Latest / Elon Musk Podcast / Tesla Q4 and full year 2024 Financial Results and Q&A Webcast

Transcript

Tesla Q4 and full year 2024 Financial Results and Q&A Webcast · Elon Musk Podcast

0:01
Hey, everybody. Welcome back to the Elon Musk
0:04
Podcast. This is a show where we discuss
0:07
the critical crossroads, the Shape, SpaceX, Tesla X, The
0:11
Boring Company and Neurolink. I'm your host, Will Walden.
0:18
Good afternoon, everyone and welcome to Tesla's fourth
0:20
quarter 2024 Q&A webcast. My name is Travis Axelrod, the
0:24
Head of Investor Relations here at Tesla and I'm joined today by
0:28
Elon Musk and Bebop, Tunisia and a number of other executives.
0:32
Our Q4 results were announced at about 3:00 PM Central Time in
0:36
the update deck we published at the same link as this webcast.
0:40
During this call, we will discuss our business outlook and
0:43
make forward-looking statements. These comments are based on our
0:46
predictions and expectations as of today.
0:49
Actual events or results could differ materially due to a
0:52
number of risks and uncertainties, including those
0:54
mentioned in our most recent filings with the SEC.
0:58
During the question and answer portion of today's call, please
1:00
limit yourself to one question and one follow up.
1:03
Please use the raise hand button to join the question queue.
1:07
Before we jump into Q&A, Elon has some opening remarks.
1:10
Elon, thank you. So in summary, in Q4 we set a
1:14
record and limit vehicles at an annualized rate of only two
1:17
million a year. So congratulations to the Tesla
1:21
team on excellent work. We're achieving record
1:24
production and deliveries. We Model Y was the best selling
1:29
vehicle of any kind for 2024. That's worth noting that the not
1:33
just the best electric vehicle, the best vehicle of any kind on
1:36
Earth, number one was Model Y. We're staying focused on on
1:41
maximizing volumes and obviously doubling down for I don't know
1:45
what it's really, I was going to say doubling down on autonomy,
1:50
but really it's like autonomy is like 10 Xing.
1:54
Frankly, doubling is not even enough.
1:57
We made many critical investments in 2024 in
2:00
manufacturing, AI and robotics that will bear immense fruit in
2:05
the future. Immense.
2:06
Like it's in fact to such a scale that it is difficult to
2:10
comprehend. And I've said this before and
2:14
I'll stand by it. I see a path, but I'm not saying
2:18
it's an easy path, but I see a path for Tesla being the most
2:21
valuable company in the world, world by far, Not even close.
2:24
Like maybe several times more than I mean, there is a path
2:31
where Tesla is worth more than the next top five companies
2:34
combined. There's a path to that.
2:38
I mean, I, I think it's like an incredibly, it's like a
2:42
difficult path, but it is an achievable path.
2:47
So, and that is overwhelmingly due to autonomous vehicles and
2:54
autonomous humanoid robots. So our focus is, is actually
3:01
building towards that. And then that that's what we're
3:06
laying the ground. We will lay the groundwork for
3:07
that in 2024. We'll continue to lay the
3:09
groundwork for that in 2025. In fact, more than lay the
3:13
groundwork actually be building the structure, be we're building
3:17
the manufacturing lines. And like, yeah, I'd like setting
3:26
up for what I think will be an epic 2026 and a ridiculous 27
3:33
and 28. Ridiculously good.
3:37
That is my prediction. You know, we as you have very
3:45
few, very few people understand the value of full self driving
3:49
and our ability to monetize the fleet.
3:53
You know, I've some of these things I've said for quite a
3:56
long time and I know people have said, well, you know, he owns
3:58
the boy. You cried wolf like several
3:59
times, but I'm telling you there's a damn wolf at this time
4:04
and you can drive it. In fact, it can drive you.
4:06
It's a self driving wolf. You know, for a lot of people
4:12
that they're like their experience of Tesla autonomy is
4:15
like if it's even a year old, if it's even 2 years old, that it's
4:20
like me. It's like meeting someone when
4:21
they're like a toddler and thinking that they're going to
4:24
be a toddler forever, but obviously not going to be a
4:26
toddler forever if they grow up. But if their last experience was
4:31
like, oh, FSD was a toddler, it's like, well, it's grown up
4:34
now. Have you seen it?
4:34
It's like walks and talks, and that's really what we've got.
4:41
And it's difficult to for people to understand this because human
4:44
intuition is linear as opposed to what we're seeing is
4:48
exponential progress. So that's why my number one
4:54
recommendation for anyone who doubts is simply try it.
4:58
Have you tried it? When's the last time you tried
4:59
it? And the only people who are
5:02
skeptical, the only people who are skeptical are those who have
5:05
not tried it. So, you know, a car goes, a
5:12
passenger car typically has only about 10 hours of utility per
5:15
week out of 168, a very small percentage.
5:20
Once that car is autonomous, my rough estimate is that it is it
5:25
is in use for at least a third of the hours per week.
5:28
So call it 50, maybe 35 hours of the week.
5:32
And it can be used for both cargo delivery and people's
5:36
delivery. So even let's say people are
5:38
asleep, but you can deliver packages in the middle of the
5:42
night or resupply restaurants or whatever the case may be,
5:45
whatever people need at all hours of the day or night.
5:51
That same asset, the thing that that is these things that
5:54
already exist with no incremental cost change, just a
5:58
software update, now have five times or more the utility that
6:03
they currently have. I, I, I think this will be the,
6:09
the largest asset value increase in human history.
6:14
Maybe there's something bigger, but I'm, I just don't know what
6:16
it is. And so people who look in the
6:19
rearview mirror are looking for past precedent, except I, I
6:24
don't think there is one. So, but you know, there's there.
6:32
But like, the reality of autonomy is upon us.
6:35
And I repeat my advice, try driving the car or let it drive
6:41
you. So now it works very well in the
6:47
US, but of course, it will, over time, work just as well
6:51
everywhere else. Yeah.
6:59
So we're working hard to grow our annual volumes, our
7:06
constraint this year, our current constraint is battery
7:09
packs this year, but we're working on addressing that
7:12
constraint and I think we will make progress in addressing that
7:17
constraint. And then things are really going
7:21
to go ballistic next year and really ballistic in 27 and 28.
7:27
So, yeah, so a bit more on full self driving.
7:36
Our Q4 vehicle safety report shows continued year over year
7:39
improvement in safety for vehicles so that the safety
7:42
numbers, if somebody has supervised full self driving
7:46
turn on or not the the safety differences are gigantic.
7:51
So and people have seen the the immense improvement with version
7:57
13 and with incremental versions in version 13 and then version
8:02
14 is going to be yet another step beyond that.
8:05
That is very significant. We launched the Cortex training
8:09
cluster at Bigger Factory Austin, which was a significant
8:12
contributor to FSD advancement and we continue to invest in
8:16
training infrastructure out of Texas headquarters.
8:21
So the, the, the training needs for Optimist or Optimist
8:25
humanoid robot are are probably at least ultimately 10X what
8:30
what's needed for the car at at least to get to the full range
8:34
of of of useful roles. You can say like how many
8:38
different roles are there for a humanoid robot versus a car
8:43
Humanoid robot is has probably what 1000 times more uses and,
8:51
and more complex things than in a car.
8:54
That that doesn't mean that the training scales by 1000, but
8:59
it's probably, you know, 10X now you can do this progressively.
9:03
So it doesn't mean like hotels are going to spend like $500
9:05
billion in training compute because we would obviously train
9:10
the optimists to do enough tasks to match the output of Optimus
9:17
robots. But and obviously the cost of
9:22
training is dropping dramatically with time.
9:25
So, but it is it is like it is one of those things where I
9:31
think long term optimists will be optimist has the potential to
9:35
be north of $10 trillion in revenue.
9:37
Like it's really bananas. So that you can obviously afford
9:43
a lot of training compute in that situation.
9:45
In fact, even $500 billion of training compute in that
9:48
situation would be quite a good deal.
9:53
Yeah, the future is going to be incredibly different from the
9:59
past, that's for sure. We live at this unbelievable
10:03
infection point in human history.
10:06
So yeah, so in the proofs, the proof is in the pudding.
10:14
So we're going to be launching unsupervised full Self driving
10:20
as a paid service in Austin in June.
10:25
So if I've talked to the team, we feel confident in being able
10:32
to do an initial launch of unsupervised no one in the car
10:37
Full Self driving in Austin IN June.
10:40
We already have Tesla's operating autonomously
10:44
unsupervised for self driving at our factory in Fremont and will
10:48
soon be doing that at our factory in Texas.
10:52
So thousands of cars every day are driving with no one in them
10:57
at our Fremont factory in in California.
11:01
They will soon be doing that in Austin and then elsewhere in the
11:05
world at the rest of our factories, which is pretty cool.
11:10
And the cars aren't just driving to exactly the same spot because
11:15
obviously they're all when it's light at the same spot.
11:20
The, the cars are actually programmed with, with where,
11:23
with what lane they need to park into to be picked up for
11:27
delivery. So they, they, they drive from
11:30
the factory end of line to their, to their destination
11:35
parking spot and, and that, and that could be picked up for
11:41
delivery to customers. And then doing this reliably
11:45
every day, thousands of times a day.
11:50
It's pretty cool. Like I said, these Teslas will
11:53
be in the wild with no one in them in June in Austin.
12:00
So what I'm saying is this is not some far off mythical
12:04
situation. It's literally, you know, 5-6
12:10
months away, five months away sort of thing.
12:15
And while we're stepping into putting our toe in the water,
12:19
gently at first, just to make sure everything's cool, our
12:24
solution, our sort of our solution, our solution is a
12:28
generalized AI solution. It is not, does not require high
12:31
precision maps of a locality. So we just want to be cautious.
12:38
It's not that it doesn't work beyond awesome.
12:40
In fact it does. We just want to be put a toe in
12:43
the water, make sure everything is OK, then, you know, put a few
12:47
more toes in the water, then put a foot in the water with safety
12:52
of the general public as and those in the car as our top
12:56
priority. With regard to optimist,
13:04
obviously I'm making you know these revenue predictions,
13:07
predictions that sound absolutely insane.
13:09
I realized that, but but they are, I think they will prove to
13:15
be accurate. Yeah, yeah.
13:23
Now with Optimus, there's there's a lot of uncertainty on
13:26
on the exact timing because it's not like a train arriving at the
13:31
station. For Optimus.
13:33
We are designing the train and the station and in real time
13:37
while also bullying the tracks. And, and so they're like, why
13:42
did people just like, why didn't the train arrive exactly at
13:45
12:05? And like we're literally
13:46
designing the train and the tracking station in real time
13:49
while you're seeing like, how can we predict this thing with
13:51
absolute precision? It's impossible.
13:54
You know, the, the, the normal internal plan calls for roughly
13:58
10,000 optimist robots to be built this year.
14:00
Will we succeed in building 10,000 exactly by the end of
14:03
December this year? Probably not.
14:06
We we but I but will we succeed in making several thousand Yes,
14:10
I think we will will those several 1000 optimists robust be
14:14
doing useful things by the end of the year?
14:16
Yes, I'm confident they will they will do useful things that
14:22
those the optimist in use at the Tesla factories for production
14:28
design one will inform how would we change for production sign 2,
14:33
which we expect to launch next year.
14:36
And our goal is to ramp prop Optimus production faster than
14:41
maybe anything's ever been ramped meaning like
14:45
aspirationally an order of magnitude ramp per year.
14:50
Now if we aspire to an order of magnitude ramp per year, perhaps
14:53
we only end up with a half order of magnitude per year, but
14:57
that's the kind of growth that we're talking about.
15:01
It doesn't take very many years before, you know, we're, we're
15:05
making 100 million of these things a year.
15:10
If you go up by let's say a factor, you know, by 5X per
15:14
year, insane. Not 50%, five, 100%.
15:24
So, you know, these are big growth numbers.
15:31
Yeah, but we do need to be. This is an entirely new supply
15:34
chain, is entirely new technology.
15:37
There's nothing off the shelf to use.
15:39
We tried desperately with Optimist to to use any existing
15:43
motors act, you know, any actuators, sensors, nothing.
15:48
Nothing worked for a humanoid robot at any price.
15:52
We had to design everything from from physics first principles to
15:56
work for humanoid robot and with the most sophisticated hand that
16:02
has ever been made before by far right Optimus will be also like
16:08
play the piano and be able to thread a needle.
16:18
I mean, this is the level of precision no one has been able
16:21
to achieve. And so it's really something
16:26
special so. Yeah, so in my prediction long
16:38
term is the optimist will be overwhelmingly the value of the
16:40
company regarding energy backed up back to us.
16:48
It was the Elon can you come back here for a minute?
16:53
OK, back to US. Energy storage is a big deal and
16:58
will become it's already super important, will become
17:02
incredibly important in the future.
17:04
And and it is something that enables far greater energy
17:07
output to the grid than is currently possible.
17:12
Because the the grid, the the grids are, are the vast majority
17:17
of the grid has no energy storage capability.
17:20
So they have to design the power plants to, you know, for very
17:23
high peaks and assuming that there's no energy storage, once
17:28
you have grid, grid energy storage and home based energy
17:30
storage, the actual total energy output per year of the grid is
17:34
dramatically greater than people think.
17:36
Maybe it's at least double. This will drive the demand of
17:41
stationary battery packs and especially the grid scale ones
17:46
to insane. Basically as much demand as we
17:49
could possibly make. So we have our second factory
17:57
which is in Shanghai that's starting operation and we're
18:00
building a third factory. So we're trying to ramp output
18:03
of the stationary battery storage as quickly as possible.
18:08
Now there, there is a challenge here where we have to be careful
18:11
to that we're not robbing take robbing from one pocket to take
18:14
to another pocket. Because for a given GW hours per
18:18
year of of the cell output, where if still does it go into
18:21
stationary applications or mobile applications, it can't go
18:24
both into both. So we have to make that trade
18:27
off yeah. But overall the demand for total
18:39
GW hours of batteries whether mobile or stationary that will
18:43
grow to in a very, very big way over time.
18:51
So in conclusion, 2025 really is a pivotal year for Tesla and I.
18:59
And when people look back on 2025 and the launch of
19:02
unsupervised full self driving, true real world AI that that
19:07
actually works, I think they may regard it as the biggest year in
19:11
Tesla history, maybe even bigger than our first car, the Roadster
19:18
or or the model ISO, the Model 3 or Model 1.
19:23
In fact, I think it probably will be viewed 25 as maybe the
19:27
most important year in Tesla's history.
19:31
There is no company in the world that is as good at real world AI
19:34
as Tesla. I don't even know if he's in 2nd
19:38
place like you say. Like who's in 2nd place for real
19:42
world? AII would need a very big
19:44
telescope to see them. That's how Fall behind they are,
19:50
all right. Great.
19:52
Thank you very much, Elon. And Vaibhav has some updating
19:54
remarks as well. Yeah, I'll talk about things on
19:56
Earth. As Elon mentioned in Q4, we set
20:01
records and vehicle deliveries and energy storage deployments
20:04
in an uncertain macro environment.
20:06
We were able to grow auto and energy storage volumes both
20:09
sequentially and on a year on year basis.
20:12
For this, I would like to thank the efforts of everyone at Tesla
20:15
to make this a reality and our customers who helped us achieve
20:19
this speed. Coming into the fourth quarter,
20:22
our focus was to reduce inventory levels in the
20:25
automotive business and we accomplished that by ending the
20:28
quarter with the lowest finish quote inventory in the last two
20:31
years. This was a result of offering
20:34
not only attractive financing options, but also other
20:37
discounts and programs which impacted ASPs.
20:40
While we saw volume growth in almost all regions that we
20:43
operate in, we hit a new record of for deliveries in the Greater
20:47
China market. This is an encouraging trend
20:50
since we grew volume in a highly competitive Bev market.
20:55
On the automotive margin front, we saw a quarter over quarter
20:58
decline primarily due to lower ASPs and due to the recognition
21:03
of FSD related revenue in Q3 from feature releases.
21:07
Our journey on cost reduction continues and we were able to
21:10
get our overall cost per car down below $35,000, driven
21:15
primarily by material costs. This was despite increased
21:19
depreciation and other costs as we prepare for the transition to
21:23
the new Model buy for which we recently started taking orders
21:26
in all markets. All of that case will start
21:29
producing the new Model Y next month.
21:32
While we feel confident in our team's abilities to ramp
21:36
production quickly, note that it is an unprecedented change and
21:39
we are not aware of anybody else taking the best selling car on
21:44
the planet and updating all factories at the same time.
21:49
This changeover will result in several weeks of last production
21:53
in the quarter. As a result, margins will be
21:56
impacted due to ID capacity and other land related costs as is
22:00
common in any launch, but will be overcome as production
22:03
present. We will be introducing several
22:06
new products throughout 2025. We are still on track to launch
22:10
a more affordable model in the first half of 2025 and we'll
22:13
continue to expand our lineup from there.
22:17
On a dollar for dollar basis, we believe we have the most
22:20
compelling lineup today compared to the industry and it will
22:23
continue to get better from there.
22:26
As always, all our products come with the best software in the
22:29
industry, autonomy features and capable of full autonomy in the
22:33
future. And despite the premium
22:35
experience, the total cost of ownership is close to mass
22:38
market less premium competitors. Energy storage deployments
22:43
reached an all time high in Q4 and this and resulted in but
22:49
declined sequentially. This was a result of higher,
22:54
this sorry growth came from Megapack and Power.
22:59
Both businesses continue to be supply constrained and like Ilan
23:02
mentioned, we're trying to ramp up production with mega Factory
23:07
Shanghai coming online this quarter onwards.
23:11
While quarterly deployments will likely continue to fluctuate
23:14
sequentially, we expect at least 50% growth in deployments year
23:18
over year in 2025. Gross profit and margins in the
23:22
service and other business was up year over year but declined
23:26
sequentially. This was the result of higher
23:29
service center costs and lower profit from used car business.
23:33
The businesses with this there was another primarily support
23:36
our new car business especially through their impact on total
23:39
cost of ownership. Therefore, while they manage
23:41
them to be positive on a GAAP business, we do not expect
23:45
similar margins as the rest of the business.
23:48
There's a lot of uncertainty around tariffs.
23:50
Over the years, we've tried to localize our supply chain in
23:54
every market, but we are still really reliant on parts from
23:57
across the world for all our businesses.
23:59
Therefore, the imposition of tariffs, which is very likely
24:04
and any of the supercosity will have an impact on our business
24:08
and profitability. Our operating expenses grew both
24:12
year over year and sequentially. The biggest driver of the
24:15
increase was R&D as we continue to invest in AI related
24:19
initiatives. The remaining increase came from
24:23
growth in our sales capabilities and marketing efforts from
24:26
preferred program. For 2025, we expect operating
24:30
expenses to increase to support our growth initiatives.
24:34
It is important to point out that the net income in Q4 was
24:37
impacted by a 600 million mark to market benefit from Bitcoin
24:42
due to the adoption of a new accounting standard for digital
24:45
assets whereby we change we will take mark to market adjustments
24:49
to other income every reporting period going forward.
24:54
Our free cash flow for the quarter was 2 billion and
24:57
despite CapEx increase of 2.4 billion in 2024, we were able to
25:01
generate free cash flow of 3.6 billion for the year.
25:05
CapEx efficiency is something we are extremely focused on.
25:10
While we have invested in AI related initiatives, we've done
25:13
so in a very targeted manner to to to utilize the spend to get
25:18
immediate benefits. The build out of Cortex was
25:22
exceeded because of the rule ex actually how explored the rule
25:28
out of FSD V version 13, a cumulative AI related CapEx
25:33
including infrastructure so far has been approximately 5
25:36
billion. And for 2025, we expect our
25:39
CapEx to be flat on a year over this year basis.
25:43
In conclusion, you know like Elon said, 2025 is going to be a
25:48
pivotal year for Tesla. There are a lot of investments
25:51
which we've made and we'll continue to make in this coming
25:54
year, which will set the pace for the next phase of growth.
25:59
And you know, it is something which now I'm getting out, out
26:03
of work. It is going to be out of this
26:05
world and we just are playing, putting the bank foundation and
26:11
that's all I have. Great.
26:13
Thank you very much, Vaibhav. Now we will move over to
26:16
investor questions. Maybe we'll start with say.com.
26:19
The first question is, is unsupervised FSD still plan to
26:23
be released in Texas and California this year?
26:25
What hurdles still exist to make that happen?
26:29
You addressed the Texas piece, I think already so.
26:31
Yeah, I'm confident that we will release unsupervised FSD in
26:35
California this year as well, you know, yeah, in fact, I think
26:42
we will at most likely release unsupervised FSD in many regions
26:49
of the country of the US by the end of this year.
26:53
Like I said, we're, we're just putting a toe in the water, then
26:56
few toes, then the foot, then leg, then, you know, make sure
26:59
everything's cool. And we're, we're looking for a
27:05
safety level that is significantly above the average
27:08
human driver. So it's not anywhere like much
27:11
safer, not like a little bit safer than human, way safer than
27:13
human. So the standard has to be very
27:16
high because the moment there's any kind of, you know, accident
27:21
with an autonomous car, this immediately gets worldwide
27:27
headlines. Even though about 40,000 people
27:30
die every year in car accidents in the US and that most of them
27:35
don't even get a mention anywhere.
27:38
But if somebody's creates a shin with an autonomous car, it's
27:41
headline news. We wouldn't want that.
27:44
Yeah, so it's really from an ex. The only thing holding us back
27:49
is an excess of caution, but people can certainly get a feel
27:53
for how well the car would perform as unsupervised FSD by
27:57
simply having a car, allowing a car to drive you around your
28:00
city. And see, how many times do you
28:03
have to intervene? Not where you wanted to
28:05
intervene or we're a little concerned, but how many times do
28:08
you have to intervene for, for definite safety reasons.
28:12
And you will find that that is currently very rare and over
28:16
time almost never great. Thank you very much.
28:22
The next question is, are there any discussions with other auto
28:25
companies about licensing FSG? Yes.
28:33
What we're saying is at this point significant interest from
28:38
a number of major car companies about licensing Tesla full soft
28:43
driving technology. What we've generally said is the
28:47
best way to know what to do is take one of our cars apart and,
28:54
and then you can see where the placement of the cameras are,
28:57
what the thermal needs are of the Tesla AI furnace computer.
29:02
That's better than us, you know, sending some CAD drawings and,
29:07
and then we're only going to entertain situations where the
29:10
volume would be very high. Otherwise it's not worth the
29:12
complexity and and we will not broaden our engineering team
29:18
with laborious discussions with other engineering teams until we
29:23
obviously have unsupervisable self driving working throughout
29:26
the United States. I think the interest level from
29:31
other manufacturers to license FSD will be extremely high once
29:35
it is obvious that unless you have FSD, you're dead.
29:42
Yeah, great. Thank you very much.
29:46
The next question is, is Optimist now mostly design
29:49
locked for 2025 production? Optimist is not design locked.
29:57
So when I say like we're, we're designing the train as it's
30:02
going to, we're, we're redesigning the train as it's
30:04
going down the tracks while redesigning the tracks and the
30:06
train stations. Yeah, it's, it's, it's rapidly
30:11
evolving. It's rapidly evolving in a good
30:13
direction. You know, it's, it's pretty,
30:17
pretty damn amazing actually. Teams doing a fantastic job.
30:20
We really have by far, I think by far the best team of humanoid
30:25
robotics engineers in the world. And, and we also have all the
30:30
other ingredients necessary, you know, because you need a, you
30:33
know, great battery pack, you need great power electronics,
30:36
you need great charging capability, you need, you know,
30:40
great communications, going to have Wi-Fi and cellular
30:43
connectivity. And, and of course, you need
30:48
real world AI and then the ability to scale that production
30:54
to significant to huge levels. So you have to design for
30:58
manufacturing the things that I mean really what what other
31:02
companies are missing is they're missing the real world AI and
31:05
they're missing the ability to scale manufacturing to millions
31:11
of units a year. I think that is an
31:14
underappreciated. I think that industrialization
31:17
and design is a whole different thing than making a design.
31:21
Yeah, prototypes are trivial. Basically, prototypes are easy,
31:25
Production is hard. I've said that for many years.
31:28
The problem is that there's like those who've never been involved
31:31
in production or manufacturing somehow think that once you come
31:35
up with some Eureka design that you magically can make a million
31:38
units a year. And this is totally false.
31:41
The there needs to be some, there's some Hollywood story or
31:44
where they show actually, the problem is manufacturing.
31:48
I never even heard of one is is, you know, it's just doesn't fit
31:51
the narrative. The Hollywood thing is like,
31:53
it's like some lone inventor in a garage goes Eureka and
31:57
suddenly it files a patent and suddenly there's millions of
32:00
units. I like, I'm listening to guys.
32:02
We're we're missing really 99% of the story, 1% is any of the
32:08
old saying one like a product is 1% inspiration, 99%
32:12
perspiration. The Hollywood church is 1%
32:15
inspiration, and mine is, which forgets about the 99%
32:18
perspiration of actually figuring out how to make that
32:21
initial prototype manufacturable and then manufacture at high
32:24
volume such that the product is reliable, low cost, consistent,
32:28
doesn't break down all the time, and that is 100 times hotter at
32:32
least than the prototype. Then you have to get it there,
32:35
deliver it. Yeah.
32:36
Yeah, you're going to read all these regulations and there's a
32:39
million regulators around the world.
32:40
It's pretty difficult. Great.
32:42
Thank you. The next question is also
32:45
optimist related. When will Tesla start selling
32:48
optimists? I mean mobile price pay.
32:53
Well, the IT may for this year, we expect to just close the loop
32:58
with optimists being used internally at Tesla because we
33:03
obviously can easily use several 1000 humanoid robots at the
33:08
Tesla for, you know, like the most boring, annoying tasks in
33:13
the factory. Like the task nobody wants to do
33:16
where we have to like beg people to do this task.
33:19
And then they that it's like like the robots totally happy to
33:24
do the boring, dangerous, repetitive task that no, no
33:26
humans want to do. And then that's also actually
33:29
some of the easiest use cases for us to, you know, have, have
33:33
optimists do things like, you know, like load, load the the
33:39
hopper, like like for, you know, like I say, lit in body line.
33:42
If you like, transporting, you know, pieces of sheet metal to
33:45
the, the, the, the robot, which is really robot, the robot
33:49
welding line for the body. And you just have to they
33:52
non-stop take things out of out of a from 1 fixture to another
33:58
fixture. And it's a very boring job.
34:02
That's the kind of thing what the optimists could do.
34:04
I'm a guy who runs around and comes all the wealth stubs and
34:08
the independence. Yeah, there's, there's a ton of
34:11
boring jobs, tedious jobs, dangerous, you know, slightly
34:15
dangerous jobs that that's perfect botiness.
34:18
So we expect to use octopus for those tasks at our factories and
34:22
and that'll help us close loop for improvement This year.
34:26
It's really with the production version 2, which I think
34:32
launches sometime next year. I'd like to be the beginning of
34:36
next year, but maybe it, it's more like the middle of next
34:39
year. And and then we have to just
34:43
with a, with a production line that is designed for, you know,
34:49
on the order of 10,000 units a month versus 1000 units a month.
34:54
So like when you design A unit for, like when you design A
34:57
production line for 1000 units a month, it takes you a while to
35:00
actually reach anywhere close to 1000 years a month.
35:04
When you design and before any given product production output,
35:09
it takes you a while to actually reach its potential.
35:12
But the current line that we're designing is for roughly 1000
35:16
years a month of of Optimist robots.
35:20
The next line would be for 10,000 years a month.
35:24
The line after that would be for 100,000 years a month.
35:30
And I think probably with product with version two, that
35:35
is a very rough guess because there's so much uncertainty
35:38
here. Very rough guess that we start
35:40
delivering optimist robots to companies that are outside of
35:45
Tesla in the maybe the second-half of next year,
35:47
something like that. But like, so this is such an
35:50
exponential round that it'll go from no one's receiving humanoid
35:54
robots to these things like coming out like crazy.
35:58
We can't build enough. Well, we're always going to be
36:00
in the, we can't build enough situation.
36:05
Demand will not be a problem even at a high price.
36:08
And, and then as I said, like once we start, once we're at a
36:11
steady state of above 1,000,000 units a year, I, I think the
36:16
production, you know, I'm confident at 1,000,000 units a
36:18
year that the production cost of optimists will be less than
36:23
$20,000. You compare the complexity of
36:26
optimists to the complexity of of a car.
36:30
So just the total mass and complexity of Optimus is is is
36:36
much less than a car. So I would expect that at
36:41
similar volumes to say the Model Y, which is the over a million
36:44
years here that you'd see Optimus be, I don't know, half
36:47
the for half the cost or something like that.
36:51
What the, what the price of Optimus is is a different
36:53
matter. The price of Optimus will be set
36:56
by the market demand. Great.
37:00
Thank you very much. The next question is what is the
37:04
status on mass production of the Tesla Semi and how will it
37:08
impact revenue and scale? I can take that one.
37:12
So we just closed up the semi factory roof and walls last week
37:15
in Reno. So that has schedule which is
37:17
great. With the weather in Reno, you
37:19
never know what's going to happen.
37:20
But we're prepping for mechanical installation of all
37:23
the equipment in the coming months.
37:25
The 1st builds of the high volume Semi design will come
37:27
late this year in 2025 and begin ramping early in 2026.
37:31
But as we've said that before, you know, the Semi is ATCO, no
37:35
brainer. It gets really similar to
37:38
optimistic, you know, set by how much people pay and you know it,
37:43
it has the total cost of ownership.
37:44
It's much, much cheaper than any other transportation you can
37:47
have. So at that point, when we're at
37:49
scale, it will meaningfully contribute to, you know, Tesla's
37:52
revenue. I think it's difficult to say
37:54
how much if you want anyone. No, I mean I I do think it tells
38:01
us semi again with autonomy is going to be incredibly valuable.
38:08
You know, that we actually have a shortage of truck that drivers
38:11
in America. That's one of the limiting
38:12
factors on transport. And, you know, and people are
38:20
human, so they get tired and sometimes, you know, there's
38:25
it's, you know, I have a lot of respect for truck drivers
38:27
because it's a tough job. But because it's a tough job,
38:31
there's not that many people that want to do it.
38:33
And there's actually fewer, I believe my saying is correct.
38:37
There are fewer people entering truck driving as a profession
38:40
than are leaving it. Yes.
38:41
So when you think. Yeah, exactly.
38:43
So. So when you consider, OK,
38:45
there's more people leaving truck driving as a profession
38:47
than entering it, well, we're going to have a real logistics
38:49
problem as time goes by. So autonomy will be very
38:54
important to meet that that need.
38:59
So like, yeah, it'll, it'll I don't know.
39:03
It's it's it's it's a several billion a year opportunity,
39:06
which I don't know in this context is that these days I do
39:11
several billion a year matter. I think it does not nothing.
39:18
It's probably, you know, it might, it's probably a good ten
39:20
billion year thing. Yes, a billion a month at some
39:24
point probably. But but it's it's you know, all
39:28
this is going to pale comparison to Optimus.
39:31
So yeah, a billion a month is, it's a lot, but it's, it's not,
39:38
it's going to be like 1% of optimism something, you know.
39:43
Great. Thank you very much.
39:44
We already covered the next question in company remarks.
39:48
So moving on, is it expected that Tesla will need to upgrade
39:53
Hardware 3 vehicles? And if so, what is the timeline
39:56
and expected impact to Tesla's CapEx?
40:00
I can refer to cost there. They're really asking the tough
40:04
questions on there. I I guess we, you know, we
40:07
haven't stopped working on Hardware 3 yet.
40:09
We are still making software releases.
40:11
We released the 12.6 release recently, which was like it's
40:15
like a baby V13, but it's a significant improvement compared
40:19
to what they had previously. And you know, people are still
40:22
finding ways. There's still larger models in
40:24
the smaller models. So we don't given up on hardware
40:26
3. We're still working on it, just
40:28
the releases will trail the hardware for releases.
40:33
Great. Thanks, Sheriff.
40:34
Yeah. I mean, I think the honest
40:38
answer is that we're going to have to upgrade people's
40:44
hardware free computer for those that have bought, bought full
40:46
self driving and that is the honest answer.
40:49
And that's going to be painful and difficult, but we'll get it
40:51
done now. I'm kind of glad that not that
40:55
many people bought the FSD package.
41:00
Thanks, Elon. The next question Has Tesla
41:03
given up on ramping their Solar Roof product?
41:09
No, we're sorry, Mike. Oh, Mike, go ahead.
41:12
Yeah. Oh, yeah, I can take it.
41:14
Yeah. Solar Roof remains a core part
41:16
of the residential product portfolio and it still remains
41:20
where it draws a lot of customer interest despite it being
41:22
premium products. We've worked on multiple
41:25
iterations of engineering to make the product easier to
41:27
install and distribute by producing the SKU count.
41:31
And you know, more more recently rather than direct installation,
41:34
we are focused on growth through our nationwide network of
41:37
certified installers and many of those they've been installing
41:41
Solar Roof with us for many years.
41:45
Yeah. That's actually turned out to be
41:46
a much better way for the like it's just let the roof just
41:50
supply product to the roofing industry.
41:53
And especially when somebody's going to is getting a new roof
41:56
anyway or building a house from scratch.
41:58
Obviously this is by far the most efficient time to put in a
42:01
solar roof as opposed to putting a solar roof on a house that,
42:05
where the roof still has, you know, 20 years of life, that
42:10
that's not economically sensible.
42:11
But, but if it's a new house or the roof needs to be replaced
42:16
anyway, then solar roof can make a lot of sense.
42:20
And, and, and, and it is, it is a premium product.
42:24
It's, it's like the, you know, it's kind of like Model S, model
42:28
X or something. It's like a, it's a premium
42:30
product. I think it looks really cool and
42:34
your house, I mean your house generates electricity and if you
42:36
combine it with the Tesla Powerwall battery, then you can
42:40
be self-sufficient. So even if the grid turns off,
42:44
even if the grid turns off for several days, your house still
42:47
works and your roof looks awesome.
42:51
So it's like, I recommend anyone go ahead, you can afford it.
42:55
Get the Tesla solar roof and the Powerwall.
42:58
Your family's life might depend on it.
43:00
Get enough and and just just in terms of convenience, you know
43:04
your kids are not going to yell at you because their computers
43:06
don't work because the power went out and they're and you
43:09
can't charge your phone first work actually happens Yeah, you
43:17
wish we can't even call in one because your phone's energies.
43:21
Thank you very much. The next question was covered in
43:25
opening remarks, so we will skip that.
43:27
And the last question from say.com, what technical break
43:30
theories will define the 14 of FSD given that the 13 already
43:37
covered photon to control? Well, we're going to help a lot
43:43
further than photons to output. We've been in sort of the
43:48
nothing, but that's the situation, nothing but neural
43:51
Nets from photons to controls for a while now.
43:54
We're just improving the neural Nets.
43:58
I guess we could get into some of the technical details to some
44:01
degree I'd say I continue to be amazed by just how effective
44:09
order aggressive Transformers are at solving a wide range of
44:11
problems. I mean, Ashok is anything you'd
44:15
like to add there without giving away the sort of family secrets?
44:18
I mean, except all things we put on next already.
44:22
Yes, they continue to scale the model size a lot.
44:25
You know, we scale a bunch of V13, but then there's still room
44:28
to grow. So we're going to continue to
44:30
scale the model size. We're going to increase the
44:32
context length even more. The memory is sort of like
44:35
limited right now. We want to increase the amount
44:37
of memory also give to even minutes of context for driving.
44:42
We're going to add audio and emergency make this better, add
44:46
like data of the tricky coral cases that we get from the
44:49
entire fleet. You know, any interventions or
44:51
any kind of like, you know, user intervention, we just add that
44:54
to the data, the data set. So scaling in basically every
44:59
access to training, compute, data set size, model size, model
45:02
context and also all the reinforcement learning
45:04
objectives. Great.
45:08
Already with that, we will move over to analyst questions.
45:13
So just as a reminder, you will need to unmute yourself to ask
45:17
your question. And the first question will be
45:19
coming from Daniel Roska from Bernstein.
45:22
Daniel, please go ahead and unmute yourself.
45:25
Hey, good evening, everybody. It's Tanya from Bernstein.
45:28
Elon Tesla's share price clearly already includes quite a few of
45:32
the anticipated benefits you talked about today.
45:35
Yet, realizing what you call a kind of difficult but achievable
45:38
will take some time. What are you pushing the Tesla
45:41
executive team to do differently now to accelerate the innovation
45:46
in order to realize the value you described for the company?
45:53
Well, I mean, we're working on perfecting real world AI and
46:01
making rapid progress week over week.
46:02
If not, you know, suddenly month over month, it's often week over
46:05
week. I spent a lot of time with the
46:10
Tesla AI team and the Tesla optimist team.
46:13
I mean, I go where the problem is essentially like not, you
46:17
know, something, you know, portion sometimes like don't
46:21
talk to Tesla executive and like, hey, we don't see you very
46:23
often. I'm like, that's because your
46:24
stuff is working awesome. Your stuff working really great.
46:28
Unfortunately, I didn't see them very often because I go where
46:30
the problem is. So, you know, whatever the
46:35
problem is what like what's, what's the greatest challenge
46:37
that lies ahead? So obviously there's, there's,
46:40
there are many challenges with optimists.
46:44
It's a hard problem to solve, many challenges with the vehicle
46:49
autonomy. But we're making rounded
46:51
progress in both, yeah. OK.
46:59
I mean, it sounds like you, you've got a conviction that the
47:03
pieces you need right are, are in place.
47:07
If we kind of go 12 months down the line and we look back and
47:10
you had some of those. But maybe what are the kind of
47:13
two or three KPIs that would tell you that, you know, you're
47:17
on track and it's going the right way and the pieces you've
47:20
put in place are the right pieces, but that's kind of what
47:22
I'm looking for or other way around.
47:27
Where would it be off most likely in your mind that you
47:30
say, hey, I need to go back there and I need to change
47:32
something to enable the team better?
47:38
Well, I mean, I think my, the predictions that I'm making here
47:41
are going to be pretty accurate, you know, and it's worth an
47:47
exercise. Sometimes girls say, oh, Elon's
47:48
always late. Well, actually, no, the problem
47:50
is that the media reports on when I'm late, but never reports
47:53
when I'm early. So sure, I'm optimistic, but I'm
47:56
not that optimistic. You know, there are many cases
48:00
in past where I actually we've been early, you know, such as
48:03
completion of the Shanghai factory or in fact the factory
48:08
completion is generally been ahead of schedule, not behind.
48:13
So, yeah, so the, but I like I said, I'm very confident we'll
48:22
have released unsupervised full self driving fully autonomous
48:26
Teslas in Austin and several other cities in America by the
48:30
end of this year. That's probably everywhere in
48:34
America next year that everywhere in North America at
48:38
least. I think in terms of next year
48:41
our constraints, I think it's likely to be just regulatory,
48:46
you know, like like Europe really has for example.
48:49
Europe is a layer cake of regulations and bureaucracy
48:52
which that really needs to be addressed.
48:56
You know, there's this, you know, joke like America
49:00
innovates, Europe regulates. It's like guys, there's too many
49:05
reps on the field. I mean, for example, for us just
49:09
to release just to release supervised full self driving in
49:12
Europe, even though it works really well.
49:15
We have we have to go through a mountain of paperwork with the
49:19
Netherlands, which is our primary regulatory authority.
49:22
Then the Netherlands presents this to the EU and I think May,
49:28
and there's like this big EU country committee, we expect it
49:33
to be approved at that time, but there's nothing we can do to
49:36
make that May happen sooner. In fact, nobody seems to do.
49:41
If I guess all the countries would have to somehow vote in
49:46
some way to have it happen sooner than May, otherwise it
49:50
won't happen sooner than May. So then well then when is
49:54
unsupervised ever see a lot in Europe?
49:57
I'm like maybe next year maybe, I don't know to find out when
50:04
the EU is meeting again. Sometimes it's a 12 month
50:07
cadence, sometimes a six month cadence.
50:10
Then in China, which is the gigantic market, we do have some
50:16
challenges because they weren't currently allow us to transfer
50:20
training video outside of China and then U.S. government won't
50:24
let us do training in China. So we're in a bit of a blind
50:27
there. So but like what are a quandary?
50:32
So, So what we're already solving then is by literally
50:37
looking at videos of streets in China that are available on the
50:42
Internet to understand, and then feeding that into our video
50:48
training. So that publicly available video
50:51
of street signs and traffic rules in China can be can be
50:57
used for training. And then also putting it in a
51:01
very accurate simulator. And so it will train using SIM
51:06
for, you know, bus lanes in China, like bus lanes in China.
51:10
By the way, what about the biggest challenges in the FSD
51:16
work in China is the bus lanes are very complicated and there's
51:21
like literally like hours of the day that you're allowed to be
51:23
there and not be there. And then if you accidentally go
51:25
in that bus lane at the wrong time, the the you get an
51:28
automatic ticket instantly. So it's kind of a big deal bus
51:34
lanes in China. So we're going to put that into
51:36
our simulator train on that. The car has to know what time of
51:41
day it is, read the sign anyway, we'll get the solved.
51:45
But you know, I think we'll have anticipated FSD in almost every
51:53
market this year limited simply by regulatory issues, not
52:02
technical capability. And then and then unsupervised
52:06
FSD in the US this year in many cities, but nationwide next
52:11
year. And and hopefully we have
52:15
unsupervised FSD in most countries by the end of next
52:18
year. That that's the best my
52:21
prediction with best data that I have right now.
52:25
Great. Thank you very much.
52:26
The next question will come from Adam Jonas at Morgan Stanley.
52:30
Adam, please feel free to unmute yourself.
52:33
Thanks, everybody. So, Elon, you've said in the
52:36
past about LIDAR that for AVS at least, that LIDAR is a crutch, a
52:42
fool's errand. I think you even told me once,
52:44
even if it was free, you'd say you wouldn't use it.
52:47
You still feel that way? Yes.
52:51
Care to elaborate or just I have another question?
52:56
Look, we even have a radar in the car and we turned it off.
53:01
I got it. All right.
53:02
So you're you're still. Yeah.
53:04
People think you're crazy, you know, but for not.
53:10
Obviously humans drive without shooting lasers out of their
53:13
eyes. I mean, what's your Superman,
53:17
you know? But, but like humans drive just
53:19
with passive visual, it was drive with, with eyes and a
53:24
neural net and a brain neural net.
53:25
So sort of biological. So the digital equivalent of
53:29
eyes and a brain are cameras and, and, and digital neural
53:34
Nets or AI. So that's the entire Rd. system
53:39
was designed for passive optical neural Nets.
53:43
That's how the whole world system was not designed and what
53:46
everyone's expecting other that's how we expect other cars
53:48
to behave. So therefore that is very
53:51
obviously the solution for full self driving in it as a
53:55
generalized for the generalized solution for full self driving
53:59
as opposed to the very specific, you know, neighborhood by
54:01
neighborhood solution, which is very difficult to maintain,
54:04
which is what our competitors are doing.
54:05
I got it. Yeah, Yeah.
54:12
I mean, Lidar doesn't work in the fall guys.
54:15
Lidar has a lot of issues. Like I don't absolutely like the
54:19
SpaceX Dragon docks with the space station using LIDAR.
54:22
That a a program that I've personally spearheaded.
54:26
I don't have some fundamental bizarre dislike of Lidar.
54:30
It's simply the wrong solution for driving cars on roads.
54:34
Right you you understand how how Lidar works I I guess.
54:37
I naturally designed in both the R and red Lidar.
54:40
I oversaw the project, the engineering thing.
54:43
I, it was my decision to use Lidar and Dragon and I, I always
54:47
saw the, that, that engineering project directly.
54:50
And so I'm like, we literally designed a major, a lighter to
54:54
dock with the space station. If I thought it was the right
54:56
solution for cars I would do that, but it isn't.
54:59
Yeah. All right, just as a follow up
55:00
at CES, you said I, I, I'm paraphrasing that any AI will be
55:05
able to do any cognitive tasks not involving atoms within the
55:10
next three or four years. And that would imply, Elon, that
55:12
before the end of President Trump's term in office, that AI
55:15
would be moving pretty damn quickly into the physical world.
55:18
What are the world of photons and atoms?
55:21
And I'm, I'm thinking, given your work with the
55:24
administration, how confident are you that the US has will
55:28
have the manufacturing and the supply base to, to make good on
55:33
your excitement about physical AI by the, you know, end of by
55:36
the latter this decade? We seem pretty vulnerable right
55:38
now. I've seen you tweeting about or
55:40
sorry X saying, excuse me, Elon, about China Freudian slip about
55:45
China having like making more drones in a day than the US
55:49
makes in a year and all the entanglement of the supply.
55:52
So what has to happen in the US to make that possible?
55:55
What's your message and what can what can you do about it?
55:59
And what's relevant for Tesla shareholders?
56:01
Thanks, Elon. Well, at Tesla, obviously we
56:04
think manufacturing is cool. SpaceX, we think manufacturing
56:08
is cool, but in general for talented Americans, they need to
56:15
beyond, you know, my beyond my companies, beyond me and the
56:19
people and my teams here in general, we need to make
56:21
manufacturing full again in America.
56:25
And you know, like, I honestly think people should move from
56:29
like law and finance into manufacturing.
56:30
That's my honest opinion. We have too much.
56:33
This is both a complement and a criticism.
56:35
We have too much talent in law and engineering, law and finance
56:38
in America, and there should be more of that talent in
56:41
manufacturing. So, yeah, I mean, it tells us
56:54
we're making sure that we can continue to manufacture our
56:58
stuff even in the event of geopolitical tensions rising to
57:05
very high levels. Great.
57:09
Thank you very much. The next question will come from
57:12
Pierre Fergu at New St. Pierre.
57:15
Please feel free to unmute yourself.
57:19
Hey, thanks guys for taking the question.
57:22
So I have a question, You learn on deploying like robotaxis in
57:27
in June and in Austin. So that's great news.
57:30
And I was wondering if it means I can, you know, drive down to
57:33
Austin in June and try and try unsupervised by myself with my
57:38
car. Oh, it's going to be more like
57:42
your fleet testing it. It'll it'll be our fleet testing
57:47
it. But that's not sort of toe in
57:50
the water. You know, we'll be scrutinizing
57:53
it very carefully. Make sure it's not something we
57:56
missed. But it will be, you know, right,
58:01
autonomous ride hailing for money in, in Austin in June and
58:07
then as shortly as possible other cities in America.
58:11
And I expect us to be operating during unsupervised activity
58:17
with, with, with our internal fleet in several cities by the
58:21
end of the year. Then it's probably next year
58:25
when people are are able to add or subtract their car from the
58:29
fleet. So you know, kind of like Airbnb
58:32
where you can sort of add or subtract your house or your
58:35
guest room. You know, if you can say like
58:38
add it to the Airbnb inventory or don't add it to the Airbnb
58:42
inventory. You know, if you're, if you're
58:47
traveling for a month, you can or whatever in case maybe you
58:52
can let other people use your house and then you can anyway.
58:57
So like that's probably next year because we want to just
59:00
make sure we've ironed out any, any kinks.
59:04
And a lot of it is not like we're not splitting the atom
59:07
here. It's just a bunch of work that
59:09
needs to be done to make sure the whole thing works
59:11
efficiently, that people can order the car, it comes, you
59:16
know, it's the right spot, does exactly the right thing.
59:19
All the payment systems work. The billing works.
59:25
Yeah. OK.
59:28
But then like, so my sort of question would be, you know, I
59:32
have a Tesla, I have FSD and I have to keep my eyes on the on
59:36
the road all the time. It's super boring because I
59:38
don't really need to intervene anymore.
59:41
And the really annoying thing is that I can't just check my
59:47
emails. And so are you working also on
59:50
introducing, you know, like a kind of like free unsupervised
59:53
where I could be eyes off and, and I would be able to check my
59:58
e-mail and we just need to with a 5 second notice.
1:00:01
I have to go back and keep an eye on what's happening or?
1:00:04
Is that something you're working on as well because it feels so
1:00:07
close with with be certain that I wonder if it's something we'd
1:00:10
expect for this year. It's a very sensitive question I
1:00:12
asked for myself, to be honest. Yes, we just need, we need to be
1:00:17
very confident that the probability of injury is low
1:00:20
before we allow people to to check, check, check with their
1:00:25
e-mail and text messages. In fact, right now we're in this
1:00:29
perverse situation, which you may have encountered yourself,
1:00:31
where people will actually go to manual driving to check their
1:00:36
text messages so the computer doesn't yell at them and then go
1:00:41
then put it back on autonomous mode once they've checked the
1:00:45
text messages, which is obviously less safe,
1:00:48
significantly less safe, significantly less safe,
1:00:51
significantly less safe than just letting people check their
1:00:54
texts once in a while without the computer yelling at them.
1:00:59
But we just want to be cautious about that, the advent of that
1:01:01
we're in this sort of, you know, neither here nor there, but just
1:01:07
for, I mean, I think it's not for many months longer.
1:01:12
But yeah, we're in this perverse situation where people will turn
1:01:15
the car off autopilot so the computer doesn't yell at them,
1:01:17
check the text, text messages, well, while staring the car with
1:01:20
their knee and not looking out the window and.
1:01:26
There like like Elon said, right?
1:01:29
If you have any problems with the system and when people are
1:01:33
not looking, that is a dangerous and that's what we're trying to
1:01:37
avoid. It's the capabilities getting
1:01:40
there, but it's not fully there. That's why he was using the
1:01:43
Thermo, tipping a toe in the water, then getting comfortable.
1:01:47
Thank you. Anyway, it's it's not far off,
1:01:54
but we wouldn't want to prove to ourselves, improve, improve to
1:01:58
ourselves and obviously prove to regulators that the car is
1:02:03
unequivocally safer in autonomous mode than than not.
1:02:11
And that's we're not far off. So this is like low single digit
1:02:16
months. To to the safety aspect, we did
1:02:19
publish our safe vehicle safety report today and then.
1:02:23
Q 4 is. 1. Crash for every. 5.9 million
1:02:26
miles driven compared to a crash every 700,000 miles without all
1:02:30
kind of. Right.
1:02:31
So we're we're getting to the point where it's an order of
1:02:34
magnitude. Yeah, it's like 8.5 tons, so
1:02:36
it's just about there. You know, it's amazing.
1:02:40
Great. Alrighty.
1:02:41
And our last question will be coming from Dan Levi at
1:02:45
Barclays. Dan, feel free to unmute
1:02:47
yourself. Great, good evening.
1:02:50
Thank you for taking other questions.
1:02:53
Elon, you've talked about the need for proliferation of
1:02:57
sustainable transport in the past as part of a sort of
1:02:59
broader push to sustainable energy.
1:03:03
OK. I know we've heard a lot about
1:03:05
President Trump's plans to reverse the the EV mandate.
1:03:08
And I think there's a view that given regulation is a driver of
1:03:12
EV uptake, this could slow EV uptake in the US.
1:03:15
So what would be your view on on the right policy in in the US
1:03:22
given your comments in the past the need to push for sustainable
1:03:27
transport? You know, at this point, I think
1:03:33
that that sustainable transport is inevitable.
1:03:38
I'm highly confident that all transport will be autonomous
1:03:44
electric, including aircraft, and that it's simply it's it
1:03:49
can't be stopped any more than one could have stopped the
1:03:52
advent of the external combustion engine, steam engine,
1:03:55
or or one could have stopped the advent of the internal
1:04:00
combustion engine. Like like even if you've been
1:04:03
the biggest horse advocate on Earth, it like it like a all
1:04:08
your courses are the way got these newfangled car
1:04:12
automobiles. You can't stop the advent of the
1:04:15
automobile. It's going to happen and you
1:04:19
can't. You can't stop the advent of
1:04:21
electric cars. It's going to happen.
1:04:24
The only thing holding back electric cars was range, and
1:04:27
that is a solved problem. Great.
1:04:35
And then as a follow up, you know, in the past, Elon, you had
1:04:40
made a comment that you know, you'd be willing to sell cars
1:04:43
effectively no margin to get the cars out there.
1:04:46
And there's a comment in the release today of the rate of
1:04:50
acceleration of autonomy efforts does impact volume growth.
1:04:53
So perhaps you could just talk about, you know, with your
1:04:57
efforts on FSD, how we should think about your desire to put
1:05:03
more vehicles out in, in the market to take advantage of
1:05:08
your, of your tech advances. So I'm not sure I understand the
1:05:14
question, but we have a lot of cars.
1:05:16
I mean, you've got millions of cars out there.
1:05:19
So is, is there a question Dan that how do we, how do we marry
1:05:24
our future growth aspects with FSD?
1:05:29
And yeah, go. Ahead and meet yourself, Dan.
1:05:34
Yeah, more, more so just how much more aggressively you would
1:05:38
be willing to you know, sell your cars versus versus, you
1:05:44
know, in light of your, of your improvements on FSD.
1:05:50
Well, right now the constraint we're trying to solve is battery
1:05:54
production as opposed to demand. So you're and now Q1, we've got
1:06:01
this massive factory retooling for the new Model Y, for
1:06:04
example, that obviously has a sort of impact on output.
1:06:09
But the problem reversing with in fact we're talking the
1:06:14
executive team and I were talking about just before this
1:06:17
call was how we've got to figure out how to increase total GW
1:06:21
hours of battery production this year one way or another.
1:06:25
That's the constraint on our output.
1:06:31
Great. All righty.
1:06:32
And with that, I think we are all doing for today.
1:06:36
So thanks everyone so much for all your questions.
1:06:38
We look forward to talking to you next quarter.
1:06:41
Thank you very much and goodbye. Hey, thank you so much for
1:06:48
listening today. I really do appreciate your
1:06:50
support. If you could take a second and
1:06:52
hit the subscribe or the follow button on whatever podcast
1:06:54
platform that you're listening on right now, I greatly
1:06:57
appreciate it. It helps out the show
1:06:59
tremendously and you'll never miss an episode.
1:07:02
And each episode is about 10 minutes or less to get you
1:07:05
caught up quickly. And please, if you want to
1:07:08
support the show even more, go to patreon.com/stagezero and
1:07:14
please take care of yourselves and each other and I'll see you
1:07:17
tomorrow.