2022年7月9日 星期六

Tesla Owners Silicon Valley: Elon Musk Unfiltered: Interview Part 3 (Bonus Material)

 http://creating-cashflow.blogspot.com/2022/06/blog-post.html

看見 Ricky介紹的ELON MUSK 精彩訪問, 找來中英文 transcript, 可以快速看到內容。

,

 https://www.youtube.com/watch?v=u5w_VkAx6tc

0:11

i want to especially thank you for starman putting starman and roadster into space was the most life-changing

0:17

thing i've ever seen um also i wanted to give you this and thank you for allowing this to happen this is the starman begins episode we

0:23

just made just a gift thank you thank you thank you for i mean like the

0:29

course we're on that i was a space geek as a kid and i didn't believe anything was gonna happen because in the 90s

0:34

nothing was happening yeah and it i didn't even pursue an engineering career because there was nothing to do in space

0:40

but when i saw that i was like i knew that this was real and pretty much changed everything i was doing to how i

0:46

make this all right cool thank you that's this is great um yeah and do you you know that how how it

0:52

came to be that arose it was in the in the falcon heavy i've heard bits of internal lore but never from you so i

0:58

would love to hear this well normally when there's a new rocket that's launched uh there's something very boring like a chuck of concrete or

1:05

something is is just says because you're not going to risk an expensive satellite uh on the first launch of a new rocket

1:11

um so but i was like hey guys we we can't put a chunk of concrete that's going to be

1:17

super boring that's what boeing would do literally they did do it in fact they literally i think it was a chunk of

1:23

concrete or something that they put on their first launch uh you know if the delta iv or something um

1:28

so i'm like now we gotta spice it up um so i was talking to my friend uh jonah nolan uh who's like uh

1:35

like you wrote like the the really good batman movies and west world and stuff like crystal's brother

1:41

you know so oh wow um and i was in his kitchen and he suggested that we should put a tesla

1:47

in and i was like oh okay that's a good idea and so i was like well i've got one in my garage i can use that one so

1:54

um you know it's not serial number one or anything it's like it's a later like 1500 serial number or something like

2:00

that and um so literally the car that i was driving around la is now in orbit around earth and mars

2:06

you could track it yeah uh so

2:12

but it was basically like let's have something that looks cool instead of boring concrete and uh

2:17

and i i really thought there was like quite a good chance the rocket would fail fail you know like we've had a number of

2:23

rocket failures in the past so there's so much that can go wrong with the falcon heavy that's kind of like surprising that it all worked like it

2:28

was like holy crow i can't believe it all worked this is insane that was my reaction uh

2:33

with the falcon heavy launch and then the the star man was like well we've got a car gotta put someone in the car you know

2:40

otherwise like who's driving it you know and uh so that's where we put the star man in in the car

2:46

and then there's the tiny star man yeah and that was that was a friend of mine uh

2:52

uh nora who who was like yo she put like a small i wish i put like the matchbox tesla

2:59

on the you know on the dashboard with a tiny star man in it i'm like okay we'll do it

3:06

so then like i don't know what the aliens are going to think when they discover it like like like clearly they worship this

3:12

thing and then when they made a small version of it because they worship it so much it's like like like the stuff like

3:19

that we just discovered of ancient civilizations probably the answer might be like much more like you know less serious than we think it's

3:26

like oh this is like we just think everything's a temple that they worshipped but maybe it's not it was just like some

3:32

thing that they made you know all the way down exactly totally i also have one selfish question

3:38

car sharing obviously we're working towards full autonomy that's the main goal but there seems to be a 0.5 step

3:44

and i'm actually a car sharer i have eight teslas and one vivian okay sharing on turo and i just feel like there's so

3:50

many ways that like the model y specifically is optimized and cyber truck in the future will be even more optimized for car sharing

3:57

is there someone working on that a tesla you know you know our issues for the last few years have been

4:03

not uh you know car sharing like i mean i guess like like a rental car you know is kind

4:10

of a car sharing situation yeah yeah we've just been trying to keep the

4:17

factories operating the last couple years has been a very difficult thing and like supply shaded eruptions have

4:23

been severe like extremely severe so it has required uh all of our attention just to

4:31

uh keep the factories going not not to think of like like i think like the super hard part for a

4:38

car company is like how do you get revenue above cost so you don't go bankrupt and

4:44

so it's not really like a you know as long as as long as our demand is enough to

4:50

absorb the production uh and we have production enough to cover

4:56

our fixed expenses then we're in an okay spot that's right but i'm thinking about like downtime if these cars are

5:01

traveling 400 000 miles sitting in someone's driveway they're not going to collect 4000 miles before the paint's kind of having issues i imagine that

5:08

increasing the utilization of the existing fleet obviously full self driving does that but up until then if we have car sharing

5:16

installed to increase utilization increase the number of miles on these vehicles it's a good thing for the world

5:22

i i said that this is that that is a that's a high-class problem that that

5:28

uh you know um the the thing that like car companies

5:34

they should think of car companies as like at any given point they desperately want to go bankrupt

5:40

okay that a car company is desperately trying to go bankrupt at any given point in time

5:45

um so uh in order to have that not be the case you have to

5:50

have the fact that factories had to have to be active otherwise uh you have parts piling up in

5:55

warehouses all over all around the world and you can't ship the but but if you're missing any parts you can't finish the

6:01

car and ship it um so the past two years have been an

6:07

absolute nightmare of supply chain interruptions one thing after another and we're not out of it yet

6:13

uh so um

6:19

so that that's that's overwhelmingly our concern is how do we

6:25

keep the factories operating so we can pay people and not go bankrupt and then everything else is like nice to

6:31

have but um you know uh

6:37

yeah even recently like the kobe shutdowns in china were very very difficult

6:43

yeah you guys actually so well that we all think of you as a tech company first which has like you know like alec apple

6:49

right they make hardware but they spend way more money on software which i i'd even say i've got the perception sometimes that it's like oh yeah things

6:55

are they've got this just huge stack of bandwidth but like you're right the amount of money and hardware moving through this place is

7:02

like this factory is losing insane money right now because we can't make it

7:07

because we um

7:12

like we should be uh outputting a lot more cars from this factory versus a very puny amount of cars

7:17

um but we had challenges with the 4680 ramp and with the structural pack ramp

7:23

and the and then ironically the tooling because we had we were able to do 2170 cells but the

7:30

tooling necessary for making 2170 variant cars was stuck in china

7:35

so with this shanghai factory in op

7:40

we've got the the tooling to enable this factory stuck in port in china with no one to actually

7:47

move it uh which basically then caused this

7:52

factory's production ram just to be very tiny uh now this is all gonna get fixed

7:57

real fast you know but it requires a lot of attention um

8:03

and we'll take like i said it'll take more effort to get the production get this factory to high value production than it took to build it in

8:09

the first place and the same is true of berlin berlin's in a slightly better position because

8:15

berlin started off with the 2170 style and did not have the 4680 and structural back risk

8:21

so so but both berlin and austin factories

8:28

are gigantic money furnaces right now yeah okay there should be like a giant

8:35

roaring sound which is the sound of money on fire okay that's what dumpster fire

8:41

bigger than a dumpster is too small

8:46

gigafactory level like berlin and austin are losing billions of dollars right now

8:53

because there's a ton of expense and and hardly any output

8:58

so having getting berlin in austin functional and getting shanghai back back in the saddle fully are

9:05

overwhelmingly our concerns yeah uh everything else is a

9:12

very small thing basically understanding that this is not an immediate concern but does a plaid model

9:18

3 make sense to you in your mind uh no

9:23

could the weight even handle that power strain um like its potential

9:30

we're trying to reduce complexity [Laughter]

9:39

yeah and we're our own worst enemy in adding complexity

9:44

um i mean in the grand scheme of things the model s and x are we're doing them not

9:50

because they are huge money owners but really for uh sentimental reasons

9:56

the like they're at at uh full output we're talking about 100 000 cars a year for s and x

10:04

but three and y will do i don't know two million plus so

10:10

it's like 10x the you know or more frankly i think three and y i could probably do

10:16

three or four million and snx will top out at a hundred hundred thousand just because the

10:21

affordability drops off exponentially with price so as as price rises you

10:26

basically the number of people that can afford an s or an x it doesn't drop by a little it drops by an order of magnitude

10:33

so like a doubling of the price will drop the affordability by factor of 10

10:38

as a rough rule of thumb yeah is there any additional thing that we

10:44

can continue to do i mean we tried our best with the the solar taxing i mean it's probably hard even like i'm like

10:50

asking you like how can we help it's kind of a weird it feels like a weird question but well i i mean the public support is certainly appreciated um

10:57

and uh shooting down fight is is always very helpful um you know the

11:03

you know and there's an argument for maybe we should advertise because the you know you know the

11:10

the sort of traditional media will not run negative pieces about automotive because automotive is like one of the

11:16

biggest if not the biggest advertisers in their in the in their paper so like the ads could literally be nothing

11:23

but it would be better because the articles that they're supposed to be bad yes like this is like like so tesla is

11:29

like basically free game whereas uh it's safe to say that uh if they run

11:34

some negative piece about general motors right next to a general motors ad general motors uh marketing exec he's

11:41

going to call them up and say um why did he do that do you want me to pull it now uh yeah

11:46

exactly yeah it's like uh yeah your budget uh next year we'll be spending nothing with your publication

11:52

yeah exactly that and they don't have to do that because

11:58

uh they they they know so there is tacit there's a test and understanding uh that you don't run hit pieces on your

12:04

major appetizers yeah it's bad for business yeah right and and another people's like oh there's don't worry there's a there's a

12:11

fire there's a truck trying to eat wall between the business and the editorial i'm like yeah but they also know who's

12:17

paying their salary okay they don't need like uh they they they're not they're not idiots

12:22

there's a yeah

12:28

exactly they just know don't go there that's what they know um but they can go to they can trash

12:35

tails like it tells us an advertise so that's the issue um so maybe we should advertise and at least like you know

12:42

uh i don't know it's sort of it doesn't feel right but

12:47

that is something we could do i mean in the grand scheme of things though like the if you look at like if if if you

12:52

ignore some of the sternum drum and say just like okay well is it actually causing us to sell fewer cars

12:58

uh it doesn't seem to be um so then then why should we care land is not an

13:03

issue so they i mean they can actually i mean frankly a bunch of the tr the pieces trashing

13:09

tesla you know as like like especially like say like

13:14

2017 through 2019 the price was just a non-stop hate stream it was like

13:20

hate like tesla's a fraud test is a failure because it's a fraudulent failure

13:25

they're like you know the thumbing through the thesaurus were like bad words to use it was like insane um

13:34

you just like click on like the the phone like the stock app ticker and i remember like for three years straight

13:40

it would just be like a headlines just trashing tesla just over and over and over again and our sales went up so i'm

13:47

like okay yeah so like i guess keep trashing us you know uh

13:54

because at the end of the day people were like oh yeah i just kept reading about tesla i don't know why but i just kept reading about it i can't remember

14:01

what the article was about but i know i read about tesla 17 times [Laughter]

14:09

is good press or whatever yeah there's like this old price is good press there's some truth to that you know so because like

14:16

generally you say like uh you know of the articles that you read about a newspaper like last week how much

14:23

content can you remember about that about those articles really it's just the headline it's sentiment yeah the

14:28

headline vaguely we weren't reading the articles yeah yeah even if they are i don't think there was

14:33

a big movement on twitter like please read the article before you retweet it

14:41

everyone's just forward retweeting and forwarding based on the freaking headline even the content says the exact

14:46

opposite of what the headline is and and the headline chosen by the editor and and the you know so the writer is

14:51

like i i didn't pick the headline you know i you know literally said the total opposite of what the headline was and

14:56

people don't even know because they just boarded the headline yeah um i mean actually one of the best one of

15:02

the things that should happen in twitter is like it's like uh like like every time you click on a news article you get a paywall and it's like

15:08

you do yeah yeah it's super annoying yeah yeah exactly

15:14

so you can just see the headline and you literally can't read through that article [Laughter]

15:20

so there's a little sign there it says can't read article yeah and it's like it's like i i don't want

15:26

to get a subscription to philadelphia inquirer but i'm sure they have everything in a good article that i'd like to read but i don't want like 18

15:33

like 12 000 subscriptions and and also they don't even make it easy to get a subscription like any like these this

15:39

like you know 18-page process where you like tell us everything about your entire life is what's required to subscribe to xyz

15:46

newspaper in some whatever like man who's going to do that so imagine what

15:52

it would do and then i keep losing my password for the ones and then it opens like twitter's emulated browser yeah it doesn't even

15:58

log in anyway yes exactly it's like i like actually like like i have a subscription to the new york times in

16:04

the wall street journal but like i like open i click on the twitter link and i'm like uh it's saying that i'm not a

16:09

subscriber yeah so then after like click go to the web web server have it redirect to the app on my phone and it

16:14

says cockamamie nonsense right this feels like a problem for apple though i was literally just about to say where it should be i wonder

16:20

it's just

16:27

[Laughter]

16:35

even as you have the subscription it doesn't figure it out and i mean yeah this is totally

16:40

dismissed if twitter got in the middle and algorithmically started um um de-prioritizing content that did that

16:48

eventually i'll say what he was saying about apple twitter could ultimately end up like apple which is like you want your content that's subscribed to be

16:54

that way great users can subscribe to twitter to whatever news is and when you click it's good yeah twitter can i mean

16:59

honestly would make a better user experience if you did get in the middle of that i would might not pay i might pay a dollar a month to the atlantic is

17:04

part of a five dollar a month twitter plan i paid yeah it gets me all the subscribed news right now clicking a link and going to it i just lose my

17:10

[ __ ] mind yes i'm going to reddit to go find the [ __ ] link with somebody like pokemon totally

17:16

uh be cool to see if that ever happened one day no i i think that's the thing that should happen no that's not the irony of that

17:21

like there's definitely like like a a loss of value here that is is simply missing

17:28

uh where it's like i understand that all these publications want to maximize their subscribers and kind of like own

17:33

their customer and stuff but there's there's a huge number of people who are never going to subscribe to that

17:40

publication like let's say you know not a particular philosophy required but let's say uh you're live in philadelphia

17:46

so most of the time you don't want to read the articles but every now and again they've got like some reporter on the scene at some news event and you and

17:53

and they they're actually the source article everyone else is just copying them and printing their stuff and so you want to

17:59

so you want to just like read the philadelphia enquirer article and but but you're not there's no chance you would become a subscriber so they're not

18:05

losing a subscriber but they are annoying everyone who's not going to be a subscriber so there's got to be some way to solve that sort of

18:12

quarter yeah ten cents for this article or something

18:17

exactly one dogecoin for this article uh you know it should be like i'll pay

18:22

i'll pay 10 cents for this article whatever 15 cents 20 cents it's like fine um

18:28

and and uh you know and that should work music that's how it should work music artists get paid like that already it's

18:34

more listen based on like yeah scaled out over a lot of dimensions exactly news is going to have to go micro payments there's just like this

18:40

exactly and and there's no there's there's no losers here like in fact the media company will be a winner because

18:45

they will get incremental revenue over and above those who are never subscribed to their publication that's right uh and and then we can all read news and not

18:51

have to be stuck behind pale worlds over mad house i mean we don't even have uh a a an app that's as good as wechat in

18:58

china uh and like in china you can like live on wechat basically uh yeah it's like

19:03

yeah everyone everyone's like they're like you live on wechat you do payments you do everything it's like yep it's great basically checks kick ass um and

19:10

we don't have anything like wechat outside of china that's true so i was like my idea would be like how about if

19:15

we just copy wechat hey copy them

19:21

twitter copies wechat yeah i guess better user experience

19:27

well before you have to go we should probably get a selfie or a picture with you okay thank you

19:33

thank you for watching what are you doing these days what am i doing youtube and uh i review electric cars for a

19:38

living i break out those cars on turo car sharing and i'm considering what it would look like to build an app

19:45

that makes that experience better because turo is good but it is not specialized and so

19:50

no there's so many things that could do better for the tesla experience yeah if you build that i'll use it and autonomy is just such a monumental

19:57

breakthrough that it's just it's just car sharing what does not matter really i get that yeah i get that

20:02

but the point five step of what other learnings maybe do you have there as far as maintaining cars who's responsible

20:08

those kinds of things that's where i think it happens yeah no i think this i mean there's there will need to be like sort of a shepherd of cars exactly

20:15

that's why i'm trying to be at this point car shepard yeah years ago not to try that

20:21

no i mean so this is not like this until we're there it's not like this like cars still need to be like maintained

20:26

like if it's like you know you gotta clean the car you gotta you know if it gets scuffed or damaged

20:32

it's gotta be it's like literally like tending a flock of cars yeah very personally yeah exactly so

20:38

uh they'll still it's it's like i mean i think it will effectively like for people that are currently like uh uber

20:44

or lyft drivers it'd be like okay like instead of you driving yourself you manage i don't know 50 cars or something

20:49

like that and you're taking 10 to your flock exactly yeah basically ryan has 52 cyber trucks on

20:56

order oh really okay complete in the bay area we need to get on that i mean

21:01

the design at least finally is is locked and and we're for sure locked yeah it's locked

21:07

uh for sure locked pencils down uh yeah seriously though

21:12

we got too carried away with the anyway that that features features

21:18

blade runner um blade runner yeah just kept changing yeah um

21:24

i mean last year the chip shortage was so bad that even if we had uh it wouldn't matter you could just you're just robbing one

21:30

pocket to pay the other pocket like yeah so you could and we even couldn't make the stationary storage stuff because

21:35

stationary storage batteries were they used uh chubs too yup so then we had

21:41

like basically a star of the powerwall line uh to make cars and even then uh couldn't make all the

21:48

cars we wanted to make so and then go straight from that into this

21:53

china cover shutdown which was really like not just affecting

21:58

shanghai production but also like there's still some parts that are made in for cosmetic california that are with

22:05

the parts made in china and it was and we had tooling stuck in china as well so

22:10

i think i'm not sure everyone knows how just how serious that the co-acceptance were trained up there

22:16

and it's not done yet by the way it's not like major green light yet it's

22:22

still bumping bumpy yeah it's tough to ramp up if you were to guesstimate

22:28

when would the cyber truck hit production i think middle next year okay roughly

22:35

12 months ish okay i'm excited excited we are for it i'm so excited

22:44

oh really yeah absolutely i stopped having kids at four because i

22:49

can't grow i can't have a cyber charge going past that so stop that four smart man

22:57

six seats you outdid your contribution to society exactly yeah i mean good for you it's like i mean man so many of the friends i

23:04

know have like zero or one kid yeah uh that's why i'm like i'm always banging the baby drum

23:09

i'm like man civilization's gonna you know collapse and no big deal

23:14

yeah yeah like where do you think people come from like some magical [ __ ] people factory

23:21

it's the dork the stork the stork the baby baby store i mean like

23:28

it's they gotta come from somewhere and they take ages to grow uh you know so unless uh

23:36

you know uh like the thing that tends to happen is like once the birth rate starts going down it seems to accelerate in a

23:42

downward trajectory like japan is leading indicator here yeah saying they're about to collapse

23:47

no population nutty they lost six hundred thousand people last year that negative six hundred thousand

23:53

and japan is like pretty close to being the uh the long most like they're live expectancies like 85.

23:59

so like that's the only thing that was more towards keeping the population plummeting before

24:05

but now despite having like the world's longest lifespan they're still losing the population oh wow saying something

24:10

yeah i was saying something so uh seems to me the rate of that decline is

24:16

such one that's going to be not really able to be turned around we kind of got a plan for what we do in spite of it

24:21

happening yeah like well if that train continues japan is going to flat out disappear as as well

24:27

most countries in the world so it's not like you know

24:33

i'm just saying if that trend continues hopefully it doesn't but if it does this is going to be i don't know what chad's going to do

24:39

like chad's going to have like a period of time of great prosperity of like the next 10

24:44

years uh and then but transit like roughly half replacement rate you know maybe

24:51

because of the one child policy right they they got rid of the one child policy several years ago yeah then made

24:57

it a two child policy now it's a three child policy okay growth rate unchanged really yes well it's so ingrained in

25:03

them for whom you know interesting things matter too oh and japan never had a one-time policy and

25:08

they still weren't that bad yep so uh it's just look at the trends it's

25:14

like basically with with with education and openness urbanization yeah both great plummets yeah and it's

25:22

there's people like oh people can just afford it nope actually the places that have the best social safety nets have

25:27

the worst birth rate yeah okay yeah that's not the reason yeah it says inverse is true right the

25:32

poorer you are the more kids you have yeah and mostly i think there's a social economic study that says the

25:38

reason why is because parents need comfort i i mean you can say like one could like say

25:44

figure out what whatever causes it's hard to say what the cause would be but but the the

25:51

the correlation cannot be argued with yeah and the the more religious the less educated and

25:57

the poorer the higher the growth rate yeah so like if for lower if like this lower religious

26:05

uh low on the religion high on education uh and at high on income that has the

26:11

lowest birth rate yeah yeah i think if we parse the data there's probably an argument to be said

26:16

how we got to the side religious had a large impact on getting to the size of population we did because if you study religion you know like most of them

26:22

preach the fruitful go go grow numbers yeah go forth and multiply yeah

26:36

i like that [Laughter]

26:44

so yeah hopefully the things turn around um yeah um

26:50

i think the current trend is is not good from a birthday standpoint i'm trying to set a good example

26:57

another reason why we did another civilization on another planet yeah

27:02

the number of humans is it's not training well um

27:08

i think a lot of people are under the impression that like you know the current number of humans is unsustainable on the planet and this is

27:13

totally untrue yeah uh we could double number humans and be okay i still keep the rainforests

27:19

people forget we're animals if we actually couldn't sustain it we would just die off yeah

27:24

like the population density is actually very low and most parts of the world have no people in them so if you like fly

27:30

over the country you know from like l.a to new york or something and he looked down and said like what percentage of

27:36

the time if you dropped a bowling ball would you hit someone here we know can i see you like round

27:41

the second oh really yeah yeah but like you can just know a few cities

27:46

and

28:00

hit a person or die you're going to die

28:05

because you're not going to succeed in any question even in la like you think la's got a lot of people but you look at the actual if

28:11

you like just look to say what cross-sectional area is human like like the cross section of like

28:19

you know you take like several blocks in la and and and make make the houses and cars translucent

28:25

and and look at okay well from looking at from above what percentage of the cross-section of the area is is humans

28:31

it's almost nothing that's below one percent in l.a wow interesting yeah have you ever seen

28:37

that giant ball of flesh they did that if you combine every human into new york city you've seen it right the big ball of goo of like all humans

28:43

pushed into one thing it would like sit in new york's uh what's the big park in new york central it would sit in central

28:49

that's it and like it's a little bit taller than the tallest skyscraper it's not very big at all like a cuboid of humans

28:57

[Laughter]

29:16

on a single floor you could fit all humans uh in new york city without on on one floor without any even having what

29:23

do you mean on one floor i mean like oh surface area yeah how thick would that human pie be no it's just one human

29:30

tall i'm gonna say there's no human without stacking without stacking it's just side by side you mean yeah just say like what's the cross-sectional error of

29:36

a human yeah say okay there's eight billion of those what's what's the area of new york city oh it's less than

29:42

you can find stanford all the humans in new york city yeah i love the phrase cross-sectional area of a human actually there's

29:47

something yeah from space it's just we're targeting and the cross-section is low

29:53

so we're going to have to just coarse in the missile because of that low cross section of humanity

29:59

that's that's basically what it comes down to um so yeah anyway yeah it's like if people

30:05

like living in some bustling city they think humans must be everywhere but actually uh

30:10

there are very few cities like like no place in america has dead the density of manhattan and then even if you've got

30:16

like it's sort of what seems like a dense city like la the actual cross-sectional area that's human is small like less than one percent of the of the

30:23

area is human so that means that if you if you dropped a cannibal exclu that doesn't bounce too much

30:30

the probability of hitting a human is extremely low wow that's interesting we've had pieces from rockets and multiple nations re-entered

30:36

the tanks and crashed uncontrolled and not hit anybody many times when we had this baseball

30:42

tragedy like that on entry it it it rained debris across the entire united states

30:49

and and i don't think there were any injuries at all yeah like the entire yeah a lot of debris

30:56

i cried and nothing people were scared oh my god it almost hit me yeah okay they weren't bowling balls

31:04

like they recovered most of the material from the shell um wow but at the end zero people were they were hit i believe

31:12

and here's an interesting part of that is it was going not vertical it's going around the earth which increases the probability so it's it's it's it's

31:18

coming in from orbital so it's going power up more or less parallel to the surface and and and and start start breaking up

31:25

uh i believe over the us like a crop dust or at least certainly like like none of the parts fell in the ocean it

31:30

all fell on the con on the land i believe uh um so uh that's it because it's it's when

31:38

when you're going orbital you're going parallel off the earth's surface you know roughly 25 times the speed of sound

31:43

so cells decelerating and breaking apart uh acid and the debris is raining down on

31:50

the u.s and no people were here so

31:55

there are actually very few humans on earth um i have one fsd related question would

32:02

it be possible to give for us to get access to some folks to give narrative feedback from the experience we're having because some of us drive a [ __ ]

32:08

ton of miles cross country somewhere about the texas and back and there's things that we saw that there's no

32:14

button or email i could possibly send to say hey here's what here's the issues we're seeing here's how they're breaking down and more importantly here's how

32:20

they're breaking down across contexts right like if there was any way to if it would be helpful if it's not then no

32:25

worries but if there's a way to give that narrative feedback to somebody i think it might help

32:32

i mean our goal right now is to get to zero interventions in in point point drives in cities yeah

32:39

so the polish is less important than that solving for zero intervention

32:46

point one drives in cities um the sheer amount of work required to do

32:51

this boggles the mind yeah um yeah and to the best my knowledge i asked the team like do we know of anyone who is

32:57

even trying to pull their nice biceps we don't know anyone maybe there is someone but

33:02

the the i mean i i've seen a lot of tough technology

33:09

problems and and solving real-world ai such that a car can drive itself is one of the hardest

33:14

problems i've ever seen and and it is way harder than i originally thought by far

33:19

um so what would you say to the people that have said

33:25

for several years you've said it it'll be this year it'll be next year what would you say to those people it'll be this year

33:31

we got it put that out quote for six months

33:37

yeah no i i think it's still tracking for this year but as far as the i'm going more for like the uh yeah why did

33:44

i where was i overly optimistic in the past uh i did not i just understand the just the scope of the problem and you

33:50

feel you do now yes um i mean the

33:57

it's it's so it it's very easy to get to get close to it working and where it feels like it's

34:04

just you know all right there just a little further you know um

34:10

and it's like that you know the donkey with the carrot over over here it's like that's right these keep you know just

34:16

pull the card a bit further and faster pull this card faster maybe that cash gonna come closer um

34:22

so the and and frankly um having radar and ultrasonics uh was a

34:30

mistake radar especially because radar uh it allows you to get close to solving it

34:37

and solve it most of the time except when you can't lock the radar and and the

34:44

visual neural net like radar radar and vision disagree which one is right

34:49

um and so you basically have to get rid of the radar um

34:55

and once we've read the radar which by the way was initially strongly opposed by the autopilot team and now nobody wants

35:01

radar back is that right yeah i have to lay down the low and say like that radar is coming out man

35:08

yeah uh it is it you can't that radar is a crutch and if you're carrying that crush you can't run so

35:14

uh it just it the radar would would it just the signal that

35:21

radar was contributing more noise than signal at the end of the day uh so so that wasn't great contradicting

35:27

signal but it was it was more noise than signal and so that's that's uh

35:33

it had to be removed um and once radar was removed then it became clear that uh actually our neural nets are much worse

35:39

than we think they are so they were being they're being helped by radar um too much

35:46

um so now yeah just having to use vision

35:52

uh you you you there's nothing you can't hide behind anything the vision has to

35:57

work the neural net vision has to work and then there was quite a profound

36:03

change to go from bag of points neural net to

36:09

having a neural net interpret that bag of points to figure out what a lane center is because previously we're doing that all

36:15

in c so you know

36:21

i mean the neural net architecture is insanely complicated

36:26

at this point it's like it's a lot of layers and and then

36:32

then you then it'll turn out oh actually we need to delete this layer put this layer on and and like we we've reacted

36:38

re-architected the neural net so many times here's a question i kind of had from the outside it's like it seems like tesla

36:44

has like one general autopilot team why not have two compete against jill you

36:49

did with boca chica and florida why not have one that you know still has radar

36:54

and all the old data points and a separate team and make them compete

37:00

i think the team does not lack for motivation level or work ethic they they work super hard and

37:07

the the tesla autopilot uh sort of software

37:13

ai team is like the best software team i've ever worked with um i'm a host judge of technical skill

37:19

and they are uh outstanding it took a while for us to get there we had a lot of false starts with the

37:25

autopilot team um i was like in silicon valley there's like i don't know like 12 people who say

37:31

they're responsible for autopilot but they're not um

37:40

some generally good people on that team too i still hang out with some of them golf with some of them ski with some of them good people

37:47

not often they don't have a lot of time

38:02

yeah the other day like the the streets were designed for

38:08

uh eyes and biological neural nets that's what the entire road system was designed for

38:14

so therefore it makes perfect sense frankly that cameras and digital neural nets

38:20

are required to solve self-driving you basically have to

38:26

repeat what the humans do but in silicon and and nothing else will actually solve

38:32

it so i mean tesla's actually i mean going to be in a situation where tesla has a self-driving solution and no

38:39

one else is even close not even for five years licensing yeah yeah sure absolutely

38:46

we're not going to stop people from doing um but the they will it

38:52

it will need to slap them in the face like that not like hey we think

38:58

like it has to be taking sales away from them oh it'll take sales away from uber it'll take sales yeah everything

39:05

so it'll be very clear it's actually delivered it'll be very clear that this

39:11

is the way everyone will know in that moment within a month this is the way if they don't already

39:17

know right anyone who has those the that you guys have the beta you you've seen the trajectory of improvement and

39:23

it's very clearly gonna uh achieve full self driving yep yeah it's just a question that there's some debate about

39:28

when when yeah but but but if you say like what is the rate of improvement per unit time yeah you can say this is clearly going to get to a point which is

39:34

where it's much safer than human yeah um is how do those you know basically there

39:40

is that point you cross and it feels as if we are kind of slowing the progression and i want to know that it

39:45

could oh okay so that the the thing that's that's sort of hard to appreciate from sort of outside the

39:52

company is that we'll re-architect the neural net like for the thousandth time

39:58

and and uh um

40:03

i mean that's like the the the things that i like they take up the most amount of

40:09

brain space for me are working on self-driving and getting starship to orbit those are

40:14

two things that take up like i don't know 70 percent of my brain

40:19

space like something like that those two things um and there's other things that

40:25

that just don't require as much sort of cp brain cpu cycles but like a lot of chores basically

40:32

if i do my chores then [ __ ] gets the fan basically i hate dog chores um

40:39

so um

40:49

anyway just like basically autopilot is absorbing a massive amount of my brain cycles that that that and getting

40:55

starship to orbit um or two things that are overwhelmingly uh

41:01

that's a super majority of brain cycles the um

41:06

so the reason that it like the say 12 that's one or whatever didn't seem like

41:12

it was that great uh was because we refactored the so many of the neural nets uh that that like to the outside it

41:19

seems like two steps forward two steps back um and so it feels like like like the improvement is flat but actually a whole

41:25

[ __ ] ton of like we just reacted with the neural net so uh like it's a word for having like

41:32

like an a and a c architecture to having a c and an a architecture but now

41:37

now you can improve room to grow yeah you got it exactly so uh you know the

41:44

they still we still don't have a unified vector space uh where all the neural nets

41:50

pour their output into a single agreed upon uh vector space for

41:56

uh both fixed and moving objects so that's huge yes that's quite

42:02

important um because like if they disagree like let's if the if the moving objects

42:08

neural net like currently this they're really so fairly separate the moving objects network and the static objects network

42:16

uh if they disagree on the position of a car then then like you could say well this this car is in a non-drive space

42:22

position because of a disagreement between the neural nets and and then if they're not running at the same frame rate you can get a time

42:29

error as well and if so if cars are you know moving at

42:35

i don't know uh 30 meters per second or something then and and you uh

42:42

you know have like a sort of 33 34 millisecond error okay like that car has

42:47

now moved a lot you know so it's moved a meter you know so

42:53

uh a meter is a lot like if we're not colliding with something it's a ton yeah it can be like so this is this car

43:00

like you're driving along let's say these two there's two this car is going on opposite directions if you

43:06

if you have if that car coming towards you is offset by a meter that could be the difference between colliding or not

43:11

so um and so there could even be like a slight a time lag error you know between one

43:17

year on the next and you can have an issue so it's uh no no any limit as these neural nets get very

43:23

good they will be much much better than a human so the it will be able to do to drive

43:29

better than james bond could possibly drive unfortunately the threshold for uh

43:36

succeeding is just being bad on a human and humans have uh really a lot of latency in general

43:41

so so this it's like not actually all that high of a bar when you realize it it's

43:46

sort of like you think like elevator operators back in the day mm-hmm and and uh yeah you can be an expert elevator operator but you gotta flip that relay

43:53

and get the and get the elevator and the floor to line up exactly it's like uh

43:59

you're not going to succeed nearly as much as the computer so now now anyone if you did have an

44:04

elevator operator with big big relay switch you'd think that's pretty backwards um

44:09

and and you'd really prefer to have a button that's how it would be with with the

44:14

autopilot that's good analogy yes like like normal human drivers will be like an elevator operator that's a good analogy yeah yeah

44:22

and the precision of waterpilot will be insane after hitting the um

44:28

no interventions on city streets what's the next benchmark that comes after that and i'm assuming it's on

44:34

some number of miles some number of drives happening because like if we're saying true zero then you've reached autonomy right like true absolutely it's

44:41

gonna be it's gonna be extremely rare uh to have an intervention i mean driving

44:47

uh from my friend's house to your factory today i had no interventions um and actually that's

44:53

austin has you know either in city streets or on the highway um so uh the one only one thing i say is like

45:00

the the the speed that the car decides on was wrong so i had to like adjust speed several times which as i

45:06

understand from feedback from the autopilot team that is considered an intervention right did you have you hit the pedal accelerator no it's not i'm

45:11

talking about accelerators okay cool the car basically was generally going too slow gotcha

45:17

like it's and there's there's definitely going to be like a collision with reality here because uh you know just like the

45:23

rolling stop in california everyone does rolling stop like who doesn't you know yeah

45:28

yeah you practically get rear-ended if you don't yeah yeah like there's no one at the end

45:34

there's like literally you know what the intersection and you squeeze to a halt it's kind of weird is that

45:39

uh you know so but then technically that's against the law so you know nasa got all up in arms about it and said oh

45:45

you have to come to a full zero hull and then like

45:52

uh because of some rule about like the speed like like like you're allowed you can

45:57

never show the speed below what it really is and so you have to earn the side of it being slightly

46:03

above what it really is like so sometimes it would show one mile an hour even for briefly even though the wheel

46:08

was stopped [Music] because of nitsa rules that's speed

46:14

and and and and then somebody sent in a complaint so then we had to literally show a picture of the

46:19

wheel a video of the wheel while showing see it says one mile an hour but that's

46:24

because of your rules that make us round up to one mile an hour but the wheel is not moving

46:30

literally it's gonna descend on the video to prove that the wheel's not the car wheel is not moving even though it

46:36

says one mph and that says one mph because of their rules so wow uh

46:43

we're getting a lot of of sort of complaints involved and whatnot from competitors about autopilot

46:49

because they have no answer to it [Music]

46:54

and they're not like thrilled about the idea of like licensing it from tesla that's not like oh yeah let's give our

47:00

competitor a ton of money yeah it's not the one their number one objective because the value is just ridiculous yes the

47:07

value of a fully self-driving car is like we've never seen before yeah yeah

47:12

why do you think everyone undervalues uh self-driving it seems like the world doesn't really

47:18

understand the value of it why is that well it's from like i think if someone's

47:26

never seen a unicorn or something you know they're like how's it happening like universe don't exist you know it's

47:31

just a horse with a horn on its head no big deal it's a magical creature you

47:38

know that's what basically self-driving sounds like some magical a fiction thing to most people

47:46

until you actually use it and then close your mind yeah so now it wouldn't be that hard to use it

47:51

they could just like talk to someone who's in the beta program in the car right now yeah seriously

47:57

it wouldn't be that hard to figure it out but they're not trying yeah um and and so like they're all just

48:03

thinking that no uh self-driving is like it's very far away slash they have invested in some

48:09

solution that that will solve it which we want no one is pursuing a pure vision

48:14

ai approach to this which also requires like a massive amount of labeling mm-hmm

48:21

like custom auto labeling software uh you know a massive array of training

48:27

computers it's not just the in-car software it's all of the training and labeling and debugging software that rule wrote us we

48:34

wrote all that too like it didn't exist got it the sheer amount of software it

48:39

tells us written uh just in software tools and debugging is is uh

48:45

insane um and you know take like being like taking the frames of video and like

48:51

analyzing this video and seeing ah this is where that it got the error and this is why when the order labeling software

48:56

needs to be corrected and this like order labeling is in and of itself a massive thing yeah

49:01

huge yeah and we have like i think 1500 human labelers wow wow yeah

49:08

but but they're amplified probably by a factor of a thousand by the order

49:13

labeling okay at least a factor of 100. wow so it's more like having 150 000 right exactly

49:20

uh it's like it's a lot yeah um

49:26

that's huge it is huge yeah uh and the speed of the aura labeling means that

49:32

like the rate at which you can order label and then so when you order label an entire video

49:37

segment uh you know that that's like thousands of frames that get order labeled and then

49:45

the all the human has to do is like say oh yes that's correct or nope um tweak that line slightly over

49:52

that's it and then and then it corrects the order label and you feed that back into the order

49:58

labeling uh basically you train the order labeler to order label and then you have fewer

50:04

fewer errors over time yeah trust that human was correct no we don't

50:09

there's a lot of there's a lot of training examples and you just trust that that aren't that that the error is statistically

50:16

yes like that it's rare for the error to occur a number of the improvements have been

50:22

us going through the order labeling software and saying like why is this

50:28

why is the probability of of this thing not much higher than it that we think it should be and and and

50:34

there are a number of cases it's been because of a very serious mislabeling error so that so then the

50:40

what the neural net will think is that the vast majority of the time the line is over there but every now again it's over here

50:46

and it's like no it's still over here that was just a labeling error so it actually tries to compensate for

50:52

for uh you know it's it's a statistic so it's like oh there's there's a it's

50:57

there's an oil distribution oh there's a little hump over here no there's not a little hub over here that's just mislabeling delete

51:05

now everything's on now it works so so some of these most like a significant

51:11

mis-labeling error uh it can have quite a significant reduction in the effectiveness of the neural net yeah

51:17

for sure one thing um so i was walking through i was at target the other literally yesterday i was walking through the parking lot and this old

51:23

lady was backing up of course wasn't looking out at all for me and she almost hit me um and what i was gonna say is is that a

51:30

lot of the people that are gonna be saved from full self-driving won't know it right whether it's my tesla

51:37

dodge all right all right we were we're at best buy we're going to best buy today already way through other people

51:42

yes they're you know a car yeah so it's crazy yeah yeah unsung

51:51

i mean when we started download autopilot path i was actually taught somebody said like oh yeah well you know

51:56

there's like somebody that said like well you're not going to be you're going to get no praise for the whatever 90 people you

52:02

save and you're going to be sued by the 10 people you don't so this is a thankless task yes exactly

52:08

so it's like uh but the other day i think we just have to put this before jury and say listen you got to believe statistics here uh

52:15

yeah because it's like the end if this is like you know i think highway safety is like an order of magnitude safer than

52:21

people and it's like it's like these are not small numbers here you know you just say objectively how many crashes were

52:27

there oh it's ten times fewer okay so like then the company that

52:32

massively reduces uh deaths and anti-accidents should not be penalized for the few that do occur

52:39

that's like that would be super messed up kill full autonomy safety is it's going to happen

52:44

even even with full autonomy there's still going to be a non-zero number of things because this you're a lot like

52:49

fall well like you're still interacting with uh you know like the it's going to take a

52:56

long time before the the fleet is autonomous it's like when the fleet is autonomous let's say at that point the problem like

53:01

accidents will be insanely rare like 100 times less or more than what they are right now like you just have to

53:07

literally leap into the traffic with no chance of the car like like practically a suicide you know um

53:15

so uh when all the cars are autonomous but for a long time uh there will be autonomous

53:21

cars mixed in with 99 point something percent non-autonomous cars

53:26

and then uh that that means people will do you know we'll swerve across traffic like so you go down the road and

53:32

somebody you know with a high speed non-divided road and a truck

53:38

just drives into you yep okay like uh were you supposed to do

53:44

um yeah um and uh but i mean i think people also

53:50

realize like just how many people are dying out there from uh you know

53:55

basically drivers falling asleep or whatever i mean jay's wife was killed by a truck driver

54:00

that fell asleep he forgot about that she was on a bike on a bike yeah yes and it's actually a

54:06

truck driver fell asleep yeah the truck veered across the road and ran her over killed him yeah

54:11

yeah and if that car if that truck had had a water pilot uh it would have stayed in late and it would not it should be alive

54:18

today wow the irony of that yeah so anyway like either way we're just going

54:23

to keep doing it um it's you know yeah yes got to do it

54:29

let's just take take the heat i don't know one last thing it won't take as much

54:36

alpha waves or mind share but a lot of people a lot of fans and owners want to

54:41

know about apple music you've been avoiding my question on that so i stopped asking

54:46

[Laughter] and what are we looking for here

55:05

yeah it could definitely be better uh come on man just thank you seriously

55:14

i mean you should just like like sync with your calendar and and you shouldn't have to type anything yeah um cloud profiles i'd have to use

55:21

my calendar for that to work yeah let's say we have the calendar sync working pretty well i don't know i just

55:26

don't i just don't i just don't use the calendar that's a joke yeah yeah for me it's a hit and miss but

55:32

yeah like for a while we had it working well i think i haven't used it in a while so but but it should basically just

55:38

automatically take you where you want to go without you ever say anything yeah and just look at your calendar and yeah yeah go based on that yeah

55:44

essentially so spotify title amazing thank you oh my god holding the new

55:50

model s when you're actually connected to wi-fi yeah yeah right before the cyber rodeo we were in a parking garage with wi-fi oh my god yeah really

55:56

unlocked yeah title is great i love it but balafan's don't have title but they do

56:02

have apple music so they've just been asking so yeah really is it like are you talking about like the

56:08

apple's like car implementation no no no no no no you're saying literally so spotify has a subscription that you can

56:14

pay to like subscribe to it yeah apple has a apple music that used to consult oh you mean just not through the phone

56:20

basically yes correct to the car yeah exactly yeah so you want to so just to avoid bluetooth yeah correct exactly

56:27

because then it's still yeah yeah there's a fair amount of customers who

56:33

use apple music okay versus spotify that's it that's true yeah

56:38

and i use apple i actually one of the rare people who i use apple music and spotify um

56:45

but i i just use spotify through bluetooth i mean so i should i choose apple music through bluetooth oh you do

56:50

so if i play apple music i just bluetooth it oh um yeah

56:55

it's doesn't bother you sound quality uh i mean the the sound quality

57:04

i it's like we can definitely improve the bluetooth data rate if that's the

57:09

the bottleneck we're talking about i mean that would make everyone happy but not just the apple music builders sure yeah

57:16

but i use spotify but i still do it through my phone because like that's just how i do it using my house i do it through my phones when i get on my sonos

57:22

so i get in the car and it goes through

57:34

no i'm i'm just sorry i'm just still in off like just this whole experience yeah yeah

57:40

sorry you caught me one of those blank moments

57:50

we already have waypoints so i have no feature to request cars

57:59

one day but who doesn't i have one so all right i'm a peasant now with

58:05

legacy s or have a legacy p 100 d so the blind spot camera when you right

58:12

click or when you use the blinker it doesn't show up so when you reverse though

58:19

you know the repeater cameras they would show up they show up in you know all the new ones but i'm on the legacy x

58:25

it's 2018 december show's up yes so it's already there it's just it

58:31

you talk about the future on the three when you hit the blinker on it'll show yes i love that i got a rental and i got

58:36

me i mean elon you responded to me to do it and then it just didn't get on my car

58:41

so it's like okay it seems like a small number of people but that's a

58:46

feature too that's why i said i'm a peasant but i'm not gonna it's fine i mean i'm not saying like our resource

58:52

allocation is is perfect because it certainly isn't but uh you i mean one does need to stack rank

58:57

these the issues by like how much good will this do times number of people yeah

59:02

um and and we we have a lot of uh long list of things uh

59:08

so so the and pretty far down the list is is like uh you're adding features to

59:13

uh legacy yeah products that's why i said i'm a peasant a different way of saying that is upgrade

59:19

oh come on i got free lifetime supercharger man it's yeah it's just it's just like it doesn't it doesn't make sense for you know

59:26

future people to yeah make sense like if there was like some

59:31

factory that just produced great software engineers like you that you just ordered like i would need a hundred

59:36

grade software engineers they're then no problem but there isn't like there's a very small number of great software

59:42

engineers makes sense and a bunch of the ones that i know are great uh got rich and stopped working

59:49

right you probably know a few of those yeah you're trying to bring the price down no stop leaving

59:54

yeah i mean i know a lot of very capable people and once they got

59:59

rich they just stopped working sure it's like i don't you know it's their life but

1:00:05

they stop coding and they just uh i don't know what they do actually

1:00:12

like watch them say they invest and stuff i but you know

1:00:18

there's yeah i mean like there's a lot there's a lot of really capable people yeah who are

1:00:23

similar retired that that i know um so

1:00:29

i'm just like the weirdo who's like i still work like a maniac because i'm just wired that way so

1:00:35

but most people i know they just pass a certain amount of money there their work amount of work they're here

1:00:42

is much less well this is the interesting thing for this interview i've never actually heard you say that like basically you're an existentialist

1:00:47

i've not heard you say that directly i've always thought that actually basically studying in high school yeah right around the same time i discovered

1:00:53

tesla and i was inspired tied to that and so it's just interesting that like yeah i don't see you ever really retiring just because of

1:00:59

the fire that is driving you to work right now and yeah yeah i just don't see it happening

1:01:04

yeah exactly so it's like trying to you know

1:01:10

i'd just try to figure out what's going on

1:01:16

um yeah yeah and then and then just like take take the set actions that are most

1:01:22

likely to enable uh us to figure out what's going on

1:01:28

and understand the universe um yeah

1:01:34

i think if you're not an alien i just realized that you're trying to become an alien you're trying to pursue and find other life potentially you're coming yes

1:01:40

we're trying to be the alien yeah yeah are you talking about the entire time i've never thought of it that way

1:01:46

yeah at some point yeah yeah astronauts are kind of just aliens yeah

1:01:53

i mean it's very it's very you know the furry paradox is a profoundly important question

1:01:59

um and there's a book that's like i don't know 50 answers to the fermi paradox or

1:02:05

um something now and again they put out a new one when they can think of some other answers to the funny paradox it's quite

1:02:11

good um because it's very troubling that we do not see scientific science of aliens

1:02:18

very very troubling like something doesn't add up the universe why is the universe 13.8 billion years old and we

1:02:23

see no evidence of of any alien visitation whatsoever

1:02:28

nor can we detect any signals or anything like that yeah nothing the only thing i can think is it's time bound

1:02:35

we're the last one that's currently alive and there may be future ones that become alive or at least can be

1:02:40

perceived as alive in the future yeah well actually like the universe is like say 13.8 you know go like basically thousand

1:02:47

eight hundred and something million years old that's when you talk about that well if you even wanna increment the

1:02:52

like the third digit past decimal point that would be a million years in the future and our civilizations as well since the

1:02:59

start of writing is five thousand years old so

1:03:04

and given current birth rates too yeah like

1:03:09

i think so do you think we're allowed

1:03:15

a million years to be 200 times the length of civilization yeah what was that one yeah do you say do you

1:03:21

think we're alone i'm certainly stating the you know i think we should adhere to the

1:03:27

scientific method which is that we have no evidence of evidence yeah like

1:03:32

it's not yeah just there is nothing that we have no evidence of aliens life whatsoever

1:03:39

nothing nothing nothing flat damn nothing yeah no signals no you know like

1:03:44

all i would have taken would be for like an alien civilization to to put like

1:03:50

like a one inch cubed of pure titanium

1:03:57

like just like just a titanium cube i'll be like [ __ ] aliens man okay because they sure as [ __ ] they sure as

1:04:03

hell didn't know how to how to how to refine deoxidize and and create like a cube of pure titanium oh what is that oh

1:04:09

pure anything pure nickel how about that so in my mind alien life it will if we if and when we find the especially the

1:04:14

first one we'll be more like a bacteria we probably won't find uh you know human-like able to modify the world

1:04:20

around them i think we'll find bacteria yeah so i think this this likely like if

1:04:26

you look at earth earth was was just basically archaic bacteria for a very long time

1:04:32

um and and it was a huge thing to have mitochondria get incorporated into a cell and then have

1:04:39

that mitochondria also divide when the cell divides that's so gigantic

1:04:44

um and you just can't have a complex organism really of any size unless

1:04:50

you've got a little power plant with the mitochondria in your cell yeah so but with the myocardium basically seems

1:04:57

to have been like uh some bacteria archaic bacteria that got captured and then essentially

1:05:03

when it's symbiosis with a thing but but then the crazy thing is like how does it replicate itself when the when the cell

1:05:10

divides the mitochondria must divide at the same time how does the mitochondria absorb

1:05:16

that so so there must have been like a crazy number of times where mitochondria was absorbed but did not

1:05:22

do not replicate itself when the cell did i think that's why i think we'll find bacteria but i guess the follow-up question is how would we even perceive

1:05:30

that if it's further than our solar system how do we perceive bacteria as life if it was outside our

1:05:36

solar system because that's what gives me hope is that it's just so small or different

1:05:42

that we just haven't yet been able to look forward yeah well and i think generally we're

1:05:48

looking for like life we can talk to you know type of thing because it doesn't matter if it's life versus life it can

1:05:54

talk to you yeah i mean if we find a planet of trilobites you know like okay that's still incredible it's not quite

1:06:00

awake it's cool it's cool can't talk to them simple shell creatures

1:06:08

and they're probably different maybe maybe different from different simple shell creatures uh okay we found

1:06:13

a wide range of different symbols shell creatures um you know

1:06:20

then i think like that seems quite likely frankly yeah um there we go yeah no no i mean i say the probability

1:06:26

from a probability standpoint i think that's likely yeah um it's like how many civilizations that

1:06:32

can actually you can actually talk to that are around at the same time as we are around

1:06:38

those appears to be a vanishingly small number maybe zero that are in this galaxy

1:06:44

and we don't really know of a way to get to other galaxies with current physics would you say given potentially all the

1:06:51

unknowns what is your thoughts on there being like a creator

1:06:59

you know in general i'm just saying like what like if you're gonna uh aim for an evidence-based uh approach

1:07:05

to existence then you simply say what is the evidence and you your conclusion should be

1:07:11

probabilistic according to the evidence um

1:07:16

so there's clearly some like like we don't know what the origins of the universe

1:07:21

are um we don't know if this is just is this a simulation always somebody's video game

1:07:27

i mean if you look at how at the progress of video games going from simple blocks with pong to

1:07:33

you know 3d re you know massively multiplayer photorealistic video games

1:07:39

in our lifetime if that trend continues video games will be instinctual from reality

1:07:46

so not saying our video games we're saying that this is a remarkable situation we witness

1:07:55

you can still be at a single turn of civilization yeah i mean that's the wild thing right

1:08:01

is there's the history of it finding the way something something causes the universe

1:08:06

to come to being you know there's like some uncaused cause or something like

1:08:12

this some like uh originate it's like it came from somewhere it's a whole several question is like is

1:08:17

is there someone some invisible creature or invisible deity uh judging our

1:08:23

actions and deciding whether we're going to go to a good place or bad place when we die that doesn't seem to be the case

1:08:33

or at least it's not it's it's not clear what the rules are for

1:08:38

because then there'd be a lot of like you know mass murderers and stuff that should have been

1:08:43

i don't know not enough in a good place so i don't know it's hard to say maybe

1:08:49

if people get get very defensive about religion because it's like you know they'll sort of identify

1:08:54

themselves by a particular religion i'll put out there if you kind of drive yourself by

1:09:00

existentialism it's like replacing that for them yeah it's replacing their purpose for

1:09:06

being well it's it's certainly uh it would be a change

1:09:13

um to say like let's pursue uh an expansion of consciousness

1:09:21

uh in order to understand the nature of reality whereas if you say no i do i do

1:09:27

understand the nature of reality because i believe in this this religion or that religion and that

1:09:32

that is that is my explanation for reality which is how a lot of people would would view a few things yeah

1:09:38

um so i'm like okay you know i mean it's

1:09:43

uh you don't one doesn't want to get into a religious debate it's just it's just

1:09:49

it's just going to be a different path that people follow if they you know take religion

1:09:55

at face value versus uh take the position that we don't know what's going on and we want to find out

1:10:01

what's going on yeah so it's very different different uh set of actions

1:10:07

and i guess what i'm saying like religion lets people believe that they know what's going on already

1:10:12

yeah um uh you know i don't really have any issue with religion uh provided that

1:10:19

we just generally go in a direction that furthers consciousness yup

1:10:27

and at the end of the day if a religion results in more people being created and then and

1:10:33

that the atheists aren't having kids then well it's going to be a self self-fulfilling prophecy

1:10:39

yeah fate loves irony yeah that's the only people around will be religious people

1:10:48

all right okay sure

English (auto-generated)

AllListenable

 

 

0:11

我要特別感謝 Starman 將 Starman 和 Roadster 送入太空是最能改變生活的

0:17

我見過的東西，嗯，我也想給你這個，謝謝你讓這發生，這是星人開始的一集，我們

0:23

剛做了一個禮物謝謝謝謝謝謝我的意思是喜歡

0:29

當然，我們認為我小時候是個太空極客，我不相信會發生任何事情，因為在 90 年代

0:34

什麼都沒有發生，是的，我什至沒有從事工程事業，因為在太空無事可做

0:40

但是當我看到我就像我知道這是真實的並且幾乎改變了我所做的一切

0:46

讓這一切都很好，謝謝你，這很好，嗯，是的，你知道它是怎麼回事

0:52

是在獵鷹重型飛機中出現的，我聽說過一些內部傳說，但從來沒有從你那裡得到

0:58

當有新的火箭發射時，通常會很樂意聽到這個，嗯，有一些非常無聊的東西，比如一塊混凝土或

1:05

有些事情只是說出來，因為你不會在第一次發射新火箭時冒著昂貴的衛星的風險

1:11

嗯，但是我就像嘿伙計們，我們不能把一塊混凝土

1:17

超級無聊，這就是波音公司真正會做的事情，他們確實做到了，事實上他們確實做到了，我認為這是一大塊

1:23

混凝土或他們在第一次發射時放置的東西，呃，你知道是 delta iv 還是什麼

1:28

所以我想現在我們得給它加點情調，嗯，所以我在和我的朋友說話，呃，喬納諾蘭，呃，他就像

1:35

就像你寫的像真正優秀的蝙蝠俠電影和西方世界，像水晶的兄弟這樣的東西

1:41

你知道哦，哇哦，我在他的廚房裡，他建議我們應該放一輛特斯拉

1:47

我當時想哦，好吧，那是個好主意，所以我覺得我的車庫裡有一個，我可以用那個，所以

1:54

嗯，你知道它不是序列號一或任何東西，它就像是後來的 1500 序列號或類似的東西

2:00

那和嗯，從字面上看，我在洛杉磯附近駕駛的汽車現在在地球和火星的軌道上

2:06

你可以追踪它是啊，所以

2:12

但基本上就像讓我們有一些看起來很酷的東西而不是無聊的混凝土，呃

2:17

我真的認為火箭失敗的可能性很大，你知道，就像我們有很多

2:23

過去的火箭失敗，所以獵鷹重型火箭可能會出現很多問題，這有點令人驚訝，它都像它一樣工作

2:28

就像神聖的烏鴉，我不敢相信這一切都奏效了這太瘋狂了，這是我的反應

2:33

獵鷹重型發射，然後明星人就像我們有車要派人上車，你知道的

2:40

否則就像誰在開車，你知道，嗯，這就是我們把明星放在車裡的地方

2:46

然後是那個小明星，是的，那是我的一個朋友

2:52

呃，諾拉，她就像你一樣，她放的像個小我希望我放的像火柴盒特斯拉

2:59

在你知道的儀表板上，裡面有一個小明星我很好，我們會做的

3:06

所以我不知道外星人發現它時會怎麼想 就像顯然他們崇拜這個

3:12

然後他們製作了一個小版本，因為他們非常崇拜它，就像像這樣的東西

3:19

我們剛剛發現的古代文明可能答案可能更像你知道的沒有我們想像的那麼嚴重

3:26

就像哦，這就像我們只是認為一切都是他們崇拜的寺廟，但也許不是它就像一些

3:32

他們讓你完全知道的事情我也有一個自私的問題

3:38

汽車共享顯然我們正在努力實現完全自主，這是主要目標，但似乎有 0.5 步

3:44

我實際上是一名汽車共享者，我在 turo 上有 8 特斯拉和 1 輛 vivian 可以共享，我只是覺得有

3:50

很多方面都特別喜歡 Model y 的優化，未來的 Cyber​​ Truck 會更加優化汽車共享

3:57

有沒有人在研究特斯拉，你知道你知道我們過去幾年的問題一直是

4:03

不，你知道汽車共享，我的意思是我想就像租車一樣，你知道是善良的

4:10

汽車共享的情況是的，是的，我們一直在努力保持

4:17

過去幾年的工廠運營一直是一件非常困難的事情，就像供應陰影爆發一樣

4:23

非常嚴重，所以它需要我們所有的注意力，只是為了

4:31

嗯，讓工廠不要不去想像我想的那樣超級困難的部分

4:38

汽車公司就像你如何獲得高於成本的收入，這樣你就不會破產

4:44

所以這並不像你知道的，只要我們的需求足夠

4:50

吸收生產，嗯，我們有足夠的生產來覆蓋

4:56

我們的固定費用然後我們處於一個好的位置這是正確的但我正在考慮如果這些車是停機時間

5:01

坐在某人的車道上行駛 400 000 英里他們不會在油漆出現問題之前收集 4000 英里我想那

5:08

增加現有車隊的利用率顯然完全自動駕駛可以做到這一點，但在那之前，如果我們有汽車共享

5:16

安裝以提高利用率 增加這些車輛的里程數 這對世界來說是件好事

5:22

ii 說這就是那那是一個高級問題

5:28

呃你知道那個喜歡汽車公司的東西

5:34

他們應該將汽車公司視為他們迫切希望破產的任何時候

5:40

好吧，一家汽車公司在任何給定的時間點都拼命地試圖破產

5:45

嗯，為了避免這種情況，你必須

5:50

有工廠必須活躍的事實，否則你有零件堆積在裡面

5:55

倉庫遍布世界各地，您無法發貨，但如果您缺少任何零件，您將無法完成

6:01

車和船，嗯，所以過去兩年一直是

6:07

供應鏈中斷的噩夢接連不斷，我們還沒有擺脫它

6:13

嗯嗯

6:19

所以這就是我們最關心的問題是我們如何

6:25

讓工廠繼續運轉，這樣我們就可以付錢給人們，而不會破產，然後其他一切都很好

6:31

有但是你知道

6:37

是的，甚至最近像中國的神戶停工一樣非常非常困難

6:43

是的，你們真的很好，我們都認為你們首先是一家科技公司，就像你們知道的亞歷克蘋果一樣

6:49

是的，他們製造硬件，但他們在軟件上花了更多的錢，我什至會說，我有時覺得這就像哦，是的事情

6:55

他們是否有這麼大的帶寬堆棧但是就像你說的那樣，通過這個地方移動的資金和硬件數量是

7:02

就像這家工廠現在正在虧損瘋狂的錢，因為我們做不到

7:07

因為我們嗯

7:12

就像我們應該呃從這家工廠生產更多的汽車而不是非常微不足道的汽車

7:17

嗯，但是我們在 4680 坡道和結構包坡道方面遇到了挑戰

7:23

然後具有諷刺意味的是工具，因為我們能夠做 2170 個單元，但是

7:30

製造 2170 變種車所需的工具被困在中國

7:35

所以有了這個上海工廠

7:40

我們擁有使這家工廠滯留在中國港口而沒有人實際操作的工具

7:47

移動它，嗯，這基本上導致了這個

7:52

工廠的生產內存只是非常小，呃現在這一切都會得到修復

7:57

你知道真的很快，但它需要很多注意力

8:03

我們會像我說的那樣，讓這家工廠實現高價值生產比建造它需要更多的努力

8:09

第一名 berlin 也是如此 berlin 的位置稍好一些，因為

8:15

柏林從 2170 風格開始，沒有 4680 和結構性後退風險

8:21

如此但柏林和奧斯汀工廠

8:28

現在是巨大的錢爐是的好吧應該像一個巨人

8:35

咆哮的聲音，這是錢著火的聲音，好吧，這就是垃圾箱著火的聲音

8:41

比垃圾箱大 太小

8:46

像柏林和奧斯汀這樣的超級工廠現在正在損失數十億美元

8:53

因為有大量的費用而且幾乎沒有任何產出

8:58

因此，讓柏林在奧斯汀發揮作用，讓上海完全回到馬鞍上是

9:05

絕大多數是我們的擔憂是的，其他一切都是

9:12

非常小的東西，基本上理解這不是一個直接的問題，而是一個格子模型

9:18

3 在你心中對你有意義 uh no

9:23

重量甚至可以像它的潛力一樣承受那種力量壓力嗎

9:30

我們正在努力降低複雜性 [笑聲]

9:39

是的，在增加複雜性方面，我們是我們自己最大的敵人

9:44

嗯，我的意思是，在大計劃中，模型 s 和 x 是我們正在做的事情，而不是

9:50

因為他們是巨額資金所有者，但真的是出於感情上的原因

9:56

就像他們在 uh full output 我們談論的是 s 和 x 每年 100 000 輛汽車

10:04

但是三個和你會做我不知道兩百萬加所以

10:10

這就像你知道的 10 倍或更坦率地說，我認為三和易可能會做

10:16

三四百萬，snx 將達到十萬，只是因為

10:21

可負擔性隨著價格呈指數下降，因此隨著價格上漲，您

10:26

基本上買得起 s 或 x 的人數不會下降一點 它下降了一個數量級

10:33

所以就像價格翻倍會使負擔能力降低 10 倍

10:38

作為一個粗略的經驗法則，是的，我們還有什麼額外的事情嗎？

10:44

可以繼續做我的意思是我們在太陽能征稅方面盡了最大努力我的意思是這可能很難，即使我喜歡

10:50

問你喜歡我們如何幫助它有點奇怪感覺就像一個奇怪的問題但是我的意思是公眾的支持當然很感激

10:57

嗯，擊落戰鬥總是很有幫助的，嗯，你知道的

11:03

你知道，有一種說法是也許我們應該做廣告，因為你知道你知道

11:10

那種傳統媒體不會對汽車進行負面報導，因為汽車就像是

11:16

最大的廣告商在他們的報紙上，所以就像廣告實際上可能什麼都不是

11:23

但它會更好，因為他們應該是壞的文章是這樣的，就像特斯拉一樣

11:29

就像基本上免費的遊戲，而呃，如果他們運行的話，可以肯定地說，呃

11:34

一些關於通用汽車的負面文章就在通用汽車廣告旁邊通用汽車營銷主管他是

11:41

會打電話給他們說，嗯，他為什麼要那樣做，你想讓我現在把它拉出來，嗯，是的

11:46

完全是的，就像嗯，是的，你的預算，嗯，明年我們不會在你的出版物上花任何錢

11:52

是的，他們不必那樣做，因為

11:58

呃他們他們知道所以有默契有一個測試和理解呃你不會在你的

12:04

主要的開胃菜 是的，這對生意不利，是的

12:11

火災有一輛卡車試圖在企業和社論之間吃掉牆我很喜歡，但他們也知道誰是

12:17

支付他們的薪水好吧他們不需要像呃他們他們不是他們不是白痴

12:22

是的

12:28

確切地說，他們只是知道不要去那裡，這就是他們所知道的，嗯，但他們可以去他們可以垃圾

12:35

像它這樣的尾巴告訴我們一個廣告，所以這就是問題，所以也許我們應該做廣告，至少就像你知道的那樣

12:42

呃，我不知道這是不是感覺不對，但是

12:47

那是我們可以做的事情，我的意思是在宏偉的計劃中

12:52

忽略一些胸骨鼓並說就像好吧好吧這實際上導致我們賣更少的汽車

12:58

呃，這似乎不是，那麼那我們為什麼要關心土地不是

13:03

問題所以他們我的意思是他們實際上可以坦率地說是一堆垃圾

13:09

特斯拉你知道喜歡喜歡特別喜歡說喜歡

13:14

從 2017 年到 2019 年，價格只是一個不間斷的仇恨流，就像

13:20

像特斯拉一樣討厭欺詐測試是失敗的，因為它是欺詐性的失敗

13:25

他們就像你知道翻閱同義詞庫就像用壞詞一樣，就像瘋了一樣

13:34

你只是喜歡像股票應用程序代碼一樣點擊電話，我記得連續三年

13:40

這就像頭條新聞，一遍又一遍地詆毀特斯拉，我們的銷量上升了，所以我

13:47

好吧，是的，所以我想我想繼續破壞我們，你知道，呃

13:54

因為在一天結束時人們就像哦，是的，我一直在讀關於特斯拉的書，我不知道為什麼，但我一直在讀它，我不記得了

14:01

這篇文章是關於什麼的，但我知道我讀了 17 次關於特斯拉的文章 [笑聲]

14:09

是好的媒體或其他什麼是的，就像這個舊價格是好的媒體有一些事實，你知道，因為喜歡

14:16

通常你會說你知道你上週讀過的關於報紙的文章有多少

14:23

內容你能記得關於那些文章的內容嗎真的只是標題它的情緒是的

14:28

標題含糊不清，我們沒有閱讀文章是的，是的，即使它們是我認為

14:33

推特上的一個大動作，比如請在轉發之前閱讀這篇文章

14:41

每個人都只是根據該死的標題轉發轉發和轉發，即使內容說的也很準確

14:46

與標題相反，編輯選擇的標題，你知道，作者是

14:51

就像我沒有選擇你知道的標題我你知道的字面意思與標題完全相反

14:56

人們甚至不知道，因為他們剛剛登上了標題是的，我的意思是實際上是最好的之一

15:02

推特上應該發生的事情就像 呃 就像每次你點擊一篇新聞文章你得到一個付費牆 就像

15:08

你這樣做是的，是的，非常煩人

15:14

所以你只能看到標題而你根本無法閱讀那篇文章[笑聲]

15:20

所以那裡有一個小標誌，上面寫著無法閱讀文章，是的，就像我不想

15:26

訂閱 philadelphiainquirer 但我確信他們在一篇好文章中包含我想閱讀但我不想要的所有內容 18

15:33

像 12 000 個訂閱，而且他們甚至無法輕鬆獲得像這樣的訂閱

15:39

就像你知道的 18 頁的過程，你喜歡告訴我們關於你整個生活的一切是訂閱 xyz 所需要的

15:46

報紙上的任何東西，比如會這樣做的人，所以想像一下

15:52

它會做，然後我不斷丟失我的密碼，然後它像推特的模擬瀏覽器一樣打開，是的，它甚至沒有

15:58

無論如何登錄 是的 就像我喜歡 就像我訂閱了《紐約時報》

16:04

華爾街日報，但我喜歡打開我點擊 twitter 鏈接，我就像 uh 它說我不是

16:09

訂閱者是的，然後在點擊後點擊轉到網絡網絡服務器讓它重定向到我手機上的應用程序，它

16:14

說 cockamamie 胡說八道，這對蘋果來說感覺像是個問題，儘管我實際上只是想說它應該在哪裡，我想知道

16:20

只是

16:27

[笑聲]

16:35

即使您訂閱了它也沒有弄清楚，我的意思是，這完全是

16:40

如果推特進入中間並通過算法開始嗯嗯取消優先級的內容，那麼就會被解僱

16:48

最終我會說他對蘋果推特所說的話最終可能會像蘋果一樣，就像你想要你訂閱的內容一樣

16:54

這樣，優秀的用戶可以訂閱 twitter 任何新聞，當你點擊它時很好，是的，twitter 是我的意思嗎

16:59

老實說，如果您確實參與其中，我可能不會付錢，我可能會每月向大西洋支付一美元，這會帶來更好的用戶體驗

17:04

我支付了每月 5 美元的 Twitter 計劃的一部分

17:10

[ __ ] 介意，是的，我要去 reddit 去尋找 [ __ ] 與 pokemon 之類的人的鏈接

17:16

呃，看看這是否曾經發生過，不，我認為那是應該發生的事情，不，這不是諷刺

17:21

就像這里肯定有一種價值損失，只是缺少

17:28

呃，我明白所有這些出版物都希望最大限度地增加他們的訂閱者，並且有點像自己的

17:33

他們的客戶和東西，但是有很多人永遠不會訂閱

17:40

出版物，例如假設您不需要特定的哲學，但假設您住在費城

17:46

所以大多數時候你不想閱讀這些文章，但他們時不時地有一些記者在一些新聞事件的現場，你和

17:53

他們實際上是源文章，其他人只是在復制它們並打印他們的東西，所以你想要

17:59

所以你想像閱讀 philadelphia enquirer 文章一樣，但你不是，你不可能成為訂閱者，所以他們不是

18:05

失去了一個訂閱者，但他們讓每個不想成為訂閱者的人感到煩惱，所以必須有某種方法來解決這種問題

18:12

四分之一，是的，這篇文章或其他東西的十美分

18:17

這篇文章只需要一個狗狗幣，嗯，你知道它應該是我會付錢的

18:22

我會為這篇文章支付 10 美分，不管 15 美分 20 美分，這就像罰款

18:28

而且，你知道，這應該奏效音樂這就是它應該如何運作音樂藝術家已經得到這樣的報酬

18:34

更多的傾聽，就像是的，擴大了很多維度，確切地說，新聞將不得不進行小額支付，就像這樣

18:40

確切地說，這裡沒有輸家，事實上媒體公司將成為贏家，因為

18:45

他們將獲得超過那些從未訂閱其出版物的人的增量收入，這是對的，然後我們都可以閱讀新聞而不是

18:51

必須被困在瘋狂房子的蒼白世界後面我的意思是我們甚至沒有一個像微信一樣好的應用程序

18:58

中國 嗯，就像在中國，你可以喜歡在微信上直播，嗯，是的，就像

19:03

是的，每個人都喜歡他們就像你住在微信上你做支付你做任何事情就像是的，它很棒基本上檢查踢屁股嗯和

19:10

我們在中國以外沒有像微信這樣的東西，這是真的，所以我想我的想法是，如果

19:15

我們只是複制微信嘿複製他們

19:21

推特複製微信是的，我想更好的用戶體驗

19:27

在你必須走之前，我們可能應該和你一起自拍或合影，好的，謝謝

19:33

感謝您觀看這些天您在做什麼

19:38

生活我在 turo 汽車共享上打破了那些汽車，我正在考慮構建一個應用程序會是什麼樣子

19:45

這使得體驗更好，因為 turo 很好，但它不是專業的，所以

19:50

不，有很多東西可以為特斯拉的體驗做得更好

19:57

突破，這只是汽車共享什麼不重要我明白了，是的，我明白了

20:02

但是關於維護負責的汽車，您可能還有其他學習的第五步

20:08

那些我認為會發生的事情是的，不，我認為這是我的意思是需要像牧羊人一樣

20:15

這就是為什麼我要在這一點上成為汽車管理員，是的，幾年前不要嘗試

20:21

不，我的意思是，在我們到達那里之前，這不是這樣的

20:26

就像你知道你必須清潔汽車 你必須知道它是否被磨損或損壞

20:32

一定是真的就像照料一群汽車是的，非常個人化，是的

20:38

呃，他們仍然會這樣，就像我的意思是我認為它會有效地喜歡那些目前像呃 uber

20:44

或者 lyft 司機，就像你自己開車而不是你自己管理我不知道 50 輛車什麼的

20:49

就像那樣，你要帶 10 到你的羊群，是的，基本上 ryan 有 52 輛網絡卡車

20:56

訂購哦，真的很好，在海灣地區完成，我們需要繼續，我的意思是

21:01

設計至少最終是鎖定的，我們肯定鎖定了是的，它被鎖定了

21:07

嗯，肯定把鉛筆鎖起來了

21:12

無論如何，我們都對具有功能的功能感到厭煩

21:18

銀翼殺手 嗯 銀翼殺手 是的 只是一直在變 是的 嗯

21:24

我的意思是去年芯片短缺非常嚴重，即使我們有，呃，那也沒關係，你可以只是你只是在搶劫一個

21:30

口袋來支付另一個口袋，這樣你就可以了，我們甚至不能做固定存儲的東西，因為

21:35

固定式蓄電池他們是否也使用了 uh chubs 是的，所以我們有

21:41

基本上就像 powerwall 線的明星一樣 uh 製造汽車，即使那樣 uh 也無法製造所有

21:48

我們想要製造的汽車，然後直接從那裡進入這個

21:53

中國封面關閉，這真的不僅僅是影響

21:58

上海生產，但也有一些零件是為加利福尼亞化妝品製造的

22:05

零件是中國製造的，而且我們的工具也卡在中國了

22:10

我想我不確定每個人都知道共同接受在那裡接受培訓有多嚴重

22:16

順便說一句，它還沒有完成，它不像主要的綠燈，但它是

22:22

仍然顛簸顛簸是的，如果您要猜測的話，很難加速

22:28

網絡卡車什麼時候能投產，我認為明年年中還可以

22:35

12 個月 好吧 我很興奮 很興奮 我們支持它 我很興奮

22:44

哦，真的，是的，我四歲就停止生孩子了，因為我

22:49

不能成長，我不能讓網絡指控超過那個，所以阻止那四個聰明人

22:57

六個席位你超過了你對社會的貢獻是的我的意思是對你有好處就像我的意思是男人這麼多朋友我

23:04

知道有像零個或一個孩子是的，這就是為什麼我喜歡我總是在敲打嬰兒鼓

23:09

我就像人類文明會讓你知道崩潰沒什麼大不了的

23:14

是的，是的，就像你認為人們來自哪裡，就像一些神奇的 [__] 人工廠

23:21

這是笨蛋鸛鸛嬰兒嬰兒商店我的意思是喜歡

23:28

是他們必須來自某個地方，他們需要很長時間才能成長，你知道的，除非呃

23:36

你知道嗎，就像傾向於發生的事情一樣，一旦出生率開始下降，它似乎會加速

23:42

像日本這樣的下降軌跡是這裡的領先指標，是的，他們即將崩潰

23:47

沒有人口 瘋了 他們去年損失了 60 萬人 那負 60 萬人

23:53

日本很接近成為最長壽的人，就像他們的預期壽命一樣，比如 85 歲。

23:59

所以這就是之前唯一更傾向於保持人口暴跌的事情

24:05

但是現在儘管擁有世界上最長的壽命，他們仍然在失去人口哦哇說些什麼

24:10

是的，我在說什麼，所以在我看來，下降的速度是

24:16

這樣的人將無法真正扭轉局面，儘管如此，我們還是有一個計劃

24:21

是的，如果那趟火車繼續下去，日本也將完全消失

24:27

世界上大多數國家，所以它不像你知道的那樣

24:33

我只是說如果這種趨勢繼續下去，希望它不會，但如果它確實如此，我不知道乍得會做什麼

24:39

就像乍得將經歷一段像下一個 10 年那樣的繁榮時期

24:44

嗯，然後，但過境率大約是更換率的一半，你可能知道

24:51

因為獨生子女政策，他們幾年前就取消了獨生子女政策，是的，然後製定了

24:57

現在是二孩政策 現在是三孩政策

25:03

他們知道有趣的事情對他們也很重要哦，日本從來沒有一次性政策

25:08

他們仍然沒有那麼糟糕，是的，所以呃，它只是看看它的趨勢

25:14

基本上就像教育和開放城市化一樣

25:22

有些人喜歡哦，人們買得起，實際上，擁有最好的社會安全網的地方有

25:27

最差的出生率

25:32

你越窮，你的孩子就越多，是的，我認為有一項社會經濟學研究表明

25:38

原因是因為父母需要安慰 ii 意味著你可以說就像一個人可以說

25:44

弄清楚是什麼原因很難說是什麼原因，但是

25:51

是的，這種相關性不能爭論，而且宗教程度越高，受教育程度越低，

25:57

越窮 增長率越高

26:05

宗教信仰低 教育高 收入高

26:11

最低出生率是的，是的，我認為如果我們解析數據，可能會有一個論點

26:16

我們如何站在宗教一邊對我們所做的人口規模產生了很大影響，因為如果你學習宗教，你就會像大多數人一樣了解

26:22

宣講富有成果的去去增長數字是的去並倍增是的

26:36

我喜歡[笑聲]

26:44

所以，是的，希望事情能好轉

26:50

我認為從生日的角度來看，目前的趨勢並不好，我試圖樹立一個好榜樣

26:57

我們在另一個星球上創造另一個文明的另一個原因是的

27:02

人類的數量是它沒有訓練好嗯

27:08

我想很多人的印像是，就像你知道的那樣，目前地球上的人類數量是不可持續的，這是

27:13

完全不真實，是的，嗯，我們可以將人類數量翻倍，沒關係，我仍然保留熱帶雨林

27:19

人們會忘記我們是動物，如果我們真的無法承受，我們就會死掉是的

27:24

就像人口密度實際上非常低，世界上大部分地區都沒有人，所以如果你喜歡飛行

27:30

你知道的國家，從洛杉磯到紐約什麼的，他低頭說像百分之多少

27:36

如果你丟了一個保齡球，你會在這裡打人嗎？我們知道我能看到你像圓形嗎

27:41

第二個哦，是的，是的，但就像你只能知道幾個城市

27:46

和

28:00

打人或死你會死

28:05

因為你在任何問題上都不會成功，即使你認為 la 有很多人，但你看看實際情況

28:11

你喜歡 只是看看說什麼橫截面積是人類喜歡 喜歡的橫截面

28:19

你知道你在洛杉磯和幾個街區裡把房子和汽車弄成半透明的

28:25

並且從上面看這個區域的橫截面中有多少百分比是人類

28:31

哇，幾乎沒有低於百分之一的東西哇，有趣的是，你見過嗎

28:37

他們做的那個巨大的肉球如果你把每個人都結合到紐約市你就會看到它就像所有人類一樣的大粘液球

28:43

被推入一件事它想坐在紐約的呃紐約中央的大公園是什麼它會坐在中央

28:49

就是這樣，就像它比最高的摩天大樓高一點，一點也不像人類的長方體

28:57

[笑聲]

29:16

在一個樓層上，您可以容納紐約市的所有人，呃，在一個樓層上，甚至沒有任何東西

29:23

你的意思是在一層嗎？我的意思是哦，表面積是的，人類餡餅會有多厚，不，它只是一個人

29:30

高我要說沒有人沒有堆疊沒有堆疊它只是並排你的意思是是的只是說像什麼是橫截面誤差

29:36

一個人是的，說好吧，有八十億，紐約市的面積是多少，哦，它小於

29:42

你可以在紐約市找到斯坦福所有的人

29:47

是的，來自太空的東西，只是我們的目標，橫截面很低

29:53

所以我們將不得不只是粗略的導彈，因為人類的低橫截面

29:59

這基本上就是歸結為嗯所以是的無論如何是的就像如果人們

30:05

就像生活在一些繁華的城市他們認為人類必須無處不在但實際上呃

30:10

很少有城市像美國沒有任何地方像曼哈頓的密度那樣死了，即使你有

30:16

就像它看起來像一個密集的城市，人類的實際橫截面積很小，不到人口的百分之一

30:23

區域是人的，所以這意味著如果你丟下一個不會反彈太多的食人族排斥

30:30

擊中人類的概率極低哇，有趣的是，我們已經從火箭和多個國家重新進入了碎片

30:36

當我們有這個棒球時，坦克並不受控制地墜毀並且沒有多次擊中任何人

30:42

像這樣的悲劇在進入它時它在整個美國下雨了碎片

30:49

而且我認為根本沒有任何傷害是的就像整個是的很多碎片

30:56

我哭了，沒有人害怕哦，我的上帝，它幾乎擊中了我，是的，好吧，他們不是保齡球

31:04

就像他們從外殼中恢復了大部分材料一樣，哇，但最後零人被擊中我相信

31:12

這是一個有趣的部分，它不是垂直的，而是圍繞地球旋轉的，這增加了概率，所以它就是它就是它

31:18

從軌道進來，所以它會或多或少平行於表面，然後開始分解

31:25

呃，我相信我們就像農作物的灰塵一樣，或者至少肯定就像沒有任何一部分掉入大海一樣

31:30

一切都落在了這片土地上

31:38

當你進入軌道時，你會平行於地球表面，你知道大約 25 倍的聲速

31:43

所以細胞減速和分解呃酸和碎片如雨點般落下

31:50

我們和沒有人在這裡所以

31:55

地球上實際上很少有人，嗯，我有一個與 fsd 相關的問題

32:02

有可能讓我們接觸到一些人，從我們的經歷中給出敘述性反饋，因為我們中的一些人駕駛 [ __ ]

32:08

噸英里穿越德克薩斯州某處和回來的地方，有些事情我們看到沒有

32:14

我可能會發送按鈕或電子郵件說嘿，這就是我們看到的問題，這裡是它們如何崩潰，更重要的是，這裡是如何

32:20

他們正在跨上下文分解，就像如果有任何方法可以幫助如果不是那麼沒有

32:25

擔心，但如果有辦法向某人提供敘述性反饋，我認為這可能會有所幫助

32:32

我的意思是我們現在的目標是在城市的點驅動器中實​​現零干預是的

32:39

所以拋光不如解決零干預重要

32:46

第一點在城市裡開車 嗯，需要做的工作量很大

32:51

這讓人難以置信

32:57

甚至試圖拉他們漂亮的二頭肌我們不認識任何人也許有人但是

33:02

我的意思是我看到了很多艱難的技術

33:09

問題和解決現實世界的人工智能，使汽車可以自行駕駛是最難的問題之一

33:14

我見過的問題，而且比我最初想像的要困難得多

33:19

嗯，你會對那些說過的人說什麼

33:25

幾年來你一直說今年會是明年你會對那些人說什麼今年會是

33:31

我們得到了它的報價六個月

33:37

是的，不，我認為它今年仍在跟踪，但就我而言，我會更喜歡呃，是的，為什麼這樣做

33:44

我過去在哪裡過於樂觀了呃我沒有我只是了解問題的範圍和你

33:50

感覺你現在是的，嗯，我的意思是

33:57

就是這樣，很容易接近它的工作方式和感覺的地方

34:04

只是你知道的，你知道的，你知道的，嗯

34:10

就像你知道驢和胡蘿蔔在那邊 就像是對的 這些讓你知道

34:16

把卡拉得更快一點

34:22

所以，坦率地說，嗯，有雷達和超聲波，呃，

34:30

錯誤的雷達，特別是因為雷達呃它可以讓你接近解決它

34:37

並在大多數情況下解決它，除非您無法鎖定雷達和

34:44

像雷達雷達和視覺這樣的視覺神經網絡不同意哪一個是正確的

34:49

嗯，所以你基本上必須擺脫雷達嗯

34:55

一旦我們讀過雷達，順便說一句，它最初被自動駕駛團隊強烈反對，現在沒有人想要

35:01

雷達回來是對的，是的，我必須放下低調，說雷達要出來了，伙計

35:08

是的，呃，你不能說雷達是一根拐杖，如果你背著那個迷戀，你就不能跑

35:14

呃它只是它雷達會它只是信號

35:21

在一天結束的時候，雷達產生的噪音比信號還多，所以這並不是很大的矛盾

35:27

信號，但它是比信號更多的噪音，所以那就是呃

35:33

它必須被移除，嗯，一旦雷達被移除，那麼很明顯，呃實際上我們的神經網絡要糟糕得多

35:39

比我們想像的要多

35:46

嗯，所以現在是的，只需要使用視覺

35:52

呃你你你沒有什麼是你無法隱藏的

35:57

工作神經網絡視覺必須工作，然後有一個相當深刻的

36:03

改變從點神經網絡包到

36:09

讓神經網絡解釋那組點以找出車道中心是什麼，因為以前我們都在做這一切

36:15

在c中，所以你知道

36:21

我的意思是神經網絡架構非常複雜

36:26

在這一點上，它就像是很多層，然後

36:32

然後你然後它會變成哦實際上我們需要刪除這個層把這個層放在上面，就像我們已經做出反應一樣

36:38

多次重新構建神經網絡這是我從外部提出的一個問題，就像特斯拉一樣

36:44

有像一個通用的自動駕駛團隊為什麼不讓兩個與吉爾競爭

36:49

與 boca chica 和 florida 合作，為什麼不擁有一個你知道還有雷達的

36:54

和所有舊數據點和一個單獨的團隊，讓他們競爭

37:00

我認為團隊並不缺乏動力水平或職業道德，他們工作非常努力並且

37:07

特斯拉自動駕駛儀呃那種軟件

37:13

人工智能團隊就像我合作過的最好的軟件團隊 嗯，我是技術技能的主持人

37:19

他們非常出色，我們花了一段時間才到達那裡，我們有很多錯誤的開始

37:25

自動駕駛團隊 嗯，我就像在矽谷 有 12 個人說我不知道

37:31

他們負責自動駕駛，但他們不是

37:40

那支球隊中也有一些通常很好的人

37:47

他們通常沒有很多時間

38:02

是的，前幾天街道的設計目的

38:08

呃眼睛和生物神經網絡，這就是整個道路系統的設計目的

38:14

因此，坦率地說，相機和數字神經網絡是完全合理的

38:20

需要解決自動駕駛你基本上必須

38:26

重複人類所做的，但在矽中，沒有其他東西可以真正解決

38:32

所以我的意思是特斯拉實際上我的意思是特斯拉有自動駕駛解決方案而沒有

38:39

其他一個甚至接近五年許可，是的，是的，絕對的

38:46

我們不會阻止人們這樣做，但他們會這樣做

38:52

它需要像我們想像的那樣打他們的臉

38:58

就像它必須從他們那里奪走銷售額哦，它會從優步那里奪走銷售額它會奪走銷售額，是的，一切

39:05

所以很明顯它實際上已經交付了很明顯這

39:11

一個月內每個人都會在那個時刻知道的方式 如果他們還不知道，這就是方式

39:17

知道正確的人誰擁有那些你們擁有測試版你們已經看到了改進的軌跡和

39:23

這很明顯會實現全自動駕駛是的，這只是一個有爭議的問題

39:28

什麼時候是的，但是但是如果你說每單位時間的改進率是多少，是的，你可以說這顯然會達到一個點

39:34

它比人類安全得多是的，嗯，你知道的那些人基本上在那裡

39:40

是不是你越過了那一點，感覺好像我們在減緩進展，我想知道它

39:45

可以哦，好吧，這樣很難從外部欣賞的東西

39:52

公司是我們將像第一千次那樣重新構建神經網絡

39:58

和嗯嗯

40:03

我的意思是這就像我喜歡的東西他們佔據了最多的東西

40:09

我的大腦空間正在研究自動駕駛並讓星際飛船進入軌道

40:14

兩件事就像我不知道我大腦的 70%

40:19

空間之類的東西 嗯，還有其他的東西

40:25

只是不需要那麼多的 cp 腦 cpu 週期，但基本上像很多家務

40:32

如果我做我的家務然後[__]得到粉絲基本上我討厭狗家務嗯

40:39

所以嗯

40:49

無論如何，就像基本上自動駕駛儀正在吸收我的大量大腦週期，並且得到

40:55

星艦環繞嗯或兩件壓倒性的東西

41:01

這是絕大多數的大腦循環，嗯

41:06

所以牠喜歡說 12 那是一個或其他看起來不像的原因

41:12

是因為我們重構了太多的神經網絡

41:19

看起來像是向前兩步，向後退了兩步，嗯，所以感覺就像進步是平的，但實際上是一個整體

41:25

[ __ ] 很喜歡我們剛剛對神經網絡做出了反應

41:32

就像 a 和 ac 架構一樣，擁有 ac 和 a 架構，但現在

41:37

現在你可以提高成長空間是的，你明白了，所以你知道

41:44

他們仍然我們仍然沒有統一的向量空間呃所有的神經網絡

41:50

將他們的輸出倒入一個商定的 uh 向量空間中

41:56

嗯，固定的和移動的物體，所以很大，是的，相當的

42:02

很重要，因為就像他們不同意一樣，如果移動物體

42:08

像目前這樣的神經網絡，它們真的將移動物體網絡和靜態物體網絡完全分開

42:16

呃，如果他們不同意汽車的位置，那麼就像你可以說這輛車在非駕駛空間

42:22

位置是因為神經網絡之間存在分歧，然後如果它們沒有以相同的幀速率運行，您可以獲得時間

42:29

錯誤也是如此，如果是這樣，如果您知道汽車在移動

42:35

我不知道呃 每秒 30 米之類的然後你呢

42:42

你知道有像那輛車一樣的 33 34 毫秒錯誤

42:47

現在移動了很多你知道所以它移動了一米你知道所以

42:53

嗯，一米很像，如果我們不與某物相撞，那是一噸，是的，它可能像，所以這就是這輛車

43:00

就像你在開車讓我們說這兩個有兩個如果你這輛車朝相反的方向行駛

43:06

如果你有如果那輛朝你駛來的汽車被一米偏移，這可能是碰撞與否之間的差異

43:11

所以嗯，所以你知道一個之間甚至可能會有一個輕微的時間滯後錯誤

43:17

明年你可能會遇到問題，所以沒有任何限制，因為這些神經網絡變得非常

43:23

很好，他們會比人類好得多，所以它可以開車

43:29

比詹姆斯邦德更好可能會不幸地推動呃的門檻

43:36

成功只是對人類不利，而人類總體上確實有很多延遲

43:41

所以當你意識到它的時候，這就像不是一個酒吧那麼高

43:46

有點像你以前認為的電梯操作員嗯嗯嗯嗯，是的，你可以成為一名專業的電梯操作員，但你必須翻轉那個繼電器

43:53

然後讓電梯和地板完全對齊就像呃

43:59

你不會像計算機一樣成功，所以現在任何人，如果你確實有一個

44:04

帶大繼電器開關的電梯操作員，你會認為這很落後

44:09

而且你真的更喜歡有一個按鈕

44:14

自動駕駛儀是一個很好的類比

44:22

並且打到um後，waterpilot的精度會瘋掉

44:28

沒有對城市街道進行干預 此後的下一個基準是什麼，我假設它正在運行

44:34

一些英里數 一些驅動器發生，因為如果我們說真正的零，那麼你已經達到了自主權，就像真的一樣，絕對是

44:41

會是非常罕見的，呃，有乾預我的意思是開車

44:47

呃今天從我朋友家到你的工廠我沒有乾預嗯實際上就是這樣

44:53

奧斯汀，你知道在城市街道還是在高速公路上，嗯，所以我說的唯一一件事就是

45:00

汽車決定的速度是錯誤的，所以我不得不多次調整速度，因為我

45:06

從自動駕駛團隊的反饋中理解，這被認為是一種干預，對你有沒有踩踏板油門不，我不是

45:11

談論加速器，好吧，很酷，這輛車基本上開得太慢了

45:17

就像它一樣，這里肯定會與現實發生碰撞，因為你知道就像

45:23

加利福尼亞州的滾動停止 每個人都會停止滾動，就像你不認識的人一樣

45:28

是的，如果你不這樣做，你實際上會被追尾是的，是的，就像最後沒有人一樣

45:34

從字面上看，您知道十字路口是什麼，然後停下來，這有點奇怪

45:39

嗯，你知道，但從技術上講，這是違法的，所以你知道美國航空航天局對此全力以赴並說哦

45:45

你必須來到一個完整的零船體，然後喜歡

45:52

呃，因為一些關於速度的規則，就像你被允許的那樣，你可以

45:57

永遠不要顯示低於實際速度的速度，所以你必須贏得它的一面

46:03

高於它的真實情況，所以有時它會顯示每小時一英里，即使是短暫的，即使車輪

46:08

被停止 [音樂] 因為 nitsa 規則那是速度

46:14

然後有人提出投訴，所以我們必須從字面上展示一張照片

46:19

車輪顯示車輪的視頻，看它說每小時一英里，但那是

46:24

因為你的規則讓我們每小時跑一英里，但輪子不動

46:30

從字面上看，它會出現在視頻中，以證明車輪不是車輪，即使它不動

46:36

說一英里每小時，因為他們的規則說一英里每小時所以哇哦

46:43

我們收到了很多來自競爭對手的關於自動駕駛儀的投訴

46:49

因為他們沒有答案[音樂]

46:54

他們對從特斯拉獲得許可的想法並不感到興奮，這不像哦，是的，讓我們給我們的

47:00

競爭對手一大筆錢是的，這不是他們的首要目標，因為價值太荒謬了

47:07

全自動駕駛汽車的價值就像我們以前從未見過的一樣

47:12

為什麼你認為每個人都低估了呃自動駕駛似乎世界並沒有真正

47:18

理解它的價值為什麼它來自我想如果有人

47:26

從未見過獨角獸或你知道的東西它們就像宇宙不存在你知道它是怎麼發生的

47:31

只是一匹頭上有角的馬沒什麼大不了它是一個神奇的生物你

47:38

知道這對大多數人來說基本上自動駕駛聽起來像是一種神奇的虛構事物

47:46

直到你真正使用它然後關閉你的想法是的所以現在使用它不會那麼難

47:51

他們可能只是想和現在在車裡參加測試版程序的人交談，是的，認真的

47:57

弄清楚它並不難，但他們沒有嘗試是的，嗯，所以就像他們都只是

48:03

認為沒有呃自動駕駛就像是很遙遠的斜線他們已經投資了一些

48:09

解決方案，我們不希望任何人追求純粹的願景

48:14

解決這個問題的方法，也需要大量的標籤 mm-hmm

48:21

就像定制的自動標籤軟件，你知道大量的培訓

48:27

計算機 它不僅僅是車載軟件 它是規則寫給我們的所有培訓、標籤和調試軟件

48:34

寫得太像它不存在一樣得到了它的絕對數量的軟件

48:39

告訴我們 uh 只是用軟件工具和調試是 uh

48:45

瘋狂的嗯，你知道就像拍攝視頻幀一樣

48:51

分析此視頻並看到啊，這就是它出錯的地方，這就是為什麼當訂單標籤軟件

48:56

需要更正，這就像訂單標籤本身就是一件大事

49:01

巨大的，是的，我們有 1500 名人類貼標籤者哇哇是的

49:08

但它們可能被放大了一千倍

49:13

標記好至少 100 倍。哇，所以它更像是 150 000 完全正確

49:20

呃，好像很多

49:26

這是巨大的，是巨大的，嗯，光環標籤的速度意味著

49:32

比如你可以訂購標籤的速度，然後當你訂購標籤整個視頻時

49:37

段，嗯，你知道這就像成千上萬的幀被標記為順序，然後

49:45

人類所要做的就是說哦，是的，那是正確的，或者不，嗯，稍微調整一下這條線

49:52

就是這樣，然後它會更正訂單標籤，然後您將其反饋到訂單中

49:58

貼標籤，嗯，基本上你訓練訂單貼標機來訂購標籤，然後你就更少了

50:04

隨著時間的推移，錯誤越來越少 是的 相信人類是正確的 不，我們不

50:09

有很多有很多訓練的例子，你只是相信那不是統計上的錯誤

50:16

是的，這樣的錯誤很少發生

50:22

我們瀏覽訂單標籤軟件並說這是為什麼

50:28

為什麼這件事發生的概率並不比我們認為它應該的高很多

50:34

有很多情況是因為一個非常嚴重的錯誤標籤錯誤，所以

50:40

神經網絡會認為，絕大多數時間線都在那邊，但時不時地又在這邊

50:46

就像沒有一樣，它仍然在這裡，這只是一個標籤錯誤，所以它實際上試圖彌補

50:52

因為呃，你知道這是一個統計數據，所以就像哦，那裡有一個它

50:57

有一個石油分配哦這裡有一個小駝峰不這裡沒有一個小樞紐只是貼錯標籤刪除

51:05

現在一切正常，現在可以正常工作，所以其中一些最像一個重要的

51:11

錯誤標記錯誤，嗯，它可能會顯著降低神經網絡的有效性，是的

51:17

肯定有一件事，嗯，所以我走過去，我在目標另一件事，昨天我穿過停車場，這個舊的

51:23

女士在倒車，當然根本沒有照顧我，她差點撞到我，嗯，我要說的是

51:30

很多將從全自動駕駛中拯救出來的人不會知道它是否是我的特斯拉

51:37

躲閃 好吧 好吧 我們是百思買 我們今天要百思買 已經通過其他人了

51:42

是的，他們是你知道一輛車是的，所以這很瘋狂是的，是的

51:51

我的意思是，當我們開始下載自動駕駛儀路徑時，我實際上被教導有人說，哦，是的，你知道的

51:56

就像有人說的好，你不會成為你的90個人，你不會得到任何讚美

52:02

保存，你會被你沒有的10個人起訴所以這是一個吃力不討好的任務是的

52:08

所以就像呃，但前幾天我想我們只需要把這個提交陪審團說聽著你必須相信這裡的統計數據

52:15

是的，因為如果這就像你知道的那樣，這就像結束一樣，我認為高速公路安全比安全一個數量級

52:21

人，就好像這些不是小數字在這裡你知道你只是客觀地說有多少崩潰是

52:27

哦，它少了十倍，好吧，就像那家公司一樣

52:32

大大減少了 uh 死亡和反事故不應該因為少數確實發生的事情而受到懲罰

52:39

就像那樣會超級混亂殺死完全自主安全會發生

52:44

即使有完全的自主權，仍然會有非零數量的東西，因為你很像

52:49

跌得很好，就像您仍在與呃互動一樣，您知道這將需要一個

52:56

在艦隊自主之前很長一段時間就像艦隊自主時讓我們說當時的問題是

53:01

事故將非常罕見，比現在的情況少或多 100 倍，就像你必須要做的那樣

53:07

從字面上跳入交通沒有汽車的機會，就像你知道的自殺一樣

53:15

所以呃當所有的汽車都是自動的但是很長一段時間呃會有自動的

53:21

汽車與 99% 左右的非自動駕駛汽車混在一起

53:26

然後呃這意味著人們會知道我們會像這樣在交通中轉彎

53:32

你認識的人有高速不分岔路和卡車

53:38

只是開車進入你是的，好吧，就像你應該做的那樣

53:44

嗯，是的，嗯，嗯，但我的意思是我認為人們也

53:50

意識到有多少人正在那裡死去，你知道的

53:55

基本上司機睡著了，或者我的意思是傑伊的妻子被卡車司機殺死了

54:00

睡著了，他忘記了她騎著自行車，是的，是的，實際上是

54:06

卡車司機睡著了 是的 卡車掉頭過馬路 撞死了

54:11

是的，如果那輛卡車有引水員，呃，它會遲到，它應該不會活著

54:18

今天哇，具有諷刺意味的是，無論如何，無論哪種方式，我們都會去

54:23

繼續這樣做，嗯，這是你知道的，是的，是的，必須這樣做

54:29

讓我們去取暖我不知道最後一件事它不會花那麼多

54:36

阿爾法波或心靈分享，但很多人很多粉絲和所有者想要

54:41

知道蘋果音樂，你一直在迴避我的問題，所以我不再問了

54:46

[笑聲] 我們在這裡尋找什麼

55:05

是的，它肯定會更好哦，拜託，真的謝謝你

55:14

我的意思是你應該喜歡與你的日曆同步，你不應該輸入任何東西是的，我必須使用雲配置文件

55:21

我的日曆可以正常工作是的假設我們的日曆同步工作得很好我不知道我只是

55:26

我不只是不我只是不使用日曆那是一個笑話是的對我來說這是一個命中和錯過但是

55:32

是的，有一段時間我們讓它運行良好我想我有一段時間沒有使用它了，但它基本上應該只是

55:38

automatically take you where you want to go without you ever say anything yeah and just look at your calendar and yeah yeah go based on that yeah

55:44

essentially so spotify title amazing thank you oh my god holding the new

55:50

model s when you're actually connected to wi-fi yeah yeah right before the cyber rodeo we were in a parking garage with wi-fi oh my god yeah really

55:56

unlocked yeah title is great i love it but balafan's don't have title but they do

56:02

have apple music so they've just been asking so yeah really is it like are you talking about like the

56:08

apple's like car implementation no no no no no no you're saying literally so spotify has a subscription that you can

56:14

付費喜歡訂閱它是的，蘋果有一個蘋果音樂，用來諮詢哦，你的意思是不是通過電話

56:20

基本上是的，對汽車是正確的，是的，是的，是的，所以你想要這樣只是為了避免藍牙是的，是的

56:27

因為那時它仍然是的，是的，有相當多的客戶

56:33

使用 Apple Music 好，而不是 Spotify 就是這樣，是的

56:38

我用蘋果我實際上是我用蘋果音樂和Spotify um的少數人之一

56:45

但我只是通過藍牙使用 spotify 我的意思是我應該通過藍牙選擇蘋果音樂哦，你做的

56:50

所以如果我播放蘋果音樂我只是藍牙它哦嗯是的

56:55

它不會打擾你音質 呃我的意思是音質

57:04

我好像我們絕對可以提高藍牙數據速率，如果那是

57:09

我們正在談論的瓶頸我的意思是這會讓每個人都開心，但不僅僅是蘋果音樂建設者肯定是的

57:16

但我使用Spotify，但我仍然通過手機進行操作，因為這就是我在家裡使用的方式，當我使用我的sonos時，我通過手機進行操作

57:22

所以我上了車，它通過了

57:34

不，我很抱歉，我只是還處於關閉狀態，就像整個經歷一樣，是的，是的

57:40

對不起，你抓住了我那些空白的時刻

57:50

我們已經有了航路點，所以我沒有請求汽車的功能

57:59

有一天，但我沒有一個，所以好吧，我現在是一個農民

58:05

legacy s or have a legacy p 100 d so the blind spot camera when you right

58:12

click or when you use the blinker it doesn't show up so when you reverse though

58:19

you know the repeater cameras they would show up they show up in you know all the new ones but i'm on the legacy x

58:25

it's 2018 december show's up yes so it's already there it's just it

58:31

you talk about the future on the three when you hit the blinker on it'll show yes i love that i got a rental and i got

58:36

me i mean elon you responded to me to do it and then it just didn't get on my car

58:41

so it's like okay it seems like a small number of people but that's a

58:46

feature too that's why i said i'm a peasant but i'm not gonna it's fine i mean i'm not saying like our resource

58:52

allocation is is perfect because it certainly isn't but uh you i mean one does need to stack rank

58:57

these the issues by like how much good will this do times number of people yeah

59:02

um and and we we have a lot of uh long list of things uh

59:08

so so the and pretty far down the list is is like uh you're adding features to

59:13

uh legacy yeah products that's why i said i'm a peasant a different way of saying that is upgrade

59:19

oh come on i got free lifetime supercharger man it's yeah it's just it's just like it doesn't it doesn't make sense for you know

59:26

未來的人是有道理的，就像如果有一些

59:31

剛剛生產出像你這樣優秀的軟件工程師的工廠，你剛剛訂購的，我需要一百個

59:36

級別的軟件工程師他們是沒有問題的，但並不是只有少數優秀的軟件

59:42

工程師是有道理的，而我認識的一些人很棒，你發了財，停止了工作

59:49

是的，你可能知道其中一些是的，你正在努力降低價格，不停地離開

59:54

是的，我的意思是我認識很多非常有能力的人，一旦他們得到

59:59

有錢他們剛剛停止工作肯定就像我不知道這是他們的生活但是

1:00:05

他們停止編碼，他們只是呃我不知道他們實際上在做什麼

1:00:12

就像看他們說他們投資和我的東西，但你知道

1:00:18

是的，我的意思是有很多有很多真正有能力的人是的

1:00:23

類似的退休，我知道，嗯

1:00:29

我就像一個怪人，我仍然像個瘋子一樣工作，因為我就是這樣連接的

1:00:35

但是我認識的大多數人他們只是把一定數量的錢在那裡他們的工作量他們在這裡的工作量

1:00:42

不太好這是這次採訪的有趣之處我從未真正聽到過你說你基本上是一個存在主義者

1:00:47

i've not heard you say that directly i've always thought that actually basically studying in high school yeah right around the same time i discovered

1:00:53

tesla and i was inspired tied to that and so it's just interesting that like yeah i don't see you ever really retiring just because of

1:00:59

the fire that is driving you to work right now and yeah yeah i just don't see it happening

1:01:04

yeah exactly so it's like trying to you know

1:01:10

i'd just try to figure out what's going on

1:01:16

um yeah yeah and then and then just like take take the set actions that are most

1:01:22

likely to enable uh us to figure out what's going on

1:01:28

and understand the universe um yeah

1:01:34

i think if you're not an alien i just realized that you're trying to become an alien you're trying to pursue and find other life potentially you're coming yes

1:01:40

we're trying to be the alien yeah yeah are you talking about the entire time i've never thought of it that way

1:01:46

yeah at some point yeah yeah astronauts are kind of just aliens yeah

1:01:53

i mean it's very it's very you know the furry paradox is a profoundly important question

1:01:59

um and there's a book that's like i don't know 50 answers to the fermi paradox or

1:02:05

um something now and again they put out a new one when they can think of some other answers to the funny paradox it's quite

1:02:11

good um because it's very troubling that we do not see scientific science of aliens

1:02:18

very very troubling like something doesn't add up the universe why is the universe 13.8 billion years old and we

1:02:23

see no evidence of of any alien visitation whatsoever

1:02:28

nor can we detect any signals or anything like that yeah nothing the only thing i can think is it's time bound

1:02:35

we're the last one that's currently alive and there may be future ones that become alive or at least can be

1:02:40

perceived as alive in the future yeah well actually like the universe is like say 13.8 you know go like basically thousand

1:02:47

eight hundred and something million years old that's when you talk about that well if you even wanna increment the

1:02:52

就像小數點後的第三位數字將是未來一百萬年以及我們的文明以及自

1:02:59

寫作的起點是五千年，所以

1:03:04

考慮到目前的出生率，是的

1:03:09

我想是的，你認為我們被允許

1:03:15

一百萬年是文明長度的200倍 是的 那是什麼 是的 你說你

1:03:21

認為我們是孤獨的我當然是在說你知道我認為我們應該堅持

1:03:27

科學方法是我們沒有證據的證據是的

1:03:32

不是的，只是沒有任何證據表明我們沒有任何外星人生命的證據

1:03:39

nothing nothing nothing flat damn nothing yeah no signals no you know like

1:03:44

all i would have taken would be for like an alien civilization to to put like

1:03:50

like a one inch cubed of pure titanium

1:03:57

like just like just a titanium cube i'll be like [ __ ] aliens man okay because they sure as [ __ ] they sure as

1:04:03

hell didn't know how to how to how to refine deoxidize and and create like a cube of pure titanium oh what is that oh

1:04:09

pure anything pure nickel how about that so in my mind alien life it will if we if and when we find the especially the

1:04:14

first one we'll be more like a bacteria we probably won't find uh you know human-like able to modify the world

1:04:20

around them i think we'll find bacteria yeah so i think this this likely like if

1:04:26

you look at earth earth was was just basically archaic bacteria for a very long time

1:04:32

um and and it was a huge thing to have mitochondria get incorporated into a cell and then have

1:04:39

that mitochondria also divide when the cell divides that's so gigantic

1:04:44

um and you just can't have a complex organism really of any size unless

1:04:50

you've got a little power plant with the mitochondria in your cell yeah so but with the myocardium basically seems

1:04:57

to have been like uh some bacteria archaic bacteria that got captured and then essentially

1:05:03

when it's symbiosis with a thing but but then the crazy thing is like how does it replicate itself when the when the cell

1:05:10

divides the mitochondria must divide at the same time how does the mitochondria absorb

1:05:16

that so so there must have been like a crazy number of times where mitochondria was absorbed but did not

1:05:22

do not replicate itself when the cell did i think that's why i think we'll find bacteria but i guess the follow-up question is how would we even perceive

1:05:30

that if it's further than our solar system how do we perceive bacteria as life if it was outside our

1:05:36

solar system because that's what gives me hope is that it's just so small or different

1:05:42

我們只是還沒有能夠很好地向前看，我認為一般來說我們是

1:05:48

尋找喜歡的生活，我們可以和你交談知道類型的事情，因為它是生活還是生活都沒有關係，它可以

1:05:54

和你談談是的

1:06:00

醒著很酷很酷不能和他們說話簡單的貝殼生物

1:06:08

它們可能不同，也可能與不同的簡單貝殼生物不同，好吧，我們找到了

1:06:13

各種不同的符號貝殼生物嗯，你知道

1:06:20

坦率地說，我認為這似乎很可能是的，嗯，我們去是的，不，我的意思是我說概率

1:06:26

from a probability standpoint i think that's likely yeah um it's like how many civilizations that

1:06:32

can actually you can actually talk to that are around at the same time as we are around

1:06:38

those appears to be a vanishingly small number maybe zero that are in this galaxy

1:06:44

and we don't really know of a way to get to other galaxies with current physics would you say given potentially all the

1:06:51

unknowns what is your thoughts on there being like a creator

1:06:59

you know in general i'm just saying like what like if you're gonna uh aim for an evidence-based uh approach

1:07:05

to existence then you simply say what is the evidence and you your conclusion should be

1:07:11

probabilistic according to the evidence um

1:07:16

so there's clearly some like like we don't know what the origins of the universe

1:07:21

are um we don't know if this is just is this a simulation always somebody's video game

1:07:27

i mean if you look at how at the progress of video games going from simple blocks with pong to

1:07:33

you know 3d re you know massively multiplayer photorealistic video games

1:07:39

in our lifetime if that trend continues video games will be instinctual from reality

1:07:46

so not saying our video games we're saying that this is a remarkable situation we witness

1:07:55

you can still be at a single turn of civilization yeah i mean that's the wild thing right

1:08:01

有沒有它找到某種事物導致宇宙的方式的歷史

1:08:06

你知道有一些沒有原因的原因或類似的東西

1:08:12

這有些像呃起源它就像它來自某個地方它是一個完整的幾個問題就像是

1:08:17

有沒有什麼看不見的生物或看不見的神 呃 判斷我們的

1:08:23

行動並決定我們死後是去好地方還是壞地方，但似乎並非如此

1:08:33

或者至少不是它不清楚規則的用途

1:08:38

因為那樣就會有很多像你知道的大屠殺者和本來應該是的東西

1:08:43

我不知道在一個好地方還不夠所以我不知道這很難說也許

1:08:49

如果人們對宗教變得非常防禦，因為就像你知道他們會認同

1:08:54

如果你自己開車，我會放在那裡的一個特定的宗教

1:09:00

存在主義就像為他們取代了它是的，它正在取代他們的目的

1:09:06

很好，這肯定是，嗯，這將是一個改變

1:09:13

嗯說像讓我們追求呃意識的擴展

1:09:21

呃，為了理解現實的本質，如果你說不，我願意

1:09:27

了解現實的本質，因為我相信這個宗教或那個宗教和那個

1:09:32

這就是我對現實的解釋，這就是很多人會如何看待一些事情是的

1:09:38

嗯，所以我很好，你知道我的意思是

1:09:43

呃，你不是不想進入宗教辯論，只是它只是

1:09:49

如果人們知道信仰宗教，那將是一條不同的道路

1:09:55

從表面上看與呃採取我們不知道發生了什麼並且我們想知道的立場

1:10:01

發生了什麼事，是的，所以這是非常不同的，不同的，呃，一組動作

1:10:07

我猜我說的就像宗教讓人們相信他們已經知道發生了什麼

1:10:12

是的，嗯，你知道我對宗教沒有任何問題，呃，前提是

1:10:19

我們通常只是朝著促進意識的方向前進

1:10:27

歸根結底，如果一種宗教導致更多的人被創造出來，然後

1:10:33

無神論者沒有孩子，那麼這將是一個自我實現的預言

1:10:39

是的，命運喜歡諷刺，是的，周圍唯一的人將是宗教人士

1:10:48

好的 好的 確定

英語（自動生成）

全部可聽

 

 

於
標籤：

沒有留言:

張貼留言

訂閱： 張貼留言 (Atom)