Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
So we've had a very exciting quarter be we're able
to successfully launch a rover taxis so providing out both
riders with no one in the drive to see with
paying customers in Austin. And then so some of they've
known we've already expanded our service area at Austin. It's
bigger and longer, and it's it's going to get even
(00:20):
bigger and longer the we we're expecting to really greatly
increase the mortion service area too, well in excess of
what com Feder's doing. And that's hopefully in a week
or so. Two weeks, yeah, a couple of weeks, a
couple of weeks or so, and we're getting the regulatory
(00:41):
permission to launch in the area about Arizona and the
number of at Florida, a number of other places. So
as we get the approvals and reprove out safety, then
we'll be launching the autonomous right here in Wislim country.
And I think we're oubly have autonomous right handling in
(01:04):
the half of the population of the US by the
end of the year. That's that's least a goal. Subject
to regulatory proofs, I think will technically be able to
it so it's being weird regulatory rubles. It's probably interesting
half the population of the US at the end of
the year. But we are very very cautious. We don't
(01:25):
want to take any chances, and so we're working out, yeah,
they'll go cautiously, but that's the service areas and the
number of vehicles and in operation will decrease a hyberate
natural rate. So you add some other notable things word
(01:46):
away and dream. We jam the best selling car in Turkey, Netherlands,
Switzerland and Austria. It is I believe the best selling
car already fived in the world still and the top
of it. Autonomy is a is a big factor there,
so give about even absent and people about supervised even
even if it's just supervised, so self driving it's it's
(02:08):
a huge selling point. And it's worth noting that we
do not actually yet have approval or supervised FSD in Europe,
where so our sales in Europe we think will improve
significantly once we are able to give customers the same
experience that they have in the US. This is why
(02:29):
this is I think a very important point to convey
and we've been working with the Big Country regulator which
is the Netherlands, and I think we're close to getting
approval with Netherlands. Then instead of go to the EU.
It's quite a cockca esque, but to copa no idea
that's looking at the EU coulds just beyond concas challenges
(02:50):
with eurocracy, but it will. We will get the approvals
and I think we'll get you know, some people in
Europe will have an experience and out of the US
in most of Europe this year, hopefully at least partially
in this quarter. And then we also have some regulator
(03:12):
regulatory challenges in China which we're hopefully to unblocked shortly
where we because we also cannot find super vice qusty
in China currently, but we got to unblock that suit.
And that's a student that's another major. It's really is
the single biggest demand drive. And then within the US,
as we get confident about safety in different geographic areas,
(03:36):
the we will loosen up on how much somebody has
to be layer of focused to have the eyes layer
of focus on the road, you know, because that's that's
been a common complaint. In fact, it does create an odd,
odd safety issue where people will senders disengage or a pilot,
then then do something change the radio or AMU look
(04:00):
at the bone drivers their knee and then re engage
water pilot, which obviously is not is less safe than
simply keeping autopol on. So anywhere, get that that that
experience will will improve in the next several weeks. The
because when I focus on Awestem with Dewain driver's seat,
(04:20):
the production release of water Pilot is actually several months
behind what people experience on a road tax and most
so not Now we have the road tax that was
launched will be provided adding back those elements so that
they will be a step change improvements and the water
Pilot experience or people outside of Worsting. This is really
(04:44):
as you can tell, this is very much sort of
an autonomy. Is the story like we need the We
need the physical product with that which you can dipatonomy.
But once you have a physical product, you need the
autonomy is what amplifies the value to stratospheric levels. You
also launched the Tesla dina, which has been a huge hit.
Actually got world wide attention, which is unusual for a
(05:05):
dina that is don't typically get headline news around Earth.
But this is a pretty special dynas and if you're
in the LA area, it's worth visitting. It's sort of
a shiny beacon of hope and otherwise sort of leap
up and landscape rightly. So it's really quite quite a
lovely experience where don't played. There's little team there up
(05:28):
on the pulls up garbage front. Continue to work about that.
We have. We're continuing to make significant improvements just with
the software. So the we're expecting to increase the parameter account. Actually,
at this point, I think we think we can probably
ten x that almost tax the Pepper prim Yeah, roughly
(05:48):
roughly ten x the prim account. So this is actually
a very tricky thing to do because if you as
you increase criameter account, you get you get to choke
on every bad wind. But we currently think we can
tenext the parameter count from what people are currently experiencing.
That's that's not just support x, actually ten x increasing
(06:11):
in primary account. And yeah, so that's still a lot
of improvement on the existing artware. To happen. Energy is
growing really well despite edwards from tariffs and spirit suppliettery
jodges the very package bagistic quickly and how we have
ouprates the bagapact that will make it even better. And
(06:32):
we had red board our powerful to plates and dat
H two. So where I think bat batteries are just
going to be a massive thing, that that the scale
of batteries battery deman is, I think not that many
people appreciate just how gigantic the scale of battery man is.
(06:54):
The way you think about it is that the US
sustained power output of the U S grid is on
one terrat wet, but average usage is less than half
at tarot wet. If you add batteries to the mix,
you can run the power plants twenty four to seven
ex soul capacity. That's double a more than doubling the
(07:15):
energy output per year of the United States just with batteries.
But that's again big deal. It's a really big deal.
Speaker 2 (07:22):
Optimist to what we're revolving the optimist design to really
get an optimist at the point where it is a
phenomenal design to pur an optimist Version two right now
sort of two and a half.
Speaker 1 (07:36):
Optimus three is if there is an exquisitive design, in
my opinion, there will be an credit. As I've said
many times before, I predicted it will be the biggest
product ever. It's a very hard problem to solve. You
have to design every part of its from physics first principles.
There's nothing that's both the shelf that actually works, so
(07:57):
you're going to design every motor, gear box, par electronics, control, electronics, sensors,
the mechanical elements also got to train optimists to use
its eleven sented sensors with a neural net, but will
be applying the same techniques that we're applied. Or a
(08:19):
car which is essentially a four wheel robot, and optimist
is a robot which arms and legs, so that the
same principles that apply to optimize eight AI inferns are
the car applied to optuments because they are both really
robots in different forms. And Tesla, it is wortant to
(08:40):
know the Tesla is my father best in the world
at real world AI. Like a clear proof point that
would be if you're compared to Tesla to way more
where it's got. You know the car is best doing
with god knows how many sensors, and yet doesn't Google
goodn't AI, Yes, but they're not good at real best
(09:00):
part the Tesla is actually much better than Google by
far and any much anyone at real world and and
I and by far, Chesla has the best infernce efficiency.
Like I think a key for AI is what is
the intelligence per gigabytes? People know about parameters blah blah blah,
(09:24):
but I think, but about toil about gigabytes because with
the parameters can have four bit parameters, eight bit parameters,
sixteen bit parameters. But the actual constraints in the hardware
are how many gig bytes run and how many gigabytes
per second he transfer from. Therefore, it is not the
parameters straight, there's it isn't there's a byte straight And
(09:48):
Tesla has the highest intelligence density of any ANI by far,
and I have a lot of insight this with with
x Ai. Xai is grock is the smartest of the
I over but it's a BROP four is a giant
beast there with the terrorbyte levels. And that's kind of
(10:08):
an important to know. It telles that has the best
collegian students intelligents density will be a very big deal
of the future. It is no so with Optimus three,
which is really the right design, it's like it doesn't
happen at this point, there's no there's a SIGNI think
claws would be Octimist three design, but we are going
(10:28):
to retool a bunch of things, so it's it'll probably
be prototypes of Optimist three end this year and then
scale production next year. We're going to try to scale
Optimist production as fast as it's seemingly possible to do. So,
try to get two million units a year as quickly
as possible. We think they give them less than five years.
That's I guess that that's a reasonable aspiration. Is the
(10:52):
million units a year five years? It seems like an
indieval tarry A conclusion. So four twenty White has been
breaks a year. A lot of major monstudents made clear
launch progress in autonomy that a lot of Masaya said
we would not achieve. It's worth noting that the we
(11:13):
have done what we said we're going to do, and
they were always on time. But we get it done
and there sitting there with egg on their face. So great, great,
great progress by it tells the team, Yeah, I do
think it tells continues to execute well with big autonomy
and and the humor or what autonomy, it will be
(11:37):
the most valuable helping in the world. There's a lot
of execution between year and there. It doesn't just happen,
it's about it excute very well. I think Tessant hasn't
shown it being most company in the world obviously, I'm
stremely optimistic about future of company. Best way to break
(11:57):
the futurist and make it happen maging. It happened here
with those team. So I'd just like to say thanks
to one of our supporters, and I think we're get
to tell you incredibly exciting future. I think it's it'll
be available for vice personal use by the end of
this year and certain geographies. We're just being very careful
about it and we're just being extremely partner. But I
(12:19):
would be a copy of that by the others year
within a number of cities in the US. We will,
we will, It'll be available to edge users like the Optimist.
Optimist three design, as I mentioned earlier, is I think
finally the right design. They'll wait for the optimizations, but
they're will. I think they're fundantal changes too that are needed.
Over the Optimist three design. It has all the degrees
(12:41):
breedom that you really want or need, so it's you.
I'll have bonotypes of that and three three months and
means suctually in production postimply start production on that next
year the production round. Simple stories conflict the s cove
(13:02):
up your production around. Something is got an entire when
everything is new, because the rate of production will move
this fast is the least lucky and least confident elements
of the entire supply chain, as well as internal processes.
So the more new STU that is a in a product,
the slower than the rank could be because of unexpected
(13:25):
supply chain interruptions or mistakes internally. Especially using it to
predict through the end of the of the escove or
or late in the S curve. In the beginning of
the ESCO and the beginning of the ESCO of the
production round is a case not really materials for revenue purposes.
The beginning of the ESCO, you're usually yours negative course, Morgan,
(13:50):
you're debuting amount of issues. So that's why it's it's
I feel like fairly confident predicting things. Well, please medium
confident in predicting where we are five years, but it's
hard ricked where we are in a year or two years.
So that's why I think five year us I think
(14:10):
it could be at the living with our be surprised
at the end of five years, sixty months from now
get three online roughly making one hundred thousand Optimus robots
in month in sixty months, I would be shocked, I think, obviously. Yeah,
we're a publicly treaded company. Sure Holders are welcome to
put forward any shareholder proposals that they'd like. I recently
(14:34):
encouraged that, and then have shareholders vote and we'll act
in accordance with the shareholder wishes. Great. Yeah, there's a
lot of exciting things happening in the science studio. It's
like static and really what's going to happen over the
next to several years is a fundamental transformation of the
company from the pre asolomy world to a post autonomy
(14:56):
and I'm working on a near master plan to articulately
that tells the team and you know there will there
will be there are some teething pains as you as
you transition from pre pre aufineries vote autonomy world, but
I think that the future tells it is incredibly exciting
and will profoundly change the world in a good way.
(15:17):
Please sound like fiable whatever, but I think we'll let's
just say if we execute on that plan effectively, which
is you have to actually do that because it will
be the most vaguable company in the world by far.
Are two. We expect to have Dojo to operating scale
SubTime next year, with scale being somewhere around one hundred
(15:40):
k a twelve hundred equivalents. And AI five, which is
really it's spectacular use those words likely spectacular to the
A five Childen being hope hopefully be buying production around
the end of next year, but that has a lot
of potential. I think, you know, thinking about DOJOH three
(16:03):
and the AI six birdst chip, it seems like, intuitively
we want to try to find conbudgets there where it's
basically the same chip that is used what were you used,
say two of them in a car or an optimist
and maybe a large a number on board account of
(16:28):
by twelve and AB or something like that. If you
want highbands communication between the chips out of course serving
they're drink it bird serving. That sort of seems like
intuitively the sansful way to go.
Speaker 3 (16:42):
Yeah, well, the the cybercap is which is really optimized
for autonomy that that I think has got probably somebody's
sent per mile potential over time by betweenty five.
Speaker 1 (16:58):
You know, it's really like if you design a car
from scratched to be a cost optimized robotic taxi like
Cybercap like for example, we're not trying to make it,
you know, corner like a look incredibly well like a
Model three would, or modeless or even mall way like
(17:20):
it's got moddel all about all about. Cars that are
driven by people are super fun drive. They've got incredible acceleration,
you know, the incredible cornering capability. But not ben We're
confident that very few people and a cybercap want to
be totaling around. So you know, so we were we
(17:41):
produce the top end speed, which means we can use
more efficient tires, we don't use as much acceleration, we
don't use as much you know, big big brake system.
We won't want stopping distance, but we're not expecting it
to be heavily used. If it's it's a gentle right.
Essentially you designed for a ride and then you you
(18:03):
have am that much optimized optimized asign point. So that
that's why it timpts probably achieve that, especially optim optimuses,
you know, so asserving you cleaning up the car and
make doing maintenance and stuff. It's during automatic charging, so
I think it's gonna actual cost for a mile of
(18:26):
cyber cab will be very low. The cost for a
while our existing I fee will be higher but still
very competing. So breviously number of fifty sets and discussing, Yeah,
this really fills road taxically. It will build from tiny
to gigantic in terms of operations in pretty short period
(18:47):
of time. Like my guess is there. It has material
impact on our financials around the end of next year.
Yeah that that that is a major CONSERVETI as I
mentioned in the past, and I hope that is addressed
to the upcoming shareholders meeting. But yeah, this is a
big deal. Like you know, I want to find that
(19:08):
I've bet like so a little control that I can
usually be ousted by active as shareholders after having bolts
you know these all if you're going robots. I think
that as I mentioned before, I think my control of
Hotaila should be enough to ensure that it goes in
a good direction, but not so much control that I
can't be thrown out if I go crazy. Yes, it's
(19:30):
a bit of a tough thing because like when we
do an aid. We find that some was literally done
it pray by bray examination of our slightes everything we say,
and they're coveyos. So you know of say like, well,
what's the traidal. It does help with's recruiting, but then
competitors look break clessly in coveyots. I mean that said
(19:53):
we should probably I mean, I guess we could consider
the shareholder beginning to be sort of an A can
do if we can maybe go to depth, some solid
amount of depth at the ill share Older meeting with
respect to Optimus and a AI and that sort of
our chip chip stuff. Perhaps, Yeah, the test also really
(20:14):
underrated in terms of a chip design as well as
AI software. So like there's a whole lot that chip
that we were that that exists that we will prefer
to bringing our car, that is, ANYI chip that then
we would prefer to putting the car over our own.
And even though it's been now out for several years
and recognan that the air bike chip will be for
(20:37):
a Vale game tream, in fact, it's it's so powerfully
we'll have to interurb it or sun grief or pockets
outside of the rings because it flows weary past the
export restrictions. So let'sten export restrictions change. We actually will
have to know if our ai bib ship is look
kind of where hopefully list hopefully we keep raising the
(20:58):
bar and the other way the kids bit silly. We'll
have a bunch of Optimist robots on stage at the
shareholder meeting. The Optimist lab is cool to see. It
looks like it basically looks like the state of Westwold
get robots in various stages, some of them, you know,
in various stages of prepared, like I don't know. It's
(21:20):
some combination of like the tattooing Chukyard and the and
West Wolds. It's pretty cool. An octopus is walking around
the office here in polls there are twenty four seven
is just walking around like it's small, I think. And
we're so optimistic, you know the t A diners that
(21:42):
we're bub born. Yeah, so it's we'll go for a
world where robots are rare to where they're so common
that you don't even look up. Okay, we haven't really
thought hard about that. We need to make sure it
works when the bagels are pull the under our control
and it's kind of one step at a time. Here
we don't want to jump the gun. As I said,
(22:03):
we're doing parallered about safety. So so it's just like
but I guess, I guess like next year, I'd say
competently next year. I'm not sure when next year, but
competently next year. People who will be able to add
to subtract their part to the Testla fleet. I'm just
like at their bab or ing prossible well as seem
(22:25):
as there is a clear cash blow stream associated with
any any product that you can get finance it. Most
people still done there the master yard. People don't know
it exists, and it's still like half of Tesla owners
who could use it, I haven't tried it even once
this They don't actually rely as this is something we
(22:47):
want to educate them on. So we've got to. When
they come in for service, we'll reach out to them,
send them like videos of how to make the work
and most much it's it's such a shocking thing. They
don't think a car is capable of this. So you
have Jack, you have just show them and get them comfortable.
Alternally don't know if it's so trivial, but it's it's
(23:10):
like saying you've got a Canada sing a dance, but
it just looked like an old cat. They're like, you know,
until you see the cat sing a dance and talk
like your series just a cat, that's that's tes I see.
Our car is intelligence. Well, we're in this like weird
transition period where we will lose a lot of intatus
(23:30):
in the US. Incentives actually ready other parts of the world,
but roll is them in the US, and it will
still have at the relative early stages of autonomy. On
the other autonomy is most advanced and most available from
a regulatory standpoint in the US. So I mean, does
that mean like we could have a few rough quarters. Yeah,
(23:50):
we probably could have a few rough quarters. And I'm
not saying the world, but we could Q you know,
Q four, Q one, maybe Q two. But once you
get to autonomy at scale in the second hand of
half of next year, certainly by the end of next year,
I think the I'm surprised to the economics, I'm not
very simpellent because it's just a model. Why they look
(24:14):
the cat out of the back there dancing cut that
consilingly bess, But it can be. It can talk and think.
But that's the cool part yeah, I mean the fundamentally,
the biggest obstacle remains that people just don't happen. Some
of the people don't know that. The desire to buy
the car is very high. Just people don't have enough
money in their bad account to buy it. Literally, that
is the issue, not a lack of desire, but lack body.
(24:38):
So the more affordable we can make the car, the
better I think it's gonna be. It will be a
very big deal when people can release their car to
the fleet and have it earned money for them, which,
like I said, i'm big. I feel confident saying that
will happen next year in the US at least, and
in the US where we're legally allowed, you appropriate displayers
(25:00):
and they'll make the affordability dramatically greater, just like you know,
if you have Airbnb and you ran out your home
when you're not there, or ran out of a guest
room or guest house or something like that, you're the
affordabilit if your home is much greater. Well, they are
doing different things here, so you know, the Xai is
doing like you know, terabyte scale models and the multi
(25:25):
terrabyt scale models. Tesla's one hundred times smaller models. So
once real walth AI and once kind of I guess
art official superintelligence tech thing. The I mean really kind
of the genesis for XAI was that there were there
were certain people who some people would not join Tesla
AI engineers because they wanted to look at ASI. They
(25:48):
would join Tesla, and I was like, well, maybe they'll
join a new company. And now I think the Tesli
bomb is extremely important, but not everyone agrees with me
on that. And so rather than have them join and
you know, open a eye or people or or some
other company, it's like, as well have them create a
company in that regard, which like say, I so that's
(26:11):
you know, and you know, pilin make it. Make a
decision do they want to work on It's like super
intelligence data center or real wealthy idens good. They're both
compelling problems, but some people work want to open one
and some one of them the other
So we've had a very exciting quarter be we're able
to successfully launch a rover taxis so providing out both
riders with no one in the drive to see with
paying customers in Austin. And then so some of they've
known we've already expanded our service area at Austin. It's
bigger and longer, and it's it's going to get even
(00:20):
bigger and longer the we we're expecting to really greatly
increase the mortion service area too, well in excess of
what com Feder's doing. And that's hopefully in a week
or so. Two weeks, yeah, a couple of weeks, a
couple of weeks or so, and we're getting the regulatory
(00:41):
permission to launch in the area about Arizona and the
number of at Florida, a number of other places. So
as we get the approvals and reprove out safety, then
we'll be launching the autonomous right here in Wislim country.
And I think we're oubly have autonomous right handling in
(01:04):
the half of the population of the US by the
end of the year. That's that's least a goal. Subject
to regulatory proofs, I think will technically be able to
it so it's being weird regulatory rubles. It's probably interesting
half the population of the US at the end of
the year. But we are very very cautious. We don't
(01:25):
want to take any chances, and so we're working out, yeah,
they'll go cautiously, but that's the service areas and the
number of vehicles and in operation will decrease a hyberate
natural rate. So you add some other notable things word
(01:46):
away and dream. We jam the best selling car in Turkey, Netherlands,
Switzerland and Austria. It is I believe the best selling
car already fived in the world still and the top
of it. Autonomy is a is a big factor there,
so give about even absent and people about supervised even
even if it's just supervised, so self driving it's it's
(02:08):
a huge selling point. And it's worth noting that we
do not actually yet have approval or supervised FSD in Europe,
where so our sales in Europe we think will improve
significantly once we are able to give customers the same
experience that they have in the US. This is why
(02:29):
this is I think a very important point to convey
and we've been working with the Big Country regulator which
is the Netherlands, and I think we're close to getting
approval with Netherlands. Then instead of go to the EU.
It's quite a cockca esque, but to copa no idea
that's looking at the EU coulds just beyond concas challenges
(02:50):
with eurocracy, but it will. We will get the approvals
and I think we'll get you know, some people in
Europe will have an experience and out of the US
in most of Europe this year, hopefully at least partially
in this quarter. And then we also have some regulator
(03:12):
regulatory challenges in China which we're hopefully to unblocked shortly
where we because we also cannot find super vice qusty
in China currently, but we got to unblock that suit.
And that's a student that's another major. It's really is
the single biggest demand drive. And then within the US,
as we get confident about safety in different geographic areas,
(03:36):
the we will loosen up on how much somebody has
to be layer of focused to have the eyes layer
of focus on the road, you know, because that's that's
been a common complaint. In fact, it does create an odd,
odd safety issue where people will senders disengage or a pilot,
then then do something change the radio or AMU look
(04:00):
at the bone drivers their knee and then re engage
water pilot, which obviously is not is less safe than
simply keeping autopol on. So anywhere, get that that that
experience will will improve in the next several weeks. The
because when I focus on Awestem with Dewain driver's seat,
(04:20):
the production release of water Pilot is actually several months
behind what people experience on a road tax and most
so not Now we have the road tax that was
launched will be provided adding back those elements so that
they will be a step change improvements and the water
Pilot experience or people outside of Worsting. This is really
(04:44):
as you can tell, this is very much sort of
an autonomy. Is the story like we need the We
need the physical product with that which you can dipatonomy.
But once you have a physical product, you need the
autonomy is what amplifies the value to stratospheric levels. You
also launched the Tesla dina, which has been a huge hit.
Actually got world wide attention, which is unusual for a
(05:05):
dina that is don't typically get headline news around Earth.
But this is a pretty special dynas and if you're
in the LA area, it's worth visitting. It's sort of
a shiny beacon of hope and otherwise sort of leap
up and landscape rightly. So it's really quite quite a
lovely experience where don't played. There's little team there up
(05:28):
on the pulls up garbage front. Continue to work about that.
We have. We're continuing to make significant improvements just with
the software. So the we're expecting to increase the parameter account. Actually,
at this point, I think we think we can probably
ten x that almost tax the Pepper prim Yeah, roughly
(05:48):
roughly ten x the prim account. So this is actually
a very tricky thing to do because if you as
you increase criameter account, you get you get to choke
on every bad wind. But we currently think we can
tenext the parameter count from what people are currently experiencing.
That's that's not just support x, actually ten x increasing
(06:11):
in primary account. And yeah, so that's still a lot
of improvement on the existing artware. To happen. Energy is
growing really well despite edwards from tariffs and spirit suppliettery
jodges the very package bagistic quickly and how we have
ouprates the bagapact that will make it even better. And
(06:32):
we had red board our powerful to plates and dat
H two. So where I think bat batteries are just
going to be a massive thing, that that the scale
of batteries battery deman is, I think not that many
people appreciate just how gigantic the scale of battery man is.
(06:54):
The way you think about it is that the US
sustained power output of the U S grid is on
one terrat wet, but average usage is less than half
at tarot wet. If you add batteries to the mix,
you can run the power plants twenty four to seven
ex soul capacity. That's double a more than doubling the
(07:15):
energy output per year of the United States just with batteries.
But that's again big deal. It's a really big deal.
Speaker 2 (07:22):
Optimist to what we're revolving the optimist design to really
get an optimist at the point where it is a
phenomenal design to pur an optimist Version two right now
sort of two and a half.
Speaker 1 (07:36):
Optimus three is if there is an exquisitive design, in
my opinion, there will be an credit. As I've said
many times before, I predicted it will be the biggest
product ever. It's a very hard problem to solve. You
have to design every part of its from physics first principles.
There's nothing that's both the shelf that actually works, so
(07:57):
you're going to design every motor, gear box, par electronics, control, electronics, sensors,
the mechanical elements also got to train optimists to use
its eleven sented sensors with a neural net, but will
be applying the same techniques that we're applied. Or a
(08:19):
car which is essentially a four wheel robot, and optimist
is a robot which arms and legs, so that the
same principles that apply to optimize eight AI inferns are
the car applied to optuments because they are both really
robots in different forms. And Tesla, it is wortant to
(08:40):
know the Tesla is my father best in the world
at real world AI. Like a clear proof point that
would be if you're compared to Tesla to way more
where it's got. You know the car is best doing
with god knows how many sensors, and yet doesn't Google
goodn't AI, Yes, but they're not good at real best
(09:00):
part the Tesla is actually much better than Google by
far and any much anyone at real world and and
I and by far, Chesla has the best infernce efficiency.
Like I think a key for AI is what is
the intelligence per gigabytes? People know about parameters blah blah blah,
(09:24):
but I think, but about toil about gigabytes because with
the parameters can have four bit parameters, eight bit parameters,
sixteen bit parameters. But the actual constraints in the hardware
are how many gig bytes run and how many gigabytes
per second he transfer from. Therefore, it is not the
parameters straight, there's it isn't there's a byte straight And
(09:48):
Tesla has the highest intelligence density of any ANI by far,
and I have a lot of insight this with with
x Ai. Xai is grock is the smartest of the
I over but it's a BROP four is a giant
beast there with the terrorbyte levels. And that's kind of
(10:08):
an important to know. It telles that has the best
collegian students intelligents density will be a very big deal
of the future. It is no so with Optimus three,
which is really the right design, it's like it doesn't
happen at this point, there's no there's a SIGNI think
claws would be Octimist three design, but we are going
(10:28):
to retool a bunch of things, so it's it'll probably
be prototypes of Optimist three end this year and then
scale production next year. We're going to try to scale
Optimist production as fast as it's seemingly possible to do. So,
try to get two million units a year as quickly
as possible. We think they give them less than five years.
That's I guess that that's a reasonable aspiration. Is the
(10:52):
million units a year five years? It seems like an
indieval tarry A conclusion. So four twenty White has been
breaks a year. A lot of major monstudents made clear
launch progress in autonomy that a lot of Masaya said
we would not achieve. It's worth noting that the we
(11:13):
have done what we said we're going to do, and
they were always on time. But we get it done
and there sitting there with egg on their face. So great, great,
great progress by it tells the team, Yeah, I do
think it tells continues to execute well with big autonomy
and and the humor or what autonomy, it will be
(11:37):
the most valuable helping in the world. There's a lot
of execution between year and there. It doesn't just happen,
it's about it excute very well. I think Tessant hasn't
shown it being most company in the world obviously, I'm
stremely optimistic about future of company. Best way to break
(11:57):
the futurist and make it happen maging. It happened here
with those team. So I'd just like to say thanks
to one of our supporters, and I think we're get
to tell you incredibly exciting future. I think it's it'll
be available for vice personal use by the end of
this year and certain geographies. We're just being very careful
about it and we're just being extremely partner. But I
(12:19):
would be a copy of that by the others year
within a number of cities in the US. We will,
we will, It'll be available to edge users like the Optimist.
Optimist three design, as I mentioned earlier, is I think
finally the right design. They'll wait for the optimizations, but
they're will. I think they're fundantal changes too that are needed.
Over the Optimist three design. It has all the degrees
(12:41):
breedom that you really want or need, so it's you.
I'll have bonotypes of that and three three months and
means suctually in production postimply start production on that next
year the production round. Simple stories conflict the s cove
(13:02):
up your production around. Something is got an entire when
everything is new, because the rate of production will move
this fast is the least lucky and least confident elements
of the entire supply chain, as well as internal processes.
So the more new STU that is a in a product,
the slower than the rank could be because of unexpected
(13:25):
supply chain interruptions or mistakes internally. Especially using it to
predict through the end of the of the escove or
or late in the S curve. In the beginning of
the ESCO and the beginning of the ESCO of the
production round is a case not really materials for revenue purposes.
The beginning of the ESCO, you're usually yours negative course, Morgan,
(13:50):
you're debuting amount of issues. So that's why it's it's
I feel like fairly confident predicting things. Well, please medium
confident in predicting where we are five years, but it's
hard ricked where we are in a year or two years.
So that's why I think five year us I think
(14:10):
it could be at the living with our be surprised
at the end of five years, sixty months from now
get three online roughly making one hundred thousand Optimus robots
in month in sixty months, I would be shocked, I think, obviously. Yeah,
we're a publicly treaded company. Sure Holders are welcome to
put forward any shareholder proposals that they'd like. I recently
(14:34):
encouraged that, and then have shareholders vote and we'll act
in accordance with the shareholder wishes. Great. Yeah, there's a
lot of exciting things happening in the science studio. It's
like static and really what's going to happen over the
next to several years is a fundamental transformation of the
company from the pre asolomy world to a post autonomy
(14:56):
and I'm working on a near master plan to articulately
that tells the team and you know there will there
will be there are some teething pains as you as
you transition from pre pre aufineries vote autonomy world, but
I think that the future tells it is incredibly exciting
and will profoundly change the world in a good way.
(15:17):
Please sound like fiable whatever, but I think we'll let's
just say if we execute on that plan effectively, which
is you have to actually do that because it will
be the most vaguable company in the world by far.
Are two. We expect to have Dojo to operating scale
SubTime next year, with scale being somewhere around one hundred
(15:40):
k a twelve hundred equivalents. And AI five, which is
really it's spectacular use those words likely spectacular to the
A five Childen being hope hopefully be buying production around
the end of next year, but that has a lot
of potential. I think, you know, thinking about DOJOH three
(16:03):
and the AI six birdst chip, it seems like, intuitively
we want to try to find conbudgets there where it's
basically the same chip that is used what were you used,
say two of them in a car or an optimist
and maybe a large a number on board account of
(16:28):
by twelve and AB or something like that. If you
want highbands communication between the chips out of course serving
they're drink it bird serving. That sort of seems like
intuitively the sansful way to go.
Speaker 3 (16:42):
Yeah, well, the the cybercap is which is really optimized
for autonomy that that I think has got probably somebody's
sent per mile potential over time by betweenty five.
Speaker 1 (16:58):
You know, it's really like if you design a car
from scratched to be a cost optimized robotic taxi like
Cybercap like for example, we're not trying to make it,
you know, corner like a look incredibly well like a
Model three would, or modeless or even mall way like
(17:20):
it's got moddel all about all about. Cars that are
driven by people are super fun drive. They've got incredible acceleration,
you know, the incredible cornering capability. But not ben We're
confident that very few people and a cybercap want to
be totaling around. So you know, so we were we
(17:41):
produce the top end speed, which means we can use
more efficient tires, we don't use as much acceleration, we
don't use as much you know, big big brake system.
We won't want stopping distance, but we're not expecting it
to be heavily used. If it's it's a gentle right.
Essentially you designed for a ride and then you you
(18:03):
have am that much optimized optimized asign point. So that
that's why it timpts probably achieve that, especially optim optimuses,
you know, so asserving you cleaning up the car and
make doing maintenance and stuff. It's during automatic charging, so
I think it's gonna actual cost for a mile of
(18:26):
cyber cab will be very low. The cost for a
while our existing I fee will be higher but still
very competing. So breviously number of fifty sets and discussing, Yeah,
this really fills road taxically. It will build from tiny
to gigantic in terms of operations in pretty short period
(18:47):
of time. Like my guess is there. It has material
impact on our financials around the end of next year.
Yeah that that that is a major CONSERVETI as I
mentioned in the past, and I hope that is addressed
to the upcoming shareholders meeting. But yeah, this is a
big deal. Like you know, I want to find that
(19:08):
I've bet like so a little control that I can
usually be ousted by active as shareholders after having bolts
you know these all if you're going robots. I think
that as I mentioned before, I think my control of
Hotaila should be enough to ensure that it goes in
a good direction, but not so much control that I
can't be thrown out if I go crazy. Yes, it's
(19:30):
a bit of a tough thing because like when we
do an aid. We find that some was literally done
it pray by bray examination of our slightes everything we say,
and they're coveyos. So you know of say like, well,
what's the traidal. It does help with's recruiting, but then
competitors look break clessly in coveyots. I mean that said
(19:53):
we should probably I mean, I guess we could consider
the shareholder beginning to be sort of an A can
do if we can maybe go to depth, some solid
amount of depth at the ill share Older meeting with
respect to Optimus and a AI and that sort of
our chip chip stuff. Perhaps, Yeah, the test also really
(20:14):
underrated in terms of a chip design as well as
AI software. So like there's a whole lot that chip
that we were that that exists that we will prefer
to bringing our car, that is, ANYI chip that then
we would prefer to putting the car over our own.
And even though it's been now out for several years
and recognan that the air bike chip will be for
(20:37):
a Vale game tream, in fact, it's it's so powerfully
we'll have to interurb it or sun grief or pockets
outside of the rings because it flows weary past the
export restrictions. So let'sten export restrictions change. We actually will
have to know if our ai bib ship is look
kind of where hopefully list hopefully we keep raising the
(20:58):
bar and the other way the kids bit silly. We'll
have a bunch of Optimist robots on stage at the
shareholder meeting. The Optimist lab is cool to see. It
looks like it basically looks like the state of Westwold
get robots in various stages, some of them, you know,
in various stages of prepared, like I don't know. It's
(21:20):
some combination of like the tattooing Chukyard and the and
West Wolds. It's pretty cool. An octopus is walking around
the office here in polls there are twenty four seven
is just walking around like it's small, I think. And
we're so optimistic, you know the t A diners that
(21:42):
we're bub born. Yeah, so it's we'll go for a
world where robots are rare to where they're so common
that you don't even look up. Okay, we haven't really
thought hard about that. We need to make sure it
works when the bagels are pull the under our control
and it's kind of one step at a time. Here
we don't want to jump the gun. As I said,
(22:03):
we're doing parallered about safety. So so it's just like
but I guess, I guess like next year, I'd say
competently next year. I'm not sure when next year, but
competently next year. People who will be able to add
to subtract their part to the Testla fleet. I'm just
like at their bab or ing prossible well as seem
(22:25):
as there is a clear cash blow stream associated with
any any product that you can get finance it. Most
people still done there the master yard. People don't know
it exists, and it's still like half of Tesla owners
who could use it, I haven't tried it even once
this They don't actually rely as this is something we
(22:47):
want to educate them on. So we've got to. When
they come in for service, we'll reach out to them,
send them like videos of how to make the work
and most much it's it's such a shocking thing. They
don't think a car is capable of this. So you
have Jack, you have just show them and get them comfortable.
Alternally don't know if it's so trivial, but it's it's
(23:10):
like saying you've got a Canada sing a dance, but
it just looked like an old cat. They're like, you know,
until you see the cat sing a dance and talk
like your series just a cat, that's that's tes I see.
Our car is intelligence. Well, we're in this like weird
transition period where we will lose a lot of intatus
(23:30):
in the US. Incentives actually ready other parts of the world,
but roll is them in the US, and it will
still have at the relative early stages of autonomy. On
the other autonomy is most advanced and most available from
a regulatory standpoint in the US. So I mean, does
that mean like we could have a few rough quarters. Yeah,
(23:50):
we probably could have a few rough quarters. And I'm
not saying the world, but we could Q you know,
Q four, Q one, maybe Q two. But once you
get to autonomy at scale in the second hand of
half of next year, certainly by the end of next year,
I think the I'm surprised to the economics, I'm not
very simpellent because it's just a model. Why they look
(24:14):
the cat out of the back there dancing cut that
consilingly bess, But it can be. It can talk and think.
But that's the cool part yeah, I mean the fundamentally,
the biggest obstacle remains that people just don't happen. Some
of the people don't know that. The desire to buy
the car is very high. Just people don't have enough
money in their bad account to buy it. Literally, that
is the issue, not a lack of desire, but lack body.
(24:38):
So the more affordable we can make the car, the
better I think it's gonna be. It will be a
very big deal when people can release their car to
the fleet and have it earned money for them, which,
like I said, i'm big. I feel confident saying that
will happen next year in the US at least, and
in the US where we're legally allowed, you appropriate displayers
(25:00):
and they'll make the affordability dramatically greater, just like you know,
if you have Airbnb and you ran out your home
when you're not there, or ran out of a guest
room or guest house or something like that, you're the
affordabilit if your home is much greater. Well, they are
doing different things here, so you know, the Xai is
doing like you know, terabyte scale models and the multi
(25:25):
terrabyt scale models. Tesla's one hundred times smaller models. So
once real walth AI and once kind of I guess
art official superintelligence tech thing. The I mean really kind
of the genesis for XAI was that there were there
were certain people who some people would not join Tesla
AI engineers because they wanted to look at ASI. They
(25:48):
would join Tesla, and I was like, well, maybe they'll
join a new company. And now I think the Tesli
bomb is extremely important, but not everyone agrees with me
on that. And so rather than have them join and
you know, open a eye or people or or some
other company, it's like, as well have them create a
company in that regard, which like say, I so that's
(26:11):
you know, and you know, pilin make it. Make a
decision do they want to work on It's like super
intelligence data center or real wealthy idens good. They're both
compelling problems, but some people work want to open one
and some one of them the other
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
New Heights with Jason & Travis Kelce
Football’s funniest family duo — Jason Kelce of the Philadelphia Eagles and Travis Kelce of the Kansas City Chiefs — team up to provide next-level access to life in the league as it unfolds. The two brothers and Super Bowl champions drop weekly insights about the weekly slate of games and share their INSIDE perspectives on trending NFL news and sports headlines. They also endlessly rag on each other as brothers do, chat the latest in pop culture and welcome some very popular and well-known friends to chat with them. Check out new episodes every Wednesday. Follow New Heights on the Wondery App, YouTube or wherever you get your podcasts. You can listen to new episodes early and ad-free, and get exclusive content on Wondery+. Join Wondery+ in the Wondery App, Apple Podcasts or Spotify. And join our new membership for a unique fan experience by going to the New Heights YouTube channel now!