremely useful thing that we can do for the world which is to make a useful humanoid robot that is capable of interacting with the world and helping in in many different ways uh so certainly in fact reason and really just

97:85-98:33

just i mean i think if you say like uh extrapolate to you know many years in the future it's like i i think uh work will become optional so like there's a lot of jobs that if you if people weren't paid to do it they would they wouldn't do it like it's not it's not fun you know necessarily like if you're washing dishes all day it's like you know even if you really like washing dishes you really want to do it for eight hours

98:33-98:88

a day every day probably not so um and then there's like dangerous work and basically if it's dangerous boring uh has like potential for repetitive stress injury that kind of thing um then that's really where humanoid robots would add the most value initially um so that's what we're aiming for is is to um for the human robots to do jobs that people don't don't voluntarily want to do um and then we'll have to pair that obviously with some kind of universal

98:88-99:52

basic income in the future so i think um so do you see a world when there's like hundreds of millions of tesla bots doing different performing different tasks throughout the world yeah i haven't really thought about it that far into the future but i guess that there may be something like that um so ask a wild question so the the number of tesla cars has been accelerating there's been close to 2 million produced many of them have autopilot i think we're over 2 million now yeah

99:52-100:11

do you think there will ever be a time when there will be more tesla bots than tesla cars yeah i i i i you know actually it's funny you asked this question because normally i do try to think pretty far into the future but i haven't really thought that far into the future with the with the tesla bot or it's code named optimus i call it optimus subprime because that's not it's not like a giant you know transformer robot um so uh but it's meant to be a general purpose

100:11-100:66

help help robot um and and basically like like the things that were basically like tesla i think um is the has the most advanced real-world ai uh for interacting with the real world which we've developed as a function of to to make self-driving work um and so along with custom hardware and like a lot of you know uh hardcore low-level software to have it run efficiently and be you know power efficient because because you

100:66-101:15

know it's one thing to do neural nets if you've got a gigantic server room with 10 000 computers but now let's say you just you have to now distill that down into one computer that's running at low power in a humanoid robot or a car that's actually very difficult a lot of hardcore software work is required for that um so so since we're kind of like solving the navigate the real world with neural nets problem for cars which are kind of like robots with four wheels

101:15-101:72

then it's like kind of a natural extension of that is to put it in a robot with arms and legs uh and actually you know actuators um so um like like the the two like the hard things are like you basically need to make the have the rower be intelligent enough to interact in a sensible way with the environment um so you need real real world ai and you need to be very good at um manufacturing which is a very hard problem tails is

101:72-102:37

very good at manufacturing and also uh has the real world ai so making the humanoid robot work is uh basically it means developing custom motors and sensors that that are different from what a car would use um but we also we have um i think with the the best expertise in developing advanced electric motors and power electronics so it just has to be for humanoid robot application on a car still you do talk about love sometimes

102:37-102:94

so let me ask this isn't like for like sex robots or something i love is the answer yes uh there is something compelling to us not compelling but we connect with the humanoid robots or even lego robots like with the dog and she shapes with dogs it just it seems like you know there's a huge amount of loneliness in this world all of us seek companionship and with other humans friendship and all those kinds of things we have a lot of here in austin a lot of people have dogs there seems to be a huge opportunity to

102:94-103:52

also have robots that decrease the uh the the amount of loneliness in the world or help us humans connect with each with each other so in a way that dogs can um do you think about that we test about it all or is it really focused on the problem of of performing specific tasks not connecting with humans um i mean to be honest i have not actually thought about it from the companionship standpoint but i think it actually would end up being it could be actually a very good companion

103:52-104:11

um and it could develop like a personality uh over time that is that is like unique like uh you know it's not like they're just all the robots are the same and that personality could evolve to be you know uh match match the the the owner or the you know or i guess the owner uh well whatever you want to call it uh the other the companion the other half right uh the same way their friends do see i think that's a huge opportunity i think

104:11-104:69

yeah no that's interesting like um the because you know like there's a japanese phrase i like the uh wabi-sabi you know uh the subtle imperfections are what makes something special and the subtle imperfections of the personality of the robot mapped to the subtle imperfections of the robot's human friend i don't know owner sounds like maybe the wrong word but um could actually make an incredible buddy basically and in that way the r2d2 or

104:69-105:27

like a c-3po sort of thing you know so from a machine learning perspective i think the flaws being a feature is really nice you could be quite terrible at being a robot for quite a while in the general home environment or all the in general world and that's kind of adorable and that's like those are your flaws and you fall in love with those flaws so it's it's very different than autonomous driving where it's a very high stakes environment you cannot mess up and so yeah it's more fun to be a robot in the home

105:27-105:78

yeah in fact if you think of like c-3po and r2d2 yeah like they actually had a lot of like flaws and imperfections and silly things and they would argue with each other and um were they actually good at doing anything i'm not exactly sure they definitely added a lot to the story um but but there's they're sort of quirky elements and you know that they would like make mistakes and do things like it was like uh it made them

105:78-106:29

relatable i don't know um endearing so so yeah i think that that could be something that uh probably would happen um but our initial focus is just to make it useful uh so that so um i'm confident we'll get it done i'm not sure what the exact time frame is but uh like we'll probably have i don't know a decent prototype towards the end of next year or something like that and it's cool that it's connected to tesla the car

106:29-106:84

the so so yeah it's using a lot of you know it would use the autopilot inference computer and um a lot of the training that we've done for the four cars in terms of recognizing real world things could be applied directly to the to the robot um so it but but there's there's a lot of custom actuators and sensors that need to be developed and an extra module on top of the vector space uh for love oh yeah that's missing okay okay add that to the car too

106:84-107:38

that's true um that could be useful in all environments like you said a lot of people argue in the car so maybe people can help them out uh you're a student of history fan of dan carlin's hardcore history podcast yeah that's great greatest podcast ever yeah i think it is actually i i it does it almost doesn't really count as a podcast yeah it's it's it's more like a audiobook yeah so you were on the podcast with dan i just had to chat with him about it he said you guys want military and all that kind of stuff uh yeah it's

107:38-107:92

literally uh it was basically um uh i think it should be titled engineer wars uh essentially like like when there's a rapid change in the rate of technology then uh engineering plays a pivotal role in in victory in battle um how far in back in history did you go did you go world war ii uh it was mostly well it was supposed to be a deep dive on fighters and bomber uh technology in world war ii um but that ended up being more wide-ranging

107:92-108:48

than that um because i just went down the total rat hole of like studying all of the fighters and bombers world war ii and like the constant rock paper scissors game that like you know uh one country would make this plane that would make a plane to beat that and that try to make plane to beat that and then and really what matters like the pace of innovation um and also access to high quality uh fuel and uh raw materials so like germany had like some amazing designs but they couldn't make them uh because they couldn't get their raw

108:48-109:09

materials and uh they they had a real problem with the oil and and and uh fuel basically the fuel quality was extremely uh variable so the design wasn't the bottom that goes uh yeah like the us had kick-ass fuel uh that was like very consistent like the problems if you make a very high performance aircraft engine um in order to make high performance you have to um the the the fuel the aviation gas uh has to be a consistent mixture and uh uh it has to have a high high octane um

109:09-109:65

like high octane is the most important thing but also can't have like impurities and stuff uh because you'll fill up the engine and and and german just never had good access oil like they tried to get it by invading the caucasus um but that didn't work too well [Laughter] that never works well for him that's nice for you so they always was germany was always struggling with [ __ ] with basically shitty oil um and then they could not uh they couldn't count on a on high quality fuel for their aircraft so then they had to add all they have all these additives and and stuff

109:65-110:23

uh so um uh whereas the u.s had awesome fuel um and that provided that to britain as well um so that allowed the british and the americans to design aircraft engines that were uh super high performance better than anything else in the world and germany germany could could design the engines they just didn't have the fuel and then also the likes of the the uh the quality of the aluminum allies that they were getting was also not that great and so yeah did you is this like uh you talked about all this with dan yep awesome broadly looking at history when you look

110:23-110:96

at genghis khan when you look at stalin hitler the darkest moments of human history uh what do you take away from those moments does it help you gain insight about human nature about human behavior today whether it's the wars or the individuals or just the behavior of people any aspects of history yeah i find history fascinating um there's a lot of incredible things that have been done good and bad um that they help help you understand

110:96-111:51

the nature of civilization um and individuals and does it make you sad that humans do these kinds of things to each other you look at the 20th century world war ii the cruelty the abuse of power talk about communism marxism and stalin um i mean some of these things do i mean if you like there's a lot of human history um most of it is actually people just getting on with their lives uh you know and and it's not like human history is just

111:51-112:05

uh what nonstop war and disaster is it's those are actually just those are intermittent and rare and if they weren't then you know humans would soon cease to exist um uh but it's just that wars tend to be written about a lot and whereas like uh something being like well a normal year where nothing major happened was doesn't get written about much but that's you know most people just like farming and kind of like living their life you know

112:05-112:60

being a villager somewhere and every now and again there's a war and i think so um and um you know what i'll say like that there aren't very many books that i where i just had to stop reading because it was just too too dark but uh the book about stalin the court of the reds are i could have stopped reading it was just too too bad dark rough

112:60-113:19

yeah um the 30s uh there's a lot of lessons there to me in particular that it feels like humans like all of us have that as the old soldier knits in line um that the line between good and evil runs to the heart every man that all of us are capable of evil all of us are capable of good it's almost like this kind of responsibility that um all of us have to to to tend towards the good and so like to me looking at history is almost like an example of look you have some charismatic leader

113:19-113:78

that uh convinces you of things it's too easy based on that story to do evil onto each other onto your family and to others and so it's like our responsibility to do good it's not like now is somehow different from history that can happen again all of it can happen again and yes most of the time you're right i mean the optimistic view here is mostly people are just living life and as you've often memed about the quality of life was way worse back in the day and keeps improving over time through innovation to technology but

113:78-114:44

still it's somehow notable that these blimps of atrocities happen sure yeah i mean life was really tough for most of history i mean really for most of human history a good year would be one where not that many people in your village died of the plague starvation freezing to death or being killed by a neighboring village it's like well it wasn't that bad you know it was only like you know we lost five percent this year that was uh yeah it was a good year you know that would be par for the course like just just not starving to death would have been like the primary goal of

114:44-115:00

most people in through throughout history just making sure we'll have enough food to last through the winter and not get in our freezer or whatever so [Music] um now food is is plentiful if i have an obesity problem um you know so well yeah the lesson there is to be grateful for the way things are now for for some of us we've spoken about this offline i'd love to get your thought about it here if i sat down for a long form in-person

115:00-115:66

conversation with the president of russia vladimir putin would you potentially want to call in for a few minutes uh to join in on a conversation with him moderated translated by me sure yeah sure i'd be happy to do that you've shown interest in the russian language is this grounded in your interest in history of linguistics culture general curiosity i think it sounds cool sounds cool now it looks cool so uh well it's it's you know it's it's a it takes a moment to read cyrillic um once you know what the sort like

115:66-116:19

characters stand for actually then reading russian becomes a lot easier because there are a lot of words that are actually the same like bank is bank [Laughter] um and uh so find the words that are exactly the same and now you start to understand cyrillic yeah if you can if you can sound it out then yeah it's much there's at least some commonality of words what about the culture you uh you love great engineering physics there's a tradition of the

116:19-116:83

sciences there sure you look at the 20th century from rocketry so you know some of the greatest rockets of the space exploration has been done in the soviet in the former soviet union yeah so do you draw inspiration from that history just how this culture that in many ways i mean one of the sad things is because of the language a lot of it is lost to history because it's not translated all those kinds of because it it is in some ways an isolated culture it flourishes within its within its borders um yeah so do you draw inspiration from those folks from from the history of science engineering there

116:83-117:39

i mean the soviet union russia and ukraine as well and uh have a really strong history in space flight like some of the most advanced and impressive things in history were done uh you know by the soviet union um so [Music] um one can cannot help but admire the impressive rocket technology that was developed um you know after the sort of fall of soviet union the

117:39-117:94

there there's the there's much less that that that happened um but uh still things are happening but it's not not quite at the um frenetic piece that was happening before the soviet union kind of dissolved into separate republics yeah i mean i i you know there's ross cosmos the russian agency a um i look forward to a time when those countries with china are working together uh in the united states they're

117:94-118:49

all working together maybe a little bit of friendly competition but i think friendly competition is good um you know governments are slow and the only thing slower than one government is a collection of governments so yeah the olympics would be boring if everyone just crossed the finishing line at the same time yeah nobody would watch yeah uh and and people wouldn't try hard to run fast and stuff so i think friendly competition is a good thing uh this is also a good place to give a shout out to a video titled the entire soviet rocket engine family tree by tim

118:49-119:06

dodd aka everyday astronaut it's like an hour and a half it gives a full history of soviet rockets and people should definitely go check out and support tim in general that guy is super excited about the future super excited about space flight every time i see anything by him i just have a stupid smile on my face because he's so excited about stuff yeah i love people like dad is real really great if you're interested in anything to do with space um he's in terms of uh explaining rocket technology to your average person he's awesome the best i'd say um

119:06-119:61

and um i should say like the part reason like uh i switched us from like rafter at one point was gonna be a hydrogen engine um but but hydrogen has a lot of challenges it's very low density it's a it's a deep cryogen so it's only liquid at a very you know very close to absolute zero requires a lot of insulation it's um so it is a lot of challenges there and um and i was actually reading a bit about uh russian rocket engine development and um

119:61-120:21

at least the impression i had was that that uh soviet union russia and ukraine uh primarily were uh actually in the process of uh switching to meth methylox um and there were some interesting tests and data for isp like they were able to get like up to like a 380 second isp with the meth lux engine and i was like whoa okay that's that's actually really impressive so um so i think we could you could actually get a much lower cost like in optimizing

120:21-120:83

cost per ton to orbit cost per time to mars it's uh i think methane auction is the way to go and i was partly inspired by the russian work on the test stands uh with methodolox engines and now for something completely different do you mind doing a bit of a meme review in the spirit of the great the powerful pewdiepie let's say one to eleven just go over a few documents print it out we can try let's try this i present to you document

120:83-121:62

numero uno okay vladimi paler discovers marshmallows that's so bad so you get it because uh yes are you failing things it's like i know three whatever that's not very good this is uh grounded in some engineering some history uh yeah give us an eight out of ten what do you think about nuclear power i'm in favor of nuclear power i think it's uh i i in a place that is not subject to

121:62-122:16

extreme natural disasters i think it's a nuclear power is a great way to generate uh electricity um i i don't think we should be shutting down nuclear power stations yeah but what about chernobyl exactly um so [Music] uh i think people there's like a lot of fear of radiation and stuff um and it's i guess what the problem is like a lot of people just don't unders they didn't study

122:16-122:84

engineering or physics so they don't it's just the word radiation just sounds scary you know so they don't they ha they can't calibrate what radiation means but radiation is much less dangerous than than you'd think so like for example fukushima you know um when the fukushima uh problem happened uh do the tsunami the i got people in california asking me if they should worry about radiation from fukushima i'm like definitely not not even slightly not at all

122:84-123:49

that is crazy and just to show like look this is how like the dangers is so much overplayed compared to what what it really is that i actually flew to fukushima and actually i donated a a solar power system for water treatment plant and and i made a point of eating locally grown vegetables on tv in fukushima like i'm still alive okay

123:49-123:99

so it's not even that the risk of these events is low but the impact of them is the impact is greatly exaggerated it's just human nature it's people who don't know what radiation is like i've had people ask me like what about radiation from cell phones according to causing brain cancer i'm like when you say radiation do you mean photons or particles then like that i don't know what what do you mean protons particles so do you mean uh if let's say photons what what frequency or wavelength and they're like no i have no idea um like do you

123:99-124:57

know that everything's radiating all the time like what do you mean like yeah everything's radiating all the time photons are being emitted by all objects all the time basically so um and if you want to know what it's it's what it means to stand in front of nuclear fire go outside the sun is a gigantic you know thermonuclear reactor that you're staring right at it are you still alive yes okay amazing yeah i guess radiation is one of the words that can be used as a tool to to

124:57-125:09

fear monger by certain people that's it and i think people just don't understand so i mean that's the way to fight that uh that fear i suppose is to understand is to learn yeah just say like okay how many people have actually died from nuclear accidents it's like practically nothing and uh say how many people have have died from you know coal plants and it's a very big number so like obviously we should not be starting up coal plants and shutting down nuclear plants just doesn't make any sense

125:09-125:78

at all coal plants like i don't know a hundred to a thousand times worse for for health than nuclear power plants uh you want to go to the next one it's really bad it's uh that uh 90 180 and 360 degrees everybody loves the math nobody gives a [ __ ] about 270. it's not super funny yeah i don't like two or three yeah um this is not a you know lol situation [Laughter] that's pretty good the united states oscillating between establishing and destroying dictatorships it's like a

125:78-126:37

metro is that a matchup yeah yeah yeah it's uh on a seven out of ten it's kind of true oh yeah this is uh this is kind of personal for me next one oh man this is leica yeah well no is this or it's like referring to like or something as like as like a husband husband yeah yeah hello yes this is dog your wife was launched into space and then the last one is him with his eyes closed and a bottle of vodka yeah like it didn't come back no they don't tell you the full story of you know what what the love the impact they had on the

126:37-127:04

loved ones yeah true that one gets an 11 for me oh yeah it just keeps going on the russian theme first man in space nobody cares first man on the moon well i think people do care no i know but um there is uruguay gardens names will will be forever in history i think there is something special about placing like stepping foot onto another totally foreign land it's it's not the journey like uh people that explore the oceans it's not as important to explore the oceans is to land on a whole new continent yeah

127:04-127:55

this is about you oh yeah i'd love to get your comment on this elon musk after sending 6.6 billion dollars to the u.n to end world hunger you have three hours um you know i mean obviously six billion dollars is not going to end with hunger so um so i mean the reality is at this point the world is producing uh far more food than it can really consume like we don't have a caloric uh

127:55-128:10

constraint to this point so where there is hunger it is almost always due to um it's like like civil war or strife or some like um it's it's not a thing that is it's extremely rare for it to be just a matter of like like lack of money it's like you know it's like some to the civil war in some some country and and like one part of the country is literally trying to starve the other part of the country um so it's much more complex than something that money could solve it's politics geopolitics it's it's

128:10-128:72

a lot of things it's human nature it's governments it's monies monetary systems all that kind of stuff yeah food is extremely cheap uh these days it's like it's um i mean the u.s at this point um you know among low-income families obesity is actually another problem it's not like obviously it's not hunger it's like too much it's you know too many calories uh so it's not that nobody's hungry hungry anywhere it's just it's just this is uh not not a simple matter of adding money and solving it

128:72-129:37

what do you think that one gets just this is going after empires world uh where did you get those artifacts the british museum shout out to monty python we found them yeah the british museum is it's pretty great i mean yeah it admittedly britain did take uh these historical artifacts from all around the world and put them in london but uh you know it it's not like people can't go see them uh so it is a convenient place to see these uh ancient artifacts is is london for you know for for a large segment of the

129:37-129:89

world so i think you know on balance the british museum is a net good although i'm sure a lot of countries argue about that yeah it's like you want to make these historical artifacts accessible to as many people as possible and the british museum i think does a good job of that even if there's a darker aspect to like the history of empire in general whatever the empire is however things were done this it is the history that happened you can't sort of erase that history unfortunately you could just become

129:89-130:35

better in the future that's the point yeah i mean it's like well how are we gonna pass moral judgment on these these things like it's like if uh you know uh if one is gonna judge say the rosh empire you're gonna judge you know what everyone was doing at the time and how were the british relative to everyone um and i think they would british would actually get like a relatively good grade relatively good grade not in

130:35-130:90

absolute terms but compared to what everyone else was doing um they were not the worst like i said you gotta look at these things in the context of the history at the time um and say what what were the alternatives and what are you comparing it against yes and i i do not think it will be the case that um britain would get a uh a bad grade in in when looking at history at the time you know if you judge history from you know from what is morally acceptable today you

130:90-131:44

basically are going to give everyone a failing grade yeah i'm not clear it's not i don't think anyone would get a passing grade um in in their morality uh of like you go back 300 years ago like who's getting a passing grade basically no one and we might not get a passing grade from generations but uh that come after us uh what what does that one get uh sure uh success for the monty python maybe i always love monkey python they're great uh brian and the quest for holy grail

131:44-131:99

are incredible yeah yeah yeah those serious eyebrows impressions like you know how important is facial hair to great leadership well you got a new haircut is that is that is this how does that affect your leadership i i don't know hopefully not it doesn't um yeah the second is no one there's no one competing with brush i have no one too those are like epic eyebrows so sure that's ridiculous six or seven i don't

131:99-132:69

know uh i like this like shakespearean analysis of memes he had he had a flair for drama as well like you know showmanship yeah yeah it must come from the eyebrows all right um invention great engineering look what i invented yeah that's the best thing since ripped up bread yeah because they invented they're just sliced bread am i just explaining memes at this point this is what my life has become um you know like a scribe that like runs around with the kings and just like writes down

132:69-133:13

memes i mean when was the cheeseburger inventor that's like an epic invention yeah like like wow you know that was versus just like a burger or a burger i guess a burger in general it's like you know um then there's like what is a burger what's a sandwich and then you start getting as a pizza sandwich and what is the original it's it's it gets into an ontology argument yeah but everybody knows like if you order like a burger or cheeseburger or whatever you like you

133:13-133:72

get like you know tomato and some lettuce and onions and whatever and you know mayor and ketchup and mustard it's like epic yeah but i'm sure they've had bread and meat separately for a long time and it was kind of a burger on the same plate but somebody who actually combined them into the same thing and yeah bite and hold it make makes it convenient it's a materials problem like your hands don't get dirty and whatever yeah it's brilliant well that is not what i would have guessed but everyone knows like you you if you order a cheeseburger you know what

133:72-134:29

you're getting you know it's not like some obtuse like i wonder what i'll get you know um you know uh fries are i mean great i mean they were the devil but fries are awesome um and uh yeah chip pizza is incredible uh food innovation doesn't get enough love yeah i guess is what we're getting at it's great um uh what about the uh matthew mcconaughey austinite here uh president kennedy do you know how to put men on the moon yet nasa no president kennedy be a lot cooler if you did

134:29-134:86

pretty much sure six six or seven i suppose that's the last one that's funny someone drew a bunch of dicks all over the walls sistine chapel boys bathroom sure i'll give it nine it's that's really true all right this is our highest ranking meme for today i mean it's true like how do they get away with that lots of nakedness i mean dick pics are i mean just something throughout history uh as long as people can draw things there's been a dick pic it's

134:86-135:49

a staple of human history it's a staple consistent throughout human history you you tweeted that you aspired to comedy you're friends with joe rogan might you uh do a short stand-up comedy set at some point in the future maybe um open for joe something like that is that is that really stand-up actual just blown stand-up full-on stand-up is that in there or is that i've never thought about that um it's extremely difficult if at least that's what the like joe says and the comedians say huh i wonder if i could um i mean only one way to find out you know i i have done

135:49-136:10

stand up for friends just uh impromptu you know i'll get get on like a roof uh and they they do laugh but they're friends too so i don't know if if you gotta call you know like a rumor strangers are they gonna actually also find it funny but i could try see what happens i think you'd learn something either way um yeah i kind of love um both the when you bomb and when when you do great just watching people how they deal with it is so difficult it's so you're so fragile up there it's just you

136:10-136:67

and you you think you're gonna be funny and when it completely falls flat it's just it's beautiful to see people deal with like that uh you might have enough material to do standard no no i've never thought about but i might have enough material um i don't know like 15 minutes or something oh yeah yeah do it do a netflix special netflix special sure um what's your favorite rick and morty concept uh just to spring that on you is there there's a lot of sort of scientific engineering ideas explored there there's the favorite one that's the butter robot

136:67-137:14

it's great uh yeah it's a great show yeah dragon ball is awesome somebody that's exactly like you from an alternate dimension showed up there elon tusk yeah that's right that you voiced yeah rick morty suddenly explores a lot of interesting concepts i so like what's your favorite one i know the the butter robot certainly is uh you know it's like it it's certainly possible to have too much sentience in a device um like you don't want to have your toast to be like

137:14-137:71

a super genius toaster it's gonna hate hate life because well it could just make us toast but if you know it's like you don't want to have like super intelligent stuck in a a very limited device um do you think it's too easy from uh if we talk about from the engineering perspective of super intelligence like with marvin the robot like is it j it seems like it might be very easy to engineer just a depressed robot like it sure it's not obvious to engineer a robot that's going to find a fulfilling existence same as humans i suppose but um

137:71-138:31

i wonder if that's like the default if you don't do a good job on building a robot it's going to be sad a lot well we can reprogram robots easier than we can reprogram humans so i guess if you let it evolve without tinkering then it might get sad but you can change the optimization function and have it be a cheery robot you uh like i mentioned with with spacex you give a lot of people hope and a lot of people look up to you millions of people look up to you uh if we think

138:31-138:88

about young people in high school maybe in college um what advice would you give to them about if they want to try to do something big in this world they want to really have a big positive impact what advice would you give them about their career maybe about life in general try to be useful um you do things that are useful to your fellow human beings to the world it's very hard to be useful um very hard um you know are you contributing more

138:88-139:44

than you consume you know like uh like can you try to have a positive net contribution to society um i think that's the thing to aim for you know not to try to be sort of a leader for just for the sake of being a leader or whatever um a lot of time people who a lot of times the people you want as leaders are other people who don't want to be leaders so [Music] um

139:44-140:03

if you live a useful life that is a good life a life worth having lived um you know and i like i said i would encourage people to [Music] use the mental tools of physics and apply them broadly in life there are the best tools when you think about education and self-education what do you recommend so there's the university there's uh self-study there is a hands-on sort of finding a company or a place or a set of people

140:03-140:66

that do the thing you're passionate about and joining them as early as possible um there's uh taking a road trip across europe for a few years and writing some poetry which uh which which trajectory do you suggest for in terms of learning about how you can become useful as you mentioned how you can have the most positive impact but i encourage people to read a lot of books just re like basically try to ingest as much information as you can

140:66-141:24

uh and try to also just develop a good general knowledge um so you at least have like a rough lay of the land of the the knowledge landscape um like try to learn a little bit about a lot of things um because you might not know what you're really interested how would you know what you're really interested in if you at least aren't like doing it peripheral exploration of broadly of of the knowledge landscape um and you talk to people from different walks of life and different uh

141:24-141:85

industries and professions and skills and like what occupations like just try you know learn as much as possible man search for meaning isn't the whole thing a search for meaning is yeah what's the meaning of life and all you know but just generally like i said i would encourage people to read broadly in many different subject areas um and and and then try to find something where there's an overlap of your talents and and what you're interested in so people

141:85-142:36

may be good at something but or they may have skill at a particular thing but they don't like doing it um so you want to try to find a thing where you ha you're that's a good a good uh combination of of your of the things that you're inherently good at but you also like doing um and um and reading is a super fast shortcut to to figure out which where are you you're both good at it you like doing it and it will actually have positive impact well you've got to learn

142:36-142:90

about things somehow so re reading a broad range just really read it you know the more important one is that kid i i read through the encyclopedia uh so that's pretty helpful um and uh there's also things i didn't even know existed or a lot so obviously it's like as broad as it gets encyclopedias were digestible i think uh you know whatever 40 years ago um so um

142:90-143:43

you know we read through the the condensed version of the encyclopedia britannica i'd recommend that um you can always like skip subjects or you read a few paragraphs and you know you're not interested just jump to the next one so read the encyclopedia we're just gonna skim through it um and um but you know i put a lot of stock and certainly have a lot of respect for someone who puts in an honest day's work uh to do useful things and

143:43-144:03

and just generally to have like not a zero-sum mindset um or or like have more of a grow the pie mindset like the if you if you sort of say like when when we see people like perhaps um including some very smart people kind of uh taking an attitude of uh like like doing things that seem like morally questionable it's often because they have at a base sort of axiomatic level a zero-sum mindset um and and they without realizing it they don't realize they have a zero-sum mindset or at least that they

144:03-144:65

don't realize it consciously um and so if you have a zero-sum mindset then the only way to get ahead is by taking things from others if if it's like if if if the pie is fixed then the only way to have more pie is to take someone else's pie but but this is false like obviously the pie has grown dramatically over time the economic pie so in reality you can have that so overuse this analogy if you have a lot of you can have there's a lot of pi yeah it's not fixed um so you really want to make sure you

144:65-145:13

don't you're not operating um without realizing it from a zero-sum mindset where the only way to get ahead is to take things from others then that's going to result in you trying to take things from others which is not not good it's much better to work on uh [Music] adding to the economic pie maybe you know so uh you know creating like i said creating more than you consume uh doing more than you yeah um so that that's a big deal um i think

145:13-145:74

there's like you know a fair number of people in in finance that uh do have a bit of a zero-sum mindset i mean it's all walks of life i've seen that one of the one of the reasons uh rogan inspires me he celebrates oh there's a lot there's not not creating a constant competition like there's a scarcity of resources what happens when you celebrate others and you promote others the ideas of others it uh it actually grows that pie i mean it the every like the uh the resource the resources become less scarce and that that applies in a lot of kinds

145:74-146:38

of domains it applies in academia where a lot of people are very uh see some funding for academic research is zero sum and it is not if you celebrate each other if you make if you get everybody to be excited about ai about physics above mathematics i think it there'd be more and more funding and i think everybody wins yeah that applies i think broadly yeah yeah exactly so last last question about love and meaning uh what is the role of love in the human condition broadly and more specific to you how has love romantic love or otherwise made you a better person a better human being

146:38-147:03

better engineer now you're asking really perplexing questions um it's hard to give up i mean there are many books poems and songs written about what is love and what is what exactly you know um you know what is love people don't hurt me um that's one of the great ones yes yeah you've you've earlier quoted shakespeare but that that's really up there yeah let me that was the many splendor thing uh

147:03-147:63

i mean there's um it's because we've talked about so many inspiring things like be useful in the world sort of like solve problems alleviate suffering but it seems like connection between humans is a source you know it's a it's a source of joy it's a source of meaning and that that's what love is friendship love i i just wonder if you think about that kind of thing when you talk about preserving the light of human consciousness right and thus becoming a multiplication multi-planetary species i mean to me at least um

147:63-148:20

that that means like if we're just alone and conscious and intelligent it doesn't mean nearly as much as if we're with others right and there's some magic created when we're together the uh the friendship of it and i think the highest form of it is love which i think broadly is is much bigger than just sort of romantic but also yes romantic love and um family and those kinds of things well i mean the reason i guess i care about us becoming a multi-planet species in a space frank civilization is

148:20-148:83

foundationally i love humanity um and and so i wish to see it prosper and do great things and be happy and um and if i did not love humanity i would not care about these things so when you look at the whole of it the the human history all the people has ever lived all the people alive now it's pretty we're we're okay on the whole we're pretty interesting uh bunch yeah both things considered and i've read a lot of history including the darkest worst parts of it

148:83-149:37

and uh and despite all that i think on balance i still love humanity you joked about it with the 42 uh what what do you think is the meaning of this whole thing is like is there a non-numerical oh yeah well really i think what douglas adams was saying in hitchhiker's guide to the galaxy is that um the universe is the answer and uh what we really need to figure out are what questions to ask about the answer that

149:37-149:84

is the universe yeah and that the question is the really the hard part and if you can properly frame the question then the answer relatively speaking is easy so so therefore if you want to understand what questions to ask about the universe you want to understand the meaning of life we need to expand the scope and scale of consciousness so that we're better able to understand the nature of the universe and and understand the meaning of life

149:84-150:40

and ultimately the most important part would be to ask the right question yes uh thereby elevating the role of the interviewer yes exactly as the most important human in the room good questions are you know it's a hot it's hard to come up with good questions uh absolutely um but yeah like it's like that that is the foundation of my philosophy is that um i i am curious about the nature of the universe and uh

150:40-150:99

you know and obviously i will die i don't know when i'll die but i won't live forever um but i would like to know that we are on a path to understanding the nature of the universe and the meaning of life and what questions to ask about the answer that is the universe and um and so if we expand the scope and scale of humanity and consciousness in general which includes silicon consciousness then that you know there were that that seems like a fundamentally good thing elon like i said i'm deeply grateful that you would spend your extremely valuable time with me

150:99-151:57

today and also that you are given millions of people hope in this difficult time this divisive time in this uh cynical time so i hope you do continue doing what you're doing thank you so much for talking today oh you're welcome uh thanks for your excellent questions thanks for listening to this conversation with elon musk to support this podcast please check out our sponsors in the description and now let me leave you with some words from elon musk himself when something is important enough you do it even if the odds are not in your favor

Key Themes, Chapters & Summary

Key Themes

  • SpaceX and Human Spaceflight

  • Mars Colonization and Governance

  • Tesla Autopilot and Self-Driving Challenges

  • Cryptocurrencies and Financial Systems

  • First Principles Thinking and Innovation

  • Engineering Challenges in Rocketry

  • AI, Robotics, and Perception in Driving

  • Digital Transformation of Government Systems


Chapters

  1. Introduction to Elon Musk and Lex Fridman Podcast

  2. SpaceX Achievements and Human Spaceflight

  3. The Road to Mars: Colonization and Challenges

  4. Governing Mars: A New Frontier

  5. The Evolution of Tesla Autopilot

  6. Cryptocurrency: Bitcoin, Dogecoin, and Beyond

  7. Engineering Insights: The Raptor Engine

  8. AI and Robotics in Autonomous Driving

  9. Reimagining Government and Legislation

  10. Conclusion: Innovation and Future Visions


Summary

The podcast conversation between Lex Fridman and Elon Musk covers a wide range of topics, primarily focusing on Musk's ventures with SpaceX, Tesla, and his views on AI and robotics. Musk shares insights into the complexities and achievements of SpaceX, particularly emphasizing the historic significance and challenges of human spaceflight, such as the Crew Dragon Demo-2 mission. He discusses the engineering marvels and hurdles in creating SpaceX's rockets, specifically the Raptor engine, and the overarching goal of making spaceflight more accessible and sustainable.


Musk also delves into his vision for Mars colonization, elaborating on the necessity and logistics of establishing a self-sustaining human presence on Mars. He touches on the potential for different governance models on Mars, suggesting a direct democracy with transparent and simplified legislation processes.


In the realm of AI and robotics, Musk reflects on the advancements and challenges in Tesla's Autopilot and Full Self-Driving systems. He acknowledges the complexity of replicating human driving behavior through digital systems, emphasizing the intricate process of translating visual data into actionable navigation decisions. Musk also discusses the role and potential of cryptocurrencies, particularly Bitcoin and Dogecoin, in modern financial systems, weighing their benefits and limitations.


Throughout the conversation, Musk’s approach to problem-solving is evident, showcasing his emphasis on first principles thinking, efficiency, and innovation. He stresses the importance of rethinking traditional models, whether in space travel, governance, or digital currencies, to achieve breakthroughs and address contemporary challenges.