00:00:00
[Applause]
00:00:09
okay wait I'll just I'll just read off
00:00:12
all of your companies Elon I know them
00:00:14
but I'm just going to read them to make
00:00:15
sure I don't miss one because there's so
00:00:16
many now uh founder CEO chief engineer
00:00:19
of SpaceX uh CEO product architect and
00:00:22
chairman of Tesla uh owner chairman CTO
00:00:26
of uh x uh x.com uh founder of boring
00:00:30
company co-founder of neurolink and open
00:00:32
Ai and president of the mus Foundation
00:00:35
did I get
00:00:36
everything let your Winner's
00:00:43
ride and we open source it to the fans
00:00:47
and they've just gone crazy
00:00:48
[Music]
00:00:52
with where are you me it is
00:00:56
coming it's kind of absurd where are you
00:00:58
at Star base
00:01:00
um I'm in FL currently so this is this
00:01:04
is a starlink and Flake connection are
00:01:07
you kidding me that that's oh yeah that
00:01:08
works pretty well
00:01:10
huh I think there's only
00:01:13
one wait I think it's one of of how
00:01:16
stalling Works uh in an
00:01:18
airplane at altitude there's only one of
00:01:22
those in existence right it's on your
00:01:23
plane that's it 101 there are a number
00:01:26
of airliners that have stalling and
00:01:28
there will be a lot more in the future
00:01:31
Starling connection when you're assuming
00:01:33
it's working properly is you would even
00:01:35
be able to tell you you're on the ground
00:01:36
or in the air um because unlike Jo
00:01:40
satellit the latency is you know really
00:01:44
be less than 20 millisecs so
00:01:48
um it's a in fact for a lot I think for
00:01:52
some people the Sonic connection on the
00:01:53
plane will be better than their
00:01:54
connection at their house that would be
00:01:57
pretty great um H how is um
00:02:00
the Starship doing it was incredible to
00:02:02
see the first uh launch but I understand
00:02:05
you're closing in on the second I know
00:02:06
you've been working really hard on that
00:02:07
and the team's working hard on it um
00:02:10
when do you think you're going to get
00:02:11
the next one up and um what are the
00:02:13
chances it makes it to orbit uh well we
00:02:16
have the second one stacked at Star base
00:02:19
so it's ready to go um and uh we
00:02:22
finished that up in the last
00:02:24
week we believe we've we've completed
00:02:26
the rain um I had an requested by the
00:02:29
FAA so we should get our license
00:02:31
hopefully soon um but really the only
00:02:34
thing holding back second plan of
00:02:36
Starship at this point is Right fire
00:02:38
approval wow what's your
00:02:42
expectation uh or your hope in terms of
00:02:45
the probability that it gets to orbit
00:02:47
it's just a question of timing yeah how
00:02:49
long it does take to get the approval
00:02:52
paperwork whatnot um so that's really up
00:02:54
to the FAA at this point yeah but but
00:02:57
what about making it to orbit do you
00:02:59
think oh
00:03:02
we are doing a new staging technique
00:03:03
called hot staging uh where
00:03:07
um you light the up the upstage engines
00:03:10
or the ship engines while the um boost
00:03:13
stage is still firing and um this is the
00:03:17
kind of the most efficient way to do is
00:03:19
stage separation of rocket going orbit
00:03:21
um but we did not try that on the last
00:03:23
mission and and we're trying on this
00:03:24
Mission we think it will be overall
00:03:26
better um but I think this I think
00:03:29
probably have a
00:03:30
I hope well over 50% chance of getting
00:03:33
to Stage separation um and maybe at
00:03:36
close to 50% chance of getting to Orit
00:03:39
if the hot staging um the new separation
00:03:43
method uh is it works so I'd say maybe
00:03:45
it's like a you know above i' say
00:03:47
probably above 30% chance of getting
00:03:50
toor with the sign whereas previously I
00:03:51
said below
00:03:52
50 uh is this in terms of
00:03:56
complexity how complex is this of a
00:03:59
problem compared to the other problems
00:04:00
you've worked on in your
00:04:03
career uh well so I mean making a rocket
00:04:07
that is more the twice the size of the
00:04:09
SAT 5 um you know it's a in fact with an
00:04:15
extra of the rocket it'll have roughly
00:04:16
three times the thrust of a Saturn 5
00:04:18
Moon rocket um whever it's signed to be
00:04:21
fully rapidly reusable whereas the you
00:04:23
know the SAT was completely Expendable
00:04:27
and with baline we
00:04:30
we still expand the upper stage uh but
00:04:32
we we bring back the boo stage as people
00:04:34
have probably seen the The Rocker liing
00:04:36
videos and we are also able to recover
00:04:39
the varant um with the with falc but
00:04:43
these things do land toly out to sea so
00:04:45
it takes a while to bring them back to
00:04:47
board and get them ready for figh G the
00:04:51
the thing that you know so so this's a
00:04:53
scale of Starship but then also the fact
00:04:55
that it is designed for full and R re
00:04:57
usability so both the booster and the
00:04:59
ship come back launch site they get
00:05:00
caught by these giant meaz arms youve
00:05:03
see in
00:05:04
congress's Godzilla uh it's basically
00:05:08
bad uh catches this giant rocket out of
00:05:11
you know B air and puts it back on the
00:05:14
launch stand and gets ready for launch
00:05:16
so it will be capable of you know
00:05:19
basically aircraft level flight rates um
00:05:23
but but it's much bigger than say a 747
00:05:25
or
00:05:27
8380 um you want can we um talk about
00:05:31
the events of was it last weekend the
00:05:33
whole Ukraine Starling thing can you
00:05:36
give us like a the Tik Tok of like
00:05:39
what's going on and like how you're
00:05:41
being forced to
00:05:45
decide but like what is it like in that
00:05:48
decision room if there was one or
00:05:49
wherever you were where you're trying to
00:05:51
figure out am I keeping this on do I
00:05:53
turn it off what is going on people must
00:05:54
have been bombarding you whatever you
00:05:56
can share about what that was like how
00:05:58
you made the decision
00:06:00
um yeah I I so I was actually mistaken a
00:06:04
little bit in understanding the
00:06:06
situation
00:06:08
um you know obviously we St sex have
00:06:11
provided uh starlink connectivity for
00:06:15
you know to Ukraine um really since the
00:06:17
beginning of the war uh really within a
00:06:20
I think a few days of the war starting
00:06:23
um and as the Ukrainian uh Governor said
00:06:26
the S was instrumental in the FS of
00:06:30
Ukraine so you know they've said that
00:06:32
really many times although the media
00:06:34
forgets to mention that um so and in
00:06:38
fact they've said it on Twitter you X
00:06:41
only know on his
00:06:42
Twitter it's going to take a while to
00:06:45
get that right yeah take a little time
00:06:52
[Laughter]
00:06:54
um so you know you don't have to you
00:06:57
don't have to take my words for it you
00:06:58
just read what what they put
00:07:00
um you know uh
00:07:02
so uh so son has been incredibly helpful
00:07:06
to the Ukraine war eff um we' have gone
00:07:09
out of pocket very significantly uh to
00:07:11
help them um and
00:07:15
um at at the time this happened the uh
00:07:18
region around Crimea um was actually
00:07:22
turned off now the reason it was turned
00:07:24
off was actually originally was because
00:07:26
United States had sanctions against
00:07:28
Russia um and we're not allowed to
00:07:31
actually and that includes Crimea and
00:07:34
the
00:07:34
sanctions and we're not allowed to
00:07:36
actually turn on uh connectivity to
00:07:39
sanction the country without explicit
00:07:41
government approval um we did not have
00:07:44
from the US government so
00:07:48
um so so basically the uh uh you know
00:07:52
Ukraine didn't they didn't give us any
00:07:53
any advanced warning or heads up or
00:07:55
anything um we just got the the sort of
00:07:58
uh urgent calls from the Ukrainian
00:08:01
government saying that we needed to turn
00:08:02
on primier it's like in the middle of
00:08:04
the night basically and we're like what
00:08:06
are you talking about you know
00:08:09
you which's it
00:08:11
four um you know and
00:08:15
then you know we basically um figured
00:08:19
out that this was kind of like a bhit
00:08:21
type attack on um Sasol on the Russian
00:08:26
fet Sasol so they're really asking us
00:08:28
for to to really for actually take part
00:08:30
in a major act War
00:08:33
um and um you know well we so certainly
00:08:38
have huge EMP support for the Ukrainian
00:08:39
government um the Ukrainian government
00:08:42
is not in charge us uh people or
00:08:47
companies that's not how it
00:08:51
works and and Elon if if I could just
00:08:55
but I should say that you know although
00:08:57
I'm not President Biden's biggest fan if
00:08:59
if I had received a Presidential
00:09:01
Directive to turn it on I would have
00:09:03
done so because I do regard the
00:09:05
president as the chief executive officer
00:09:07
of the country whether I want that post
00:09:10
to the president or not I S respect the
00:09:13
office and so if if you know if we
00:09:16
gotten if I gotten a request from the
00:09:18
president type of thing from the
00:09:20
American president to be clear
00:09:24
um then I I I would have turned it on
00:09:27
you know so but no such requests came
00:09:29
through
00:09:30
that that's a really that's a really
00:09:31
interesting point and um you're I mean
00:09:34
the what jamas is referring to is you're
00:09:36
now being attacked I saw there was a you
00:09:38
know there was Jake Tapper uh the other
00:09:40
day on CNN interviewing our secretary
00:09:42
state was just he was all lathered up
00:09:45
basically attacking you for this uh
00:09:47
David David I mean to his credit
00:09:51
secetary Lincoln was actually quite
00:09:53
supportive despite the Absurd uh you
00:09:56
know accusations and leading questions
00:09:58
of the Jake taper at CN yeah he didn't
00:10:02
take the bait credit to sery Lincoln in
00:10:05
this regard for not um you know taking
00:10:09
the bait at all yeah well I to me this
00:10:12
is an example of no good deed goes
00:10:14
unpunished because if you had never
00:10:16
given I hope I hope some good deeds back
00:10:19
go
00:10:21
unpunished I mean if you if you had
00:10:24
never given
00:10:26
Starling that you know
00:10:29
but yeah I mean my point is just if if
00:10:31
you had never given Starling to the
00:10:32
Ukrainian government for free
00:10:34
voluntarily you just volunteered it then
00:10:37
no one be attacking you right now for
00:10:39
not turning it on so they could do their
00:10:41
attack on Crea yeah um also one other
00:10:44
thing I'll note is that your reason for
00:10:46
not turning it on which is you don't be
00:10:48
part of what could be a major escalation
00:10:50
was exactly was exactly the reason that
00:10:53
Biden Administration did not give attam
00:10:56
attam missiles to Ukraine at that point
00:10:59
in the War now they may be changing
00:11:00
their minds but they were very worried
00:11:03
about an attack the administration was
00:11:05
an attack on Crea triggering some huge
00:11:08
escalation of this war so not only did
00:11:11
you not receive a directive from
00:11:14
President Biden your thinking was very
00:11:16
much in line with theirs at the time and
00:11:19
and you're being attacked for that
00:11:21
now there there's something you
00:11:23
mentioned which is that you did this at
00:11:25
a lot of economic cost to SpaceX can you
00:11:28
just can you just talk about that for a
00:11:30
second because I'm not sure people
00:11:31
understand who's paying for what right
00:11:33
now and who hasn't been paid and you
00:11:35
know
00:11:37
Etc there is um well as you say like a
00:11:40
lot of people are contributed to the
00:11:42
effort uh staring is the fundamental
00:11:45
communication backbone of the Ukrainian
00:11:47
uh government and and essential services
00:11:49
like First Responders and that kind of
00:11:50
thing um and you know is is used we we
00:11:54
hope peacefully relatively peacefully on
00:11:56
the warfront it is the only thing that
00:11:57
works on the warfront everything else is
00:11:59
been jamed by the Russians so it the
00:12:02
only thing that works not not one of the
00:12:05
[Applause]
00:12:10
things you know but but I think you have
00:12:13
to sort of think of say the you know um
00:12:17
you know taking the actual example of
00:12:18
Paul har and say like well how did that
00:12:21
work out for Japan it didn't work out
00:12:24
well at all right um because it was a a
00:12:28
tactical victory a strategic defeat it
00:12:31
enraged the American public um who sort
00:12:35
of naturally wanted Vengeance for for
00:12:37
the Z you know the Zach and I think that
00:12:40
that you know while I don't think it's
00:12:42
on the same scale that there was
00:12:43
certainly that potential of sort of a
00:12:46
many P Hava with results in a mass
00:12:49
escalation of uh hostilities um but not
00:12:54
this would this would not this would not
00:12:55
defeat Russia it would enrage Russia do
00:12:58
you Don do you donate the network or do
00:13:00
they pay you for
00:13:02
it
00:13:04
sorry yeah so um I'm actually not sure
00:13:07
what the final accounting is at this
00:13:09
point but uh I I I think at one point at
00:13:12
one point you calculated our sort of
00:13:15
cost of supporting things that roughly
00:13:16
$100 million now the $100 million does
00:13:18
not count um the mass of risk to the
00:13:21
entire Starling constellation uh because
00:13:24
uh Russia would like to have the entire
00:13:26
thing
00:13:27
deleted um
00:13:29
so uh you know nobody's compensating us
00:13:32
for that um and uh so if we were to get
00:13:36
say uh our control center were uh take
00:13:39
down and Cyber attack they you know they
00:13:42
command satellites to the
00:13:44
over um and destroy the entire system uh
00:13:47
or use anti satellite weapons um so you
00:13:52
know the these are this is a pretty
00:13:54
significant
00:13:56
risk um for which we have not received
00:13:58
any compensation and obviously would be
00:14:01
catastrophic to the entire stalling
00:14:03
system which is you know approaching $10
00:14:07
billion Elon do you think the current
00:14:09
government Administration saying hey 10
00:14:11
mil million and then actually I say one
00:14:14
of the BR meing things was as you as
00:14:18
you've seen there's there's a very large
00:14:20
amount of money that's been appropriated
00:14:21
for uh Ukraine you know I'm not sure
00:14:24
what the the total is at this point but
00:14:27
must be 100 close billion or somewhere
00:14:30
between 80 and 100 billion um you know
00:14:34
now all of the you know other sort of
00:14:37
providers us providers of support to
00:14:40
Ukraine are being paid so then why
00:14:43
should SpaceX be excluded that doesn't
00:14:45
make sense we we're doing one of the
00:14:48
most valuable things and yet are getting
00:14:50
the least money this
00:14:51
iser um but you know despite that we're
00:14:55
still happy to keep keep going and um
00:15:00
Elon does the Biden Administration have
00:15:02
it out for you and
00:15:04
why H what have gave you that idea
00:15:07
[Laughter]
00:15:09
yeah but let me ask you own and control
00:15:13
I don't if the whole Administration has
00:15:14
it out me I think there probably aspects
00:15:17
of the administration that are not uh or
00:15:21
or you know as aspects of you know
00:15:24
interests aligned with uh with with with
00:15:27
resent Biden who probably do not um wish
00:15:33
good things for me
00:15:35
um I don't know you know really what
00:15:37
their issue is but there does seem to be
00:15:41
um a significant increase in the
00:15:42
weaponization of government um
00:15:46
and
00:15:47
um I say really sort of misuse of
00:15:51
prosecutorial discretion in a many many
00:15:54
areas where and I think this is this is
00:15:56
really a dangerous thing for
00:16:00
um you know for I don't
00:16:05
know for that to be potis and politics
00:16:08
with with with government agencies it's
00:16:10
it's just
00:16:11
really and then I think from you know
00:16:15
from from say
00:16:17
uh you know Democratic party standpoint
00:16:20
or or say B Biden Administration
00:16:23
standpoint I think this is this the
00:16:25
danger here is that if
00:16:27
there's significant uh um must use
00:16:30
prosecutorial discretion let's say one
00:16:32
says okay everyone's equal under the law
00:16:34
yes but who are you who are you choosing
00:16:36
to
00:16:36
pursue um and if if you if you're
00:16:39
pursuing what what appeared to
00:16:41
Independent voters to be
00:16:44
uh trivial cases while ignoring serious
00:16:49
crimes um it's hard to imagine that a
00:16:52
lot of independent voters that's going
00:16:54
to win over thoughtful independent
00:16:56
voters did this did things change when
00:16:58
you bought
00:17:01
X yeah I think that they did change
00:17:04
somewhat um you know I I'll go with the
00:17:08
with with the sort of you know the
00:17:10
xplatform is really to be uh a Level
00:17:14
Playing Field a public square that is
00:17:18
supportive of um you know most of the
00:17:21
country let's say that the middle 80% or
00:17:23
something like that um
00:17:26
now um that's not been the case really
00:17:29
for all social media that all social
00:17:31
media have been really very very left
00:17:33
leading far left leading and really
00:17:35
Twitter was far left leading um you know
00:17:38
the the suspensions of of um s
00:17:41
Republican candidates or interests or
00:17:44
voices was was uh really at least 10
00:17:47
times the rate of of um suppression of
00:17:51
left left bring voices on you know on
00:17:54
all Twitter
00:17:56
um so so what what I'm trying to do is
00:18:00
move it to the middle which from
00:18:01
standpoint of say the left appears it is
00:18:05
moving to the right Everything's
00:18:07
Relative if you're standing on the
00:18:09
left but it's not it's simply moving to
00:18:12
the middle that's all um in an attempt
00:18:15
to actually represent the whole country
00:18:19
um and and not just um here half the
00:18:22
country or even maybe less than half the
00:18:24
country so that's it really so I think
00:18:27
there's like there's really to be
00:18:29
alarmed about here it's you know it's
00:18:31
just that it's intended to be a Town
00:18:34
Square inclusive of the whole country
00:18:36
and else you know and and the world
00:18:38
that's all it it's been um I guess you
00:18:41
took over um X Twitter on Halloween
00:18:45
weekend if I remember correctly uh when
00:18:48
you got to the building you got the keys
00:18:51
um and David and I were lucky to be
00:18:52
there with you when you got the keys and
00:18:54
we got to um check things out um this is
00:18:57
10 months into a
00:18:59
turnaround uh and it wasn't a high
00:19:01
functioning organization I think when
00:19:03
you took it over where is the company at
00:19:05
now and are you pleased with I guess the
00:19:08
progress because it looks like new
00:19:10
features are getting launched the
00:19:11
product velocity is great uh obviously
00:19:14
advertising's been challenging but it
00:19:16
feels like there's some green shoots so
00:19:18
so how do you feel about the purchase
00:19:21
now yeah well I should say we recently
00:19:23
seen a significant increase in
00:19:25
advertising which is great um so that's
00:19:29
a you know if that Trend continues um I
00:19:32
think the company will be in in very
00:19:33
good Financial shape on the advertising
00:19:35
front um so that in terms of positive
00:19:38
developments that that seems to be one
00:19:40
of them um and um from a feature
00:19:45
standpoint I think those who are look
00:19:48
using the system I think we' I think we
00:19:51
might have delivered more new features
00:19:53
you know in the last I know a year than
00:19:55
in the last you know all tter did in 5
00:19:58
years
00:19:59
you know there there
00:20:00
really feature the feature development
00:20:03
pace is very
00:20:05
rapid um and this is being done with
00:20:07
really about 15% of the original company
00:20:12
um maybe a little more 15 20% um so it's
00:20:16
it's really you know efficient you know
00:20:20
at the end of the day you have to say
00:20:21
you know how complicated is a system uh
00:20:25
like the X Twitter platform um you know
00:20:29
how different is it from a group chat
00:20:31
frankly it's like a group chat at scale
00:20:33
um so he I don't think you need an army
00:20:37
to maintain a group chat
00:20:43
yeah
00:20:44
yeah I mean it's it's not the
00:20:47
self-driving platform and it had maybe
00:20:50
10 times as many people working on it as
00:20:53
the self-driving platform at Tesla which
00:20:56
seems crazy the entire self-driving AI
00:21:00
software team is 200
00:21:02
people and the what they're doing is
00:21:04
much more complex than Twitter or you
00:21:07
know much a lot
00:21:09
more so you now there's other things
00:21:13
that obviously need to be done like
00:21:15
advertising sales um OB network
00:21:19
operations and um how can you talk to
00:21:22
it's really not it's not a it's not a
00:21:25
huge I don't like I said don't even know
00:21:27
me for for for what we're here and I
00:21:30
think you know the people that that are
00:21:32
um still the company are obviously being
00:21:35
very productive in uh creating and
00:21:38
deling new features um and um you know
00:21:42
we keep seeing sort of RedBoard usage
00:21:44
and the the know the most prous number
00:21:47
is really the is user seconds as
00:21:49
reported by the mobile device especially
00:21:52
iOS the the iOS uh what iOS reports as
00:21:57
the screen time is the
00:21:59
is the least gameful metric um and and
00:22:02
those numbers are exem are very good um
00:22:06
so you know I think I'm course justly
00:22:09
optimistic about where things are headed
00:22:11
and I I feel like the company's turned
00:22:14
just you know just recently turned turn
00:22:15
a corner um tell us about you know said
00:22:19
um at least moderate prosperity and and
00:22:21
hopefully significant tell us about um
00:22:25
the success of sharing Revenue why did
00:22:26
you do it and then just the the vision
00:22:29
you have for just the Creator economy
00:22:30
and what you want that to evolve into
00:22:32
and build
00:22:34
into yeah I mean it sounds to reason
00:22:36
that if if you're a Creator and you need
00:22:39
to um need to make a Livy for what
00:22:42
you're doing um so there's got to
00:22:45
be
00:22:47
um you know Fair compensation
00:22:49
competitive compensation for a Creator
00:22:52
whether they're doing you know WR their
00:22:55
writing or pictures video whatever the
00:22:57
case may be
00:22:59
um and uh so we're and we're so we're
00:23:03
not really in you know inventing
00:23:04
anything new here we're just you know as
00:23:06
YouTube does with creators they will do
00:23:08
rev show um with advertising and so
00:23:11
we're doing rev show with advertising um
00:23:14
we're also to have enabled direct
00:23:17
subscription to accounts where whatever
00:23:21
somebody you know you could be doing
00:23:23
audio video long form text anything and
00:23:25
you can subscribe to someone and um
00:23:28
that's you know OB that's way for a
00:23:30
subscriber to make a living as as well
00:23:32
you know for a Creator to make a living
00:23:34
so the intent is for the ACT platform to
00:23:37
be the best home for creators uh where
00:23:40
if you've got interesting content then
00:23:43
you you want to put it on a platform and
00:23:45
um you know there's a lot of questions
00:23:48
about like sort of the algorithm and
00:23:50
whatnot I I should mention like the the
00:23:51
algorithm is uh I think almost all of it
00:23:55
is open sourced and and we will uh I
00:23:57
think quite soon have the entire open
00:23:59
source uh the only reason I it really
00:24:01
hasn't been done entirely open source
00:24:03
yet is because we're somewhat
00:24:04
embarrassed with the code and need to
00:24:06
just clean it up before uh putting
00:24:09
something extremely embarrassing out
00:24:10
there but the point is that like we want
00:24:13
transparency boths trust and if you've
00:24:15
got um uh if you if you can recreate the
00:24:19
results um on the
00:24:21
xform of how viral a post is going to be
00:24:25
independently using the uh you know the
00:24:28
the public algorithm you know the H
00:24:30
algorithm um that that's really where we
00:24:32
want to get to um so you you kind of you
00:24:36
kind of know what to expect um and and
00:24:39
why something happened um now now I
00:24:42
should say the we are we are trying to
00:24:44
optimize for uh user time um on the
00:24:50
platform what this naturally means is
00:24:53
that um
00:24:55
posting content that someone looks at
00:24:57
longer
00:24:59
is going to get higher priority than
00:25:01
content that is
00:25:03
short uh just because the system is
00:25:05
trying
00:25:06
to Max it's it's aspiring to maximize uh
00:25:12
you unreg user remittance is what I call
00:25:14
it so like basically how do we um if
00:25:18
we're succeeding you want to spend more
00:25:20
time on the platform and you want to and
00:25:23
after having spent that time you don't
00:25:24
want to regret it um I mean speaking of
00:25:27
Tik Tok um
00:25:29
you know that's I've had a lot of people
00:25:31
tell me they spend a lot of time on Tik
00:25:32
Tok and they regret it um we don't want
00:25:35
to be we want it to be that you spend a
00:25:38
lot of time on the
00:25:39
xplatform and you learned a lot uh you
00:25:43
you're entertained and you don't regret
00:25:45
it so when you're optimizing for you
00:25:49
know user minutes and like I said
00:25:51
aspirationally under regretted use
00:25:52
minutes uh uh if you the more content
00:25:56
that you post on the system the more
00:25:58
reach that uh thing will get because the
00:26:01
system is saying oh those user is
00:26:04
spending more time in the platform
00:26:06
because they're they're you know seeing
00:26:07
say your podcast or uh reading um a long
00:26:11
form article or watching some video um
00:26:16
that's going to get a lot more time than
00:26:17
say if you if you link to a video
00:26:19
elsewhere or you link to an article
00:26:23
elsewhere that that that's just that
00:26:25
that that that means you'll you will be
00:26:26
on that post for a very short period of
00:26:28
time
00:26:29
and so the the system will be like okay
00:26:30
that did not increase uh user time so it
00:26:33
will it won't be excluded it it will get
00:26:35
less attention that than actually
00:26:38
posting content natively on the system
00:26:40
do you want to talk um
00:26:43
about uh the ADL and you uh sort of
00:26:47
where what the status of that is whether
00:26:49
you're pursuing a lawsuit or not or
00:26:51
where that
00:26:52
stands um I think we'll have to see
00:26:54
about that I
00:26:56
mean um
00:26:58
yeah I mean the fact of matter is that
00:27:00
ad did initiate a boycott they don't
00:27:03
call it a boycott they call it a pause
00:27:05
but you know po that is never ending
00:27:11
isot so and and and we just we saw a
00:27:15
massive drop in uh us advertising we saw
00:27:20
basically no change in advertising in
00:27:22
Asia but domestically with ADL is strong
00:27:25
we saw uh 60% drop in advertising so you
00:27:29
know that's uh pretty intense
00:27:32
um and um and this is despite you know
00:27:36
showing repeated uh analyses of the
00:27:39
system including third party analysis of
00:27:41
the system which actually showed that uh
00:27:44
the number of uh views of hateful
00:27:47
content uh
00:27:49
declined so you know the third parties
00:27:53
who have all the data analyz and said
00:27:55
actually there's less safe speech
00:27:58
um the issue I think with the ADL is not
00:28:01
a question of hate speech it's not a
00:28:03
question of any semitism obviously uh
00:28:06
it's that the ad um and a lot of other
00:28:08
organizations have become activist
00:28:11
organizations um which are acting far
00:28:14
beyond their uh stated mandate or their
00:28:17
original mandate and and I think far
00:28:19
beyond what donors to those
00:28:20
organizations think they are
00:28:22
doing um you know one of the things that
00:28:26
the a was extremely opposed to and in
00:28:29
fact was instrumental in in happening
00:28:31
was there the ad was instrumental in
00:28:32
getting um Donald Trump the platformed
00:28:36
um and then when we we you know we
00:28:38
restored the account um they they made
00:28:42
it super clear that they regarded simply
00:28:44
restoring his account on you know
00:28:47
Twitter now uh that that that
00:28:50
constituted hateful speech like he
00:28:52
hasn't even said anything you know um he
00:28:56
has to at least say something or post
00:28:58
something for there to be incremental
00:29:01
painful painful content this is absurd
00:29:04
um and what's this got to do with
00:29:05
anti-Semitism like you know Donald
00:29:08
Trump's son-in-law is Jewish his Jewish
00:29:09
gr I'm pretty sure he's not anti-semitic
00:29:12
okay um you know he's at the
00:29:15
wedding so um this this so so the
00:29:20
problem is that a lot of these um
00:29:22
organizations like said they've really
00:29:24
been captured by the W agenda and
00:29:25
they're they're pushing um you know
00:29:27
serious of beliefs and values that I
00:29:29
think are often contrary to their what
00:29:32
what they're done as belief and that's
00:29:35
uh that's what we have in the situation
00:29:37
well yeah I I'll note that the the two
00:29:39
positions that you've taken that have
00:29:41
brought the most heat on you number one
00:29:44
defending Free Speech number two
00:29:46
advocating peace and how dare
00:29:50
you how dare you how dare you and
00:29:54
there's there's an article like the
00:29:57
opposite world or something yeah we're
00:29:58
living in upside down World there
00:30:00
there's an article in today's New Yorker
00:30:02
calling you a super villain because
00:30:04
you're advocating peace and protecting
00:30:06
the first amendment I mean it's like
00:30:07
completely upside
00:30:10
down do you want people to eat their
00:30:13
vegetables you at this point you
00:30:15
literally cannot tell actual press from
00:30:19
parody
00:30:20
no like if that was a bon be or
00:30:24
onion no literally you're doing a you're
00:30:27
doing and and change the banner to you
00:30:30
know Babylon b or whatever onion or
00:30:31
something like that par some parody
00:30:33
thing and be like oh that's a good joke
00:30:36
you know um yeah super bones normally
00:30:38
advocate for peace that's you know of
00:30:40
course
00:30:42
um we want to get rid of all the nuclear
00:30:45
weapons hey um
00:30:49
hold that's uh what the funniest the
00:30:54
funniest skit that didn't make it on SNL
00:30:57
that we were worksh Shing was probably
00:30:59
woke James Bond and we wanted to do like
00:31:03
this woke James Bond
00:31:05
and Eli will tell you some of the jokes
00:31:08
it was pretty hilarious but then we were
00:31:10
just talking about a story that broke in
00:31:12
the guardian about the new James Bond
00:31:14
novel and short story is to woke and
00:31:17
it's literally the parody we did two
00:31:20
years ago uh Elon speaking of Peace we
00:31:24
had we had a grais reality also like the
00:31:28
you know the
00:31:30
uh you know conspiracy theories that
00:31:33
that that haven't come true list is you
00:31:35
know quite short
00:31:40
um sees that turn out quite short um and
00:31:43
we really need more conspiracies
00:31:45
generated because we're running out of
00:31:47
to find the
00:31:47
[Laughter]
00:31:50
truth okay check off is accurate um so
00:31:55
uh I don't know who's the you know
00:31:57
responsible these conspiracy theories
00:31:59
but but you know we just need some more
00:32:02
material paging Alex Jones Elon we had
00:32:05
Graham Alison here today I know you
00:32:06
talked about his book we had Ray Doo
00:32:09
here we had Ro kna um and we talked a
00:32:12
lot about China the US relationship with
00:32:14
China you are you have several
00:32:16
businesses that have deep supplier and
00:32:18
customer relation uh ships in China
00:32:21
given what's going on and clearly the
00:32:22
tenor has changed the the mood has
00:32:25
changed with respect to US policy
00:32:27
towards China
00:32:28
what it's like in DC what it's like in
00:32:29
Silicon Valley and how everyone talks
00:32:31
about the relationship with China today
00:32:33
it's pretty crazy how quick things have
00:32:34
changed um as a business leader with all
00:32:38
these business relationships with China
00:32:40
how do you make decisions and and how
00:32:41
things are changing and how do you think
00:32:43
about where this is
00:32:45
headed sure well I mean let's just
00:32:47
clarify here um you know SpaceX has no
00:32:51
uh SpaceX of Starling have no in China
00:32:55
whatsoever they're not they're not
00:32:56
allowed you know SpaceX does launch uh
00:33:00
China satellites and starlink is abandon
00:33:04
in China so to be clear SpaceX starlink
00:33:06
zero business in China um uh in the case
00:33:11
of of Tesla uh one of our of our four
00:33:15
vehicle factories uh one is in uh China
00:33:20
so um you know it's a it's a significant
00:33:23
car market uh but it is uh you know so
00:33:27
what I'm trying to say is like the by
00:33:29
far the bulk of of my business interests
00:33:31
if if I I would peing MTI which I Aspire
00:33:34
not to be um are Outsider China let's
00:33:37
just be clear about that um then with
00:33:40
respect to now that said I think I
00:33:43
understand China well I've been there
00:33:45
many times I've met with uh the senior
00:33:47
leadership um at many levels of China
00:33:50
for for many years and so I I think I've
00:33:53
got a pretty good understanding um at
00:33:56
least as an outsider of ch so and and
00:33:59
Tesla has been very successful
00:34:00
domestically in China
00:34:03
so
00:34:04
um you know the fundamental thing here
00:34:07
is is really Taiwan um
00:34:11
the China has well really since uh for
00:34:16
like half a century or so uh maybe
00:34:19
longer at this point longer at this
00:34:20
point their policy has been to to um s
00:34:25
reunite Taiwan with China uh from this
00:34:28
standpoint you know it may be just
00:34:29
analogous to like Hawaii or or something
00:34:31
like that like an integral part of China
00:34:34
that is arbitrarily not part of China um
00:34:38
mostly because of the US stop has us
00:34:42
Pacific Fleet has stopped uh any any
00:34:45
sort
00:34:46
of um reunification effort force um so
00:34:52
now really things getting to the point
00:34:55
um increasingly year-over year uh where
00:34:58
China's military strength is increasing
00:35:00
and ours is more or less uh
00:35:02
static and strategically you know you
00:35:05
can imagine trying to defend Taiwan is
00:35:09
not not easy because it's it's very
00:35:11
close to the coast of China um so there
00:35:15
will come a point if you know Pro
00:35:18
probably not not too distant future
00:35:20
where China's military strength in that
00:35:23
region far ex exceeds US military
00:35:25
strength in that region and
00:35:28
if one is to take uh China's policy
00:35:31
literally and probably one should um
00:35:34
then there will be some
00:35:36
forceful uh for forceful beuse for you
00:35:41
know uh to incorporate Taiwan into
00:35:44
China this is what they've said um that
00:35:48
if if there is not a diplomatic solution
00:35:50
there will be a solution by force let me
00:35:54
um if I can and so really what's going
00:35:57
on here you've seen you know this in in
00:35:59
many areas and I think this Tempo is
00:36:01
going to increase is that you know both
00:36:05
China and the US are preparing for a
00:36:10
potential
00:36:12
Showdown uh you know in the South China
00:36:14
CA
00:36:16
so um that's why you're seeing
00:36:19
increasing restrictions on export of us
00:36:21
technology to China uh but such as the
00:36:24
the Nvidia you know the Nvidia h100s
00:36:27
being and you not allow to ship to China
00:36:30
um and I think there'll be more and more
00:36:31
you you also not not allow to ship
00:36:33
Advanced ship making equipment to China
00:36:36
um
00:36:37
so and I suspect you know you know China
00:36:39
is going to respond with some reciprocal
00:36:41
sanctions um and you'll I think you'll
00:36:43
see this kind of a TI T reciprocal
00:36:46
sanctions increasing in the next next
00:36:48
few years so I think quite a very uh hot
00:36:52
temperature um and then we'll see this
00:36:56
is there going to be a a diplomatic
00:37:00
solution to uh re reunification or a
00:37:04
non- diplomatic solution you uh but has
00:37:08
made it clear that there will one way or
00:37:10
another be a solution from their
00:37:12
standpoint Yeah you mentioned uh Nvidia
00:37:14
so let me just talk about Ai and bring
00:37:17
it back to that for a second can you
00:37:18
tell us um your regrets but also the
00:37:24
positives of the experience you had with
00:37:26
open Ai and then what your goals are
00:37:27
with
00:37:33
X well the AI discussion is is certainly
00:37:36
a long one or could be a long one
00:37:46
um you know digital super
00:37:49
intelligence that might be the most
00:37:51
significant technology that Humanity
00:37:53
ever creates um and and it has the
00:37:56
potential to be more dangerous than and
00:37:58
um nuclear weapons
00:38:01
so
00:38:06
um you know in the case of creating
00:38:07
opening eye it was to have there not be
00:38:10
a unipolar world where um Google with
00:38:14
its subsidary deat mine uh you know
00:38:17
would control an overwhelming amount of
00:38:19
AI talent and compute and and resources
00:38:24
um which then is somewhat dependent on
00:38:27
basically how
00:38:28
how Larry paage U and serge BR um and
00:38:33
Eric believe things should go because
00:38:35
they they between three of them or two
00:38:36
out of three have control over alphabet
00:38:40
because they've got super voting rights
00:38:42
and um you know I was quite based on
00:38:45
some conversations I had with lar paig
00:38:47
uh where um you know you call me a
00:38:51
species for being pro
00:38:53
humanity and
00:38:55
um so I'm like what side of you on there
00:38:58
you know
00:39:00
uh
00:39:02
um you know I think and uh so so I felt
00:39:05
like
00:39:06
uncomfortable um having the entire
00:39:09
future of digital super intelligence be
00:39:11
in the hands of someone who equ a
00:39:12
species for being pro Humanity
00:39:15
um you know how can it not be uh so
00:39:20
that's opening eye was originally
00:39:22
created as an open source nonprofit and
00:39:25
now is a close to be it should be
00:39:28
renamed closed for Max no profit AI
00:39:32
um it is it is closed um and they are
00:39:35
aiming to I think make try to make a100
00:39:37
billion I think according to
00:39:40
S get $100 billion from somewhere for
00:39:43
some vast amount of compute uh to create
00:39:46
digital god um apparently all the waste
00:39:50
are stored in a common separated value
00:39:51
file by the way so our dig God will be a
00:39:54
CSV
00:39:55
file how do we import
00:39:58
it file
00:40:01
import um see what happens um
00:40:06
so so anyway so so now
00:40:11
opening uh is also very closely line
00:40:14
with Microsoft you know Microsoft is
00:40:16
really you um the open servers are
00:40:19
running on in Azure and Microsoft data
00:40:21
centers you know so really what you have
00:40:24
is I think at the end of day Microsoft
00:40:26
having more control than open AI they
00:40:28
have access to all the source code they
00:40:30
have access to all the weights of the um
00:40:33
you know gb4 and future versions so they
00:40:36
have all rights to this to to thing it's
00:40:38
not
00:40:40
um at any point really they could cut
00:40:42
off open AI I don't think open AI quite
00:40:44
realizes that the dependence on on
00:40:48
Microsoft and even if Microsoft does
00:40:49
break some contract they'll just be tied
00:40:51
up in litigation for you know for years
00:40:54
um so really you've got
00:40:57
a contest between kind of like Google
00:41:01
and Microsoft Google as mentioned I'm
00:41:04
concerned about you know uh sorry not
00:41:07
not caring enough about AI safety and um
00:41:11
good reason and then Microsoft just is
00:41:14
is a I think you know a profit seeking
00:41:17
organization um and I you know I think
00:41:20
such is great but um I I can't say like
00:41:24
you know that it would difficult to to
00:41:26
say that that m has a has an amazing
00:41:28
track record and moral decision-
00:41:32
making so um
00:41:38
diplomatic anyway so so so was like okay
00:41:41
look let's just so I think let's try to
00:41:44
create a third company that is
00:41:47
competitive I I I do think Tesla is
00:41:49
underated from an AI standpoint in terms
00:41:51
of real world AI Tesla has the best real
00:41:54
world AI
00:41:56
so uh you know hopefully between uh xai
00:42:00
and Tesla there's kind of a third
00:42:03
Contender or would you look you've done
00:42:07
you open source your patents at Tesla
00:42:10
you are very Pro open source your source
00:42:12
code at X would you ever considering
00:42:14
releasing dojo and FSD more as a
00:42:17
platform substrate for everybody else or
00:42:19
that's sort of off the table right
00:42:21
now
00:42:23
well I don't know if that's uh you know
00:42:26
in the case of say Dojo or our inference
00:42:31
Hardware that's in the car our inference
00:42:33
inference computer which is actually a
00:42:35
lot lot more comp than Dojo by the way
00:42:38
um you know we've got I know somewhere
00:42:41
in the order of 4 million cars that have
00:42:43
um highp speeded Ai and Par computers in
00:42:46
them
00:42:48
um like open sourcing chip designs
00:42:51
doesn't mean you you suddenly get that
00:42:52
thing yeah you know
00:42:55
so um
00:43:00
you can open open source the software
00:43:02
but I think chip designs
00:43:03
it's only ones that could actually use
00:43:07
those chips or really kind of yeah would
00:43:11
be some someone that's willing to spend
00:43:14
many billions of dollars on um on a
00:43:18
computer development
00:43:21
so anyway I think I think in the case of
00:43:24
uh you know don't saying Optimus is
00:43:26
really
00:43:27
interesting um any I think just in
00:43:31
general Tesla is uh one of the most
00:43:35
leading AI companies um and in some
00:43:37
respects the leading AI company when it
00:43:39
comes to real real world AI
00:43:42
understanding the real world and and
00:43:44
actually reacting to that with
00:43:46
self-driving
00:43:48
um and so and I think that will become
00:43:50
part of the the solution for AGI or
00:43:53
general super intelligence so
00:43:58
um in the case of Tesla I think we've
00:44:00
got a sort of a good governance
00:44:03
structure in that there's no super
00:44:04
voting rights or anything like that so
00:44:06
if I'm you know go crazy the CH of Tesla
00:44:10
can vote me out um you know I have
00:44:14
enough vote to be you know I think
00:44:16
moderately influential but not enough to
00:44:18
stay in even if I'm doing crazy stuff so
00:44:21
I think that's actually good
00:44:25
um great
00:44:27
um I was told we have to wrap him at oh
00:44:29
okay uh just on FSD before we wrap I'll
00:44:31
let you go um we were talking earlier
00:44:34
this year and you said uh hey maybe chat
00:44:37
GPT 4.0 like moment for self-driving was
00:44:41
coming and uh I've I've been playing
00:44:43
with the beta and um yeah how how close
00:44:47
does it feel to you because it it some
00:44:50
of the rides it's been doing for me are
00:44:52
pretty darn impressive the latest beta
00:44:55
is pretty incredible yeah it's pretty
00:44:57
pretty neat I you know I used to love it
00:44:58
on the highways and on the streets I'd
00:45:00
be like okay but now I'm using it
00:45:02
increasingly on the streets so where do
00:45:04
you how do you feel about it right now
00:45:06
and I I guess you made a lot of
00:45:08
predictions on it over the years um but
00:45:11
it it does feel like it's getting pretty
00:45:13
close yeah I think it's I think it's
00:45:15
very close to uh you know being in a
00:45:18
situation where even if there's no human
00:45:20
oversight or intervention that the
00:45:21
probability of uh a safe journey is is
00:45:24
higher with FSD and no supervision like
00:45:28
even if you're seep of the car than if
00:45:30
the person is driving um we're very
00:45:33
close to that you know those that have
00:45:36
the FSD beta which really anyone could
00:45:39
get at this point um
00:45:42
so
00:45:43
the the the miles we see driven under
00:45:46
the FSD beta currently are much safer
00:45:49
than the miles that are driven without
00:45:50
it so um that's uh you know that's
00:45:54
that's already very good milestone
00:45:58
um but you know you can just see that
00:46:01
it's getting better and better like um
00:46:03
if you see if you compare the uh you
00:46:06
know FS beta today versus 6 months ago
00:46:09
versus you know a year ago versus 18
00:46:12
months ago it's really the Improvement
00:46:14
is
00:46:14
dramatic um and um we've got the final
00:46:19
piece of the puzzle which is to have the
00:46:22
control part of the car uh transition
00:46:24
from about 300,000 lines of C++ code to
00:46:28
also neural network so the you know the
00:46:31
whole system will be neural net in your
00:46:33
network um phons into controls out and
00:46:38
and that that that's kind of the final
00:46:39
piece of the puzzle for fullop driving
00:46:42
being significantly better than human W
00:46:44
awesome uh thanks for taking the time
00:46:46
buddy uh fly safe and I'll see you
00:46:48
shortly uh ladies and gentlemen Elon
00:46:52
mus thanks bud
00:46:55
[Applause]
00:46:55
[Music]
00:47:02
and instead we open sources to the fans
00:47:04
and they've just G
00:47:06
[Music]
00:47:14
crazy
00:47:17
Bes myg taking
00:47:20
[Music]
00:47:22
driveway oh
00:47:24
man we should all just get a room just
00:47:27
have one big huge orgy cuz they're all
00:47:29
useless it's like this like sexual
00:47:30
tension that they just need to release
00:47:33
[Music]
00:47:38
Som we need to get
00:47:41
[Music]
00:47:47
merch I'm going all
00:47:50
[Music]
