Latest / Elon Musk Podcast / Elon Musk Talks at Tesla Q1 2024 Financial Results and Q&A Webcast

Transcript

Elon Musk Talks at Tesla Q1 2024 Financial Results and Q&A Webcast · Elon Musk Podcast

0:00
My name is Martin Vieca, VP of Investor Relations, and I'm
0:03
joined today by Elon Musk, FEI, Bafta, Necha and a number of
0:06
other executives. Our Q1 results were announced at
0:09
about 3:00 PM Central Time in the update deck, we published at
0:13
the same link as this webcast. During this call, we will
0:16
discuss our business outlook and make forward-looking statements.
0:20
These comments are based on our predictions and expectations as
0:22
of today. Actual events and results could
0:25
differ materially due to a number of risks and
0:28
uncertainties, including those mentioned in our most recent
0:31
filings with the SEC during the question and answer portion of
0:35
today's call. Please limit limit yourself to
0:37
one question and one follow up. Please use the raise hand button
0:42
to join the question queue. But before we jump into Q&A,
0:45
Elon has some opening remarks. Elon.
0:49
Thanks, Martin. So to recap, in Q1 we navigated
0:53
several unposting challenges as well as the ramp with the
0:56
updated Model 3 in Fremont. There was as as we all have
1:01
seen, the EV adoption rate globally is under pressure and
1:05
and a lot of other water manufacturers are pulling back
1:09
on E BS and pursuing plug in hybrids instead.
1:12
We believe this is not the right strategy and electric vehicles
1:16
will ultimately dominate the market.
1:19
Despite these challenges, the Telstra team did a great job
1:21
executing it, executing in a tough environment and energy
1:25
storage deployments. The Megapack in particular
1:28
reached an all time high in Q1 leading to record profitability
1:31
for the energy business and that that looks likely to continue to
1:35
increase in the quarters and years ahead.
1:39
It will increase, we actually know that it will so
1:43
significantly faster than the the car business as we expected.
1:48
We also continue to expand our AI training capacity in Q1, more
1:54
than doubling our training compute sequentially.
1:58
In terms of the new product road map, there's been a lot of talk
2:01
about our upcoming vehicle line in the next.
2:04
In the past several weeks, we've updated our future vehicle
2:08
lineup to accelerate the launch of new models.
2:10
Head of previously mentioned star production in the
2:12
second-half of of 2025. So we expect it to be more like
2:15
the early 2025 if not late this year.
2:19
These new vehicles including more affordable models, we'll
2:22
use aspects of the next generation platform as well as
2:25
aspects of our current platforms and we'll be able to produce on
2:29
the same manufacturing lines as our current vehicle lineup.
2:32
This, so it's not contingent on any new factory or massive new
2:36
production line. It'll be made on our current
2:40
production lines much more efficiently and and we think
2:46
this should allow us to get to over 3 million vehicles of of
2:52
capacity when realized to the full extent regarding FSD
2:57
version 12, which is the the pure AI based self driving
3:05
people that if you haven't experienced this, I strongly
3:07
urge you to try it out. It's profound and the rate of
3:12
improvement is is rapid. So we've and we've, we've now
3:17
turned that on for all cars with the cameras and inference
3:24
computer and everything from hardware 3 on in North America.
3:28
So it's been pushed out to I think around 1.8 million
3:32
vehicles and we're seeing about half of people use it so far and
3:37
that that percentage is increasing with each passing
3:40
week. So we now have over 300 billion
3:45
miles that have been driven with FSDV 12.
3:49
Since the launch of full self driving, supervised full self
3:52
driving, it's become very clear that the vision based approach
3:55
with end to end neural networks is the right solution for
3:58
scalable autonomy. And it's it's really how humans
4:01
drive. Our entire Rd. network is
4:04
designed for biological neural Nets and eyes.
4:08
So naturally cameras and digital neural Nets are the solution to
4:13
our current Rd. system. To make it more accessible,
4:18
we've reduced the subscription price to $99.00 a month.
4:21
So it's easy to try out. And as we've announced, we'll be
4:26
showcasing our purpose built robot taxi or cyber cap in
4:31
August. Yeah, regarding, regarding AI
4:38
compute, over the past few months we've been actively
4:41
working on expanding Tesla's core AI infrastructure.
4:45
For a while there we were training constrained in our
4:47
progress. We are at this point no longer
4:50
training constrained and so we're making rapid progress.
4:55
We've installed and and commissioned, meaning they're
4:57
actually working 35,000 H 100 computers or GPUs or GPU is
5:05
wrong with they need a new word. I always feel like a wince when
5:10
I say GPU because it's not GPU. That's G central graphics and
5:13
doesn't do graphics. But anyway, roughly roughly 35H1
5:19
hundreds are active and we expect that to be probably
5:24
85,000 or their bounce by the end of this year in training,
5:28
just for training, we are making sure that we're being as
5:33
efficient as possible in our training.
5:34
It's not just about the number of H1 hundreds, but how
5:38
efficiently they're used. So in conclusion, we're super
5:42
excited about our autonomy road map.
5:44
I think it should be obvious to anyone who's driving version 12
5:48
in a test that that is only a matter of time before we exceed
5:53
the reliability of humans and not much time with that.
5:58
And and we're really headed for an electric vehicle, an
6:04
autonomous future. And I'll go back to something I
6:09
said several years ago that in the future, gasoline, gasoline
6:16
cars that are not autonomous will be like riding a horse and
6:19
using a flip boat. And that will become very
6:22
obvious in in hindsight. We continue to make the
6:26
necessary investments that will drive growth and profits for
6:28
Tesla in the future. And I wanted to thank the Tesla
6:31
team for incredible execution during this period and look
6:34
forward to everything that we have planned ahead.
6:36
Thanks. Thank you very much and Abhay
6:38
Bap has some comments as well. Thanks.
6:41
You know, it's important to acknowledge what Elon said.
6:47
From our auto business perspective, we did see a season
6:51
decline in revenues quarter over quarter and there's a primarily
6:55
because of seasonality uncertain macroeconomic environment and
7:00
the main and the other reasons which Elon had mentioned
7:02
earlier. Auto margins declined from 18.9
7:06
to 18.5% excluding the impact of cyber truck.
7:10
The impact of pricing actions was largely offset by reductions
7:14
in per unit cost and the and the recognition of revenue from auto
7:18
part feature for certain vehicles in the US that
7:22
previously did not have that functionality.
7:24
Additionally, while we did experience higher cost due to
7:28
the ramp of Model 3 in Fremont and disruptions in Berlin, these
7:32
costs were largely offset by cost reduction initiatives.
7:36
In fact, if you exclude Cybertruck and Fremont and
7:40
Fremont Model 3 Ram cost, the revenue from Auto Park auto
7:44
margins improved slightly. Currently normalized Model Y
7:48
cost per vehicle in Austin and Berlin are already very close to
7:52
that of Fremont. Our ability to reduce costs
7:56
without sacrificing on quality was due to the amazing efforts
7:59
of the team in executing Tesla's relentless pursuit of efficiency
8:03
across the business. We've also witnessed that as
8:08
other OEMs are pulling back on their investments in EV, there
8:11
is increasing appetite for credits and that means a steady
8:15
stream of revenue for us. Obviously, seeing others pull
8:18
back from EV is not the future we want.
8:21
We would prefer it. The whole industry went all in
8:25
on the demand front. We've undertaken a variety of
8:28
initiatives, including lowering the price of both the purchase
8:32
and subscription options for FSD, launching extremely
8:36
attractive leasing specials for the Model 3 in the US for 299 a
8:40
month and offering attractive financing options in certain
8:44
markets. We believe that our awareness
8:46
activities paired with attractive financing will go a
8:49
long way in expanding our reach and driving demand for our
8:52
products. Our energy business continues to
8:56
make meaningful progress with margins reaching a record of
8:59
24.6%. We expect the energy storage
9:02
deployments for 2024 to grow at least 75% higher from 2023 and
9:09
accordingly this basis will begin contributing significantly
9:12
to our overall profitability. Note that there is a bit of
9:16
lumpiness in our storage deployments due to a variety of
9:18
factors that are outside of our control, so deployments may
9:22
fluctuate quarter over quarter. On the operating expense front,
9:26
we saw a sequential increase from our AI initiatives,
9:32
continued investment in future projects, marketing and other
9:35
activities. We had negative free cash flow
9:38
of 2.5 billion in the first quarter.
9:40
The primary driver of this was an increase in inventory from a
9:43
mismatch between bills and deliveries as discussed before
9:47
and our elevated spend on CapEx across various initiatives
9:51
including AI compute. We expect the inventory build to
9:55
reverse in the second quarter and free cash flow to return to
9:58
positive again. As we prepare the company for
10:01
the next phase of growth, we are to make the hard but necessary
10:05
decision to reduce our head count by over 10%.
10:09
The savings generated are expected to be well in excess of
10:12
1.1 excess of 1 billion on an annual 100 basis.
10:18
We are also getting hyper focus on CapEx efficiency and
10:21
utilizing our installed capacity in a more efficient manner.
10:26
The savings from these initiatives, including our cost
10:29
reductions, will help improve our overall profitability and
10:33
ultimately enable us to increase the scale of our investments in
10:36
AI. In conclusion, the future is
10:40
extremely bright and the journey to get there, while challenging,
10:45
will be extremely rewarding. Once again, I would like to
10:48
thank the whole Tesla team for delivering great results and we
10:53
can open it up to Q&A. Thank you.
10:56
OK, Let's start with investor Q&A.
10:58
The first question is what is the status of 4680?
11:02
What is the current output? Lars sure. 4680 production
11:07
increased about 1820% over from Q4, reaching greater than one KA
11:12
week for Cybertruck, which is about 7 GW hours per year.
11:15
As we posted on X, we expect to stay ahead of the Cybertruck
11:18
ramp with the cell production throughout Q2 as we ramp the
11:22
third of four lines in phase one while maintaining multiple weeks
11:25
of cell inventory to make sure we're ahead of the ramp because
11:29
we're ramping. COGS continues to drop rapidly
11:31
week over week driven by yield improvements throughout the
11:34
lines and production volume increases.
11:36
So our goal and we expect to do this is to beat supplier cost of
11:41
nickel based sales by the end of the year.
11:44
Thank you. The the second question is on
11:47
Optimus. So what is the current status of
11:49
Optimus? Are they currently performing
11:51
any factory tasks? When do you expect to start mass
11:54
production? We are able to do simple factory
12:01
tasks or at least I should say factory tasks in the lab.
12:05
The in terms of actually we do, we do think we will have Optimus
12:11
in limited production in the factory and the actual factory
12:14
itself doing useful tasks before the end of this year.
12:18
So, and then I think we, we may be able to sell it externally by
12:26
the end of next year. These are just just guesses.
12:31
As I've said before, I think Optimus will be more valuable
12:36
than everything else combined. Because if, if you've got a, a
12:41
sentient humanoid robot that is able to navigate reality and do
12:48
tasks at request, there is no meaningful limit to the size of
12:54
the economy. So that's, that's what's going
13:00
to happen. And I think Tesla is best
13:03
position of any humanoid robot maker to be able to reach volume
13:10
production with efficient inference on the robot itself.
13:17
The I mean, this perhaps is a point that is worth emphasizing.
13:24
Tesla's inference AI inference efficiency is vastly better than
13:28
anyone, any other company. There's no company even close to
13:32
the inference efficiency of Tesla.
13:35
We've we've had to do that because we were constrained by
13:37
the inference hardware in the car.
13:39
We'd never choice but that that will pay dividends in many ways.
13:47
Thank you. The third question is what is
13:50
the first current assessment of the pathway towards regulatory
13:53
approval for unsupervised FSD in the US and how should we think
13:58
about the appropriate safety threshold compared to human
14:00
drivers? Sure.
14:03
I can start. There are a handful of states
14:05
that already have adopted autonomous vehicle laws.
14:08
These states are paving the way for operations while the the
14:11
data for such operations guides A broader adoption of driverless
14:15
vehicles. I think a show can talk a little
14:17
bit about our safety methodology, but expect that
14:20
these States and the work ongoing as well as the data that
14:23
we're providing will pave the way for a broad based regulatory
14:27
approval in, in, in in the US at least and then other countries
14:31
as well. Yeah, it's actually been pretty
14:35
helpful that other autonomous car companies have been cutting
14:40
a path through the regulatory jungle, but the which is so
14:45
that's, that's actually quite helpful.
14:48
And they, they have obviously been operating in temperature
14:50
scope for a while. I think they got approval for
14:53
city of LA. So these, these approvals are
14:56
happening rapidly. I, I think if you've got at
14:59
scale, it's a statistically significant amount of data that
15:04
shows conclusively that the autonomous car has let's say
15:11
half the accident rate of a human driven car.
15:15
I think that's difficult to ignore because at that point
15:18
stopping autonomy means killing people.
15:22
So I actually do not think that there will be significant
15:26
regulatory barriers, provided there is conclusive data that
15:29
the autonomous car is safer than a human driven car.
15:34
And in my view this will be much like elevators.
15:37
Elevators used to be operated by a guy with a relay switch, but
15:41
sometimes that guy would get tired or drunk or just make a
15:45
mistake and cheer somebody in half between floors.
15:48
So now we just have, we just get in an elevator and press button.
15:51
We don't think about it. In fact, it's kind of weird if
15:55
somebody's standing there with relay switch.
15:58
That'll be how cars work. You just some in a car using
16:03
your phone. You get in, it takes you to a
16:05
destination, you get out you. Don't even think about it.
16:08
You don't even think about it. Just like an elevator, it takes
16:12
you to A to your floor. That's it.
16:15
Don't think about how the elevator is working or anything
16:17
like that. And, and something I should
16:25
clarify is that Tesla will be operating the fleet.
16:30
So you can think of like how Tesla think of Tesla, like under
16:34
some combination of Airbnb and Uber, meaning that you know,
16:39
there'll be some number of cars that Tesla owns itself and
16:42
operates in the fleet. There'll be some number of cars
16:45
and then there'll be a bunch of cars where they're owned by the
16:49
end user. But that end user can add or
16:53
subtract their car to the fleet whenever they want.
16:56
And they can decide if they want to only let the car be used by
17:00
friends and family or only by 5 star users or by anyone.
17:06
And at any, at any time, they could have the car come back to
17:09
them and be exclusively theirs, like an Airbnb.
17:13
You could rent out your guest room or not anytime you want.
17:20
So as our fleet grows, we have 7 million cars going to 9 million
17:27
cars going to, you know, eventually 10s of millions of
17:30
cars worldwide with a, with a, with a constant feedback loop.
17:38
Every time something goes wrong, that that gets added to the
17:42
training data and you get this training flywheel happening in
17:48
the same way that Google search has the sort of flywheel.
17:51
It's very difficult to compete with Google because people are
17:54
constantly doing searches and clicking and, and Google's
17:57
getting that feedback loop. It's the same with with Tesla,
18:01
but at A, at a scale that is maybe difficult to comprehend,
18:06
but ultimately it'll be 10s of millions.
18:12
I, I think there's also some potential here for an AWS
18:16
element down the road where if we've got very powerful
18:23
inference, you know, because we've got a hardware 3 in the
18:29
cars not, but now all cars are being made with hardware 4.
18:32
Hardware 5 is pretty much designed and should be in cars
18:37
hopefully towards the end of next year.
18:41
And there's there's the potential to have for the to to
18:46
run when the car is not moving to to actually run distributed
18:51
inference So kind of like AWS, but it distributed inference
18:57
like it takes a lot of computers to train an AI model, but many
19:02
orders 92 less compute to run it.
19:05
So if you can imagine the Future Past where there's a fleet of
19:11
100 million Teslas and on average they've got like maybe a
19:16
kilowatt of inference compute, that's 100 gigawatts of
19:19
inference compute distributed all around the world.
19:24
It's pretty hard to put together 100 gigawatts of AI compute.
19:30
And even in an autonomous future where the the car is perhaps
19:35
used instead of being used 10 hours a week is used 50 hours a
19:38
week, that still leaves over 100 hours a week where the car
19:41
inference computer could be doing something else and it
19:47
seems like it'll be a waste not to use it.
19:52
Actually, do you want to chime in on the air process and
19:54
safety? Yeah, we have like multiple
19:56
tiers validating the safety for like every in any given week we
20:01
train hundreds of neural networks that can produce, you
20:04
know, different trajectories for how to drive the car.
20:06
They repay them through the millions of clips that we have
20:09
already collected from our users and our own QA.
20:13
Those are like critical events, you know, like someone jumping
20:15
out in front or like other critical events that we have
20:18
gathered database or many, many years and we replay through all
20:21
of them to make sure that we are net improving safety.
20:24
And top of it, we have simulation systems that also try
20:26
to recreate this and test this in close to fashion.
20:29
Once all of this is validated, we give it to our own QA
20:33
reverse. We have hundreds of them in
20:35
different cities in San Francisco, Los Angeles, Austin,
20:38
New York, lot of different locations.
20:40
They are also driving the some collecting real world miles and
20:44
we have an estimate of what are the critical events, are they
20:47
net improvement compared to the previous week's bills.
20:49
And once we have confidence that the build is a net, net
20:53
improvement, then we start shipping to early users, like
20:56
2000 employees initially that they would like get the build
20:59
they would give feedback on like if it's an improvement or are
21:01
they noting some new issues that we did not capture in our own QA
21:04
process. And only after all of this is
21:06
validated, then we go to external customers.
21:09
And even when we go external, we have like live dashboards of
21:12
monitoring every critical event that's happening in the feed
21:14
sorted by, you know, the criticality of it.
21:18
So we are having a constant pulse on the build's quality and
21:21
the safety improvement along the way.
21:23
And then any failures like you don't alluded to, you get the
21:26
data back, add it to the training and that improves the
21:29
model in the next cycle. So we have this like constant
21:32
feedback loop of issues, fixes, evaluations and then rinse and
21:36
repeat. And especially with the new
21:38
virtual architecture, all of this is automatically improving
21:41
without requiring much engineering interventions in the
21:44
sense that people, engineers don't have to be creative in
21:48
like how they code the algorithms.
21:49
It's mostly learning on its own based on data.
21:52
So you see that OK, we failure or like this is how a person
21:56
shows this is how you drive this intersection or something like
21:58
that. They get the data back, we add
22:00
it to the neural network and it learns from that training it
22:02
automatically instead of some engineers saying that oh, here
22:06
you must rotate the steering wheel by this much or something
22:08
like that. There's no hard if it's
22:10
conditions, it's everything is neural network.
22:12
It's very soft, it's probabilistic.
22:15
So it will adapt its probability probability distribution based
22:18
on the new data. That it's getting, yeah.
22:21
And we, we do have some insight into how good the things will be
22:26
in like let's say three or four months, because we have advanced
22:29
models that are, are far more capable than what is in the car,
22:34
but have have some issues with them that we need to, to fix.
22:37
So they're like there'll be there'll be a step change
22:41
improvement in the capabilities of the car, but it'll have some
22:44
quirks that are that need to be addressed in order to release
22:49
it. As Shrock was saying, we have to
22:51
be very careful in what we release the fleet or to to
22:55
customers in general. So, so like if we look at say
23:00
12.4 and 12.5, which are really could arguably even be version
23:04
13 and version 14 because it's pretty close to total retrain of
23:10
the neural Nets and in each case or substantially different.
23:16
So we have been insight into where the where the models,
23:21
where, how the call, how well the call will perform in say
23:25
three or four months. Yeah.
23:26
It turns out like scaling loss. People in the AI community
23:28
generally talk about model scaling loss, where they
23:30
increase the model size a lot and then they have corresponding
23:34
gains in performance. But we have also figured out
23:37
scaling loss and other access. In addition to the model size
23:40
scaling, we can also data scaling.
23:42
You can increase the amount of data you use to train the neural
23:44
network and that also gives similar gains.
23:48
And you can also scale up by training compute.
23:50
You can train it for much longer and get or like one more GPUs or
23:54
more Dojo nodes and that also gives better performance.
23:58
And you can also have architecture scaling where you
24:01
come up with better architectures that for the same
24:04
amount of compute produce better results.
24:06
So a combination of model size scaling, data scaling, training
24:09
compute scaling and the architecture scaling, we, we can
24:12
basically like OK with the continuous scaling based on
24:16
these, this ratio, we can sort of predict future performance.
24:20
Obviously it takes time to do the experiments because it
24:22
takes, you know, few weeks to train, it takes few weeks to
24:25
collect 10s of millions of video clips and process all of them.
24:29
But you can estimate what's going to be the future progress
24:32
based on the trends that we have seen in the past and they're
24:35
generally help through based on the past data.
24:40
OK. Thank you very much.
24:41
Let's go to the next question, which is can we get an official
24:44
announcement of the timeline for the $25,000 vehicle?
24:49
I think we Elon mentioned it in the opening remarks, but as we
24:55
as you mentioned, we're updating our future vehicle line up to
24:57
accelerate the launch of our low cost vehicles in a more CapEx
25:00
efficient way. That's our mission to get the
25:03
cheap, the most affordable cars to customers as fast as
25:06
possible. These new vehicles we built on
25:09
our existing lines and open capacity and that's a major
25:11
shift to utilize all our capacity with marginal CapEx
25:14
before we go spend high CapEx to do anything.
25:16
Yeah, we'll talk about this more on August 8th.
25:19
So, but, but really the, the way to think of Tesla is, is almost
25:25
almost entirely in terms of solving autonomy and being able
25:28
to turn on that autonomy for a gigantic fleet.
25:32
And I think it it it might be the biggest asset value
25:34
appreciation in history when that day happens when you can do
25:37
unsupervised full self driving. 5 million cars.
25:41
Yeah, a little less. Well, Tesla was having Yeah,
25:45
yeah, you know, it'll be, it'll be 7 million cars, you know, in
25:52
a year or so. Yeah.
25:55
And then 10 million and then, you know it eventually we're
25:58
talking about 10s of millions of cars.
26:01
Not eventually. It's like pulling out of this.
26:05
Yeah, pulling in the decade. It's several 10s of millions of
26:07
cars I think. Thank you.
26:09
The next question is what is the progress of Cybertruck ramp?
26:15
I can take that one to Cybertruck at one KA week.
26:18
Just a couple weeks ago. This happened in the first four
26:20
to five months since we SOP D late last year.
26:23
Of course, volume production is what matters, that's what drives
26:25
costs. And so our costs are dropping.
26:28
But the ramp still faces like a lot of challenges with so many
26:30
new technologies, some supplier limitations, etcetera and will
26:33
continue to ramp this year just focusing on cost efficiency and
26:37
and quality. OK.
26:41
Thank you. The next question have any of
26:44
the legacy automakers contacted Tesla about possibly licensing
26:47
FSD in the future? We're in conversations with one
26:52
major automaker regarding licensing FSD.
26:57
Thank you. The next question is about the
27:00
robotaxi unveil. Elon already talked about that,
27:03
so we'll have to wait till August.
27:06
The following question is about the next generation vehicle.
27:09
We already talked about that. So let's go to the semi.
27:12
What is the timeline for scaling semi?
27:17
I think awesome. So we're finalizing the
27:21
engineering of the semi to enable like a super cost
27:24
effective high volume production with our learnings from our
27:26
fleet and pilot, you know our pilot fleet and Pepsi's fleet
27:30
which we were you know expanding this year marginally in parallel
27:34
as we showed in the the shareholders deck, we have
27:37
started construction on the factory in Reno.
27:40
Our first vehicles are planned for late 2025 with external
27:43
customers starting in 2026. OK, a couple more questions.
27:49
So our favorite, can we make FSD transfer permanent until FSD is
27:55
fully delivered with level 5 economy?
27:58
OK, next question, what is the getting, what is getting the
28:05
production ramp at Lathrop? Where do you see the Megapack
28:08
run rate at the end of the year, Mike?
28:11
Yeah, Yeah, Lathrop is ramping as planned.
28:14
We have our second GA line allowing us to increase our exit
28:17
rate from 20 GW hours per year to at the start of this year to
28:21
40 GW hours per year by the end of the year.
28:24
That's that Lions commissions, there's really nothing limiting
28:27
the ramps. It's, you know, given the longer
28:29
sale cycles for these large projects, we typically have
28:32
order visibility 12 to 24 months prior to ship dates.
28:35
So, so we're able to plan the build plan several quarters in
28:39
advance. So this allows us to ramp the
28:42
factory to align with the business and order growth.
28:45
And lastly, we'd like to thank our customers globally for their
28:47
trust and Tesla as a partner for these incredible projects.
28:53
OK. Thank you very much.
28:54
Let's go to analyst questions. The first question comes from
28:57
Tony Sakanagi from Bernstein. Tony, please go ahead and
29:00
unmute. Thank you for taking the
29:03
question. I I was just wondering if you
29:06
could elaborate a little bit more on kind of the new vehicles
29:11
that you talked about today. Are these like tweaks on
29:16
existing models giving that they're going to be running on
29:22
the same lines or these like new models and how should we think
29:27
about them, You know, in the context of like the Model 3
29:31
Highland update, what will these models be like relative to that?
29:37
And given the QuickTime frame, you know, Model 3 Highland is
29:40
required a lot of work and a lot of retooling.
29:43
Maybe you can help put that all in context.
29:45
Thank you. And I have a follow up please.
29:48
I think we've said all we we will on that front.
29:51
So what's your follow? It's a more personal one for
29:57
you, Elon, which is that you're leading many important companies
30:00
right now. Maybe you can just talk about
30:03
where your heart is at in terms of your interests and do you
30:07
expect to lessen your involvement with Tesla at any
30:10
point over the next three years? Well, as a constitutes the
30:18
majority of my work time and I work pretty much every day of
30:24
the week, it's rare for me to take a slightly afternoon
30:26
afternoon off. So I'm going to make sure Tesla
30:34
is very prosperous and it is like it is prosperous and it
30:39
will be very much so in the future.
30:44
OK. Thank you.
30:45
Let's go to Adam Jonas from Morgan Stanley.
30:48
Adam, please go ahead. Go ahead and unmute.
30:53
OK, great. Hey, Elon, So you, you and your
30:57
team on volume expect a 2024 growth rate notably lower than
31:03
that achieved in 2023. But what's your team's degree of
31:06
confidence on growth above 0 percent?
31:09
Or in other words, does that statement leave room for
31:12
potentially lower sales year on year?
31:17
No, I think we'll have higher sales this year than last year.
31:21
OK, my follow up Elon on future product.
31:26
If you had nailed execution, assuming that you nail execution
31:29
on your next Gen. cheaper vehicles, you know, more
31:33
aggressive giga castings. I don't want to say one piece,
31:36
but getting closer to one piece structural pack unbox 300 mile
31:40
range, $25,000 price point. Putting aside robotaxi, those
31:44
features unique to you, how long would it take your best Chinese
31:49
competitors to copy a cheaper and better vehicle that you
31:54
could offer a couple years from now?
31:56
How long would it take your best Chinese competitors to copy
31:58
that? Thanks.
32:02
I mean, I don't know what our competitors could do except
32:09
we've we've done relatively better than they have.
32:12
You know, if you look at the drop in our competitors in China
32:15
sales versus our drop in sales, our drop is less than theirs.
32:18
So we're doing, we're doing well.
32:24
But you know, I, I, I think, you know, Kathy Wood said it best
32:30
like really we should be thought of as an AI robotics company.
32:33
If, if you value Tesla as a, as just like a, an order company,
32:37
you would just have to fundamentally it's just the
32:40
wrong framework. And it will come to be if you,
32:42
if you ask the wrong question, then the right answer is
32:45
impossible. So I, I mean, I, if, if somebody
32:53
doesn't believe Tel's going to solve autonomy, I, I, I think
32:55
they should not be an investor in the company like that.
33:00
That is, but we will and we are. And then you you, you have a car
33:06
that goes from 10 hours of use a week, like an hour and a half a
33:10
day to probably 50, but it costs the same.
33:18
I, I think that's the key thing to remember, right?
33:21
Especially if you look at FSD supervised, if you didn't
33:25
believe in autonomy, they should, they should give you a
33:29
preview that this is coming. It's actually getting better day
33:32
by day. Yeah, if if you if you've not
33:35
tried the FSD 4.3 and like I said, 12.4 is going to be
33:42
significantly better than 12.5, even better than that and we
33:46
have visibility into those things, then you really don't
33:48
understand what's going on. It's not possible.
33:52
Yeah. And and that's why we can't just
33:54
look at just as a car company because a car company would just
33:57
have a car. But here we have more than a car
34:00
company because the cars can be autonomous.
34:03
And like I said, it's happening. Yeah, this is all in addition to
34:07
test the sorting. The overall AI community is just
34:10
like increasing like improving rapidly.
34:13
Yeah, yeah. I mean, we're putting the actual
34:14
auto in automobile. So, you know, so sort of like,
34:20
well, sort of like tell us about future horse carriages you're
34:23
making. I'm like, well actually it
34:25
doesn't need a horse. That's the whole point.
34:29
That's that's really the whole point.
34:33
OK. Thank you.
34:34
The next question comes from Alex Potter from Piper Sandler.
34:38
Alex, please go ahead and unmute.
34:42
Great. Thanks.
34:44
Yeah, So couldn't agree more. The thesis hinges completely on
34:48
AI, the future of AI, full self driving, you know, neural net
34:52
training, all of these things. In that context, Elon, you've
34:57
spoken about your desire to obtain 25% voting control of the
35:02
company. And I, I understand completely
35:05
why that would be so I'm not necessarily asking about that.
35:11
I'm asking if you've come up with any mechanism by which you
35:14
can ensure that you'll obtain that level of voting control.
35:17
Because if not, then the core part of the thesis could
35:20
potentially be at risk. So any additional commentary you
35:24
might have on that topic. Well, I think no matter what
35:29
tells the, you know, even if I kidnap my aliens tomorrow, Tesla
35:35
will solve autonomy maybe a little slower, but it would
35:38
solve autonomy for vehicles at least.
35:39
I don't know if it would win on with respect to Optimus or with
35:43
respect to future products, but it would.
35:48
But there's not momentum or Tesla to solve autonomy.
35:52
Even if I disappeared for for vehicles now there's there's a
35:59
whole range, a whole range of things we can do in the future.
36:01
Beyond that, I'd be more reticent with respect to
36:04
Optimus. You know, if we have a super
36:08
sentient humanoid robot that can follow indoors and you bet you
36:13
can't escape, you know, we're talking Terminator level risk,
36:16
then yeah, I'd be uncomfortable with, you know, if there's not
36:23
some meaningful level of influence over how that is
36:26
deployed. And you know, if, if there's
36:35
shareholders have an opportunity to ratify or re ratify the, the,
36:41
the sort of competition, I guess I can't say that, but that is a
36:45
fact. They have an opportunity.
36:47
OK, very good. And yeah, we'll see if, if, if
36:53
company generates a lot of positive cash flow, we can
36:55
obviously buy back shares. All right, that's, that's
36:59
actually all Very helpful context.
37:00
Thank you. Maybe one final question, I'll
37:03
pass it on OpEx reductions. Thank you for quantifying the
37:08
impact there. I'd be interested also in
37:12
potentially more qualitative discussion of what the
37:15
implications are for these headcount reductions.
37:17
What are the types of activities that you're presumably
37:21
sacrificing as a result of parting ways with with these
37:25
folks? Thanks very much.
37:29
So, you know, like we said, we've done these headcount
37:33
reductions across the board and you know, as companies grow over
37:37
time, you've, you know, there are certain redundancies,
37:40
there's some duplication of efforts which happens in certain
37:43
areas. So you need to go back and look
37:46
at where, where, where all these pockets are, get rid of it.
37:50
So we're basically going through that exercise wherein we're
37:53
like, hey, how do we, how do we set this company right for the
37:57
next phase of growth? And the way to think about it
38:02
is, you know, any tree which grows it needs pruning.
38:07
This is the pruning exercise which we went through.
38:10
And at the end of it will be much stronger and much more
38:14
resilient to deal with the future because the future is
38:17
really bright. Like I said in my opening
38:19
remarks, we just have to get through this period and get
38:23
there. Yeah, we're, we're not giving up
38:28
anything that it's significant that I'm worried of.
38:32
So we've just we've had a lot long period of prosperity from
38:37
2019 to now and you know so if a company sort of organizationally
38:43
is 5% wrong per year, you know that accumulates to 2530% of
38:50
inefficiency. We've made some corrections
38:52
along the way, but but it is, it is time to reorganize the
38:55
company for the next phase of growth.
38:59
And you really need to reorganize it just like a, you
39:02
know, a human when you restart off with one cell and kind of
39:06
zygote and, you know, blastocyst and you start growing arms and
39:12
legs and briefly you have a tail and, you know, so.
39:18
But you shed the tail. You.
39:19
Shed the tail hopefully. And then your baby and you, you
39:23
know, you basically you, you have to be different.
39:27
The Organism, a company is kind of like, you know, creature
39:31
growing and it, if you don't reorganize it for different
39:34
phases of growth, it will fail. You can't have the same
39:39
organizational structure if you're, you know, 10 cells
39:44
versus 100 versus 1,000,000 versus a billion versus a
39:47
trillion. You know, where humans are like
39:51
around 35 trillion cells doesn't feel like, it feels like, you
39:57
know, like one person, but, but you know, you're, you're
40:02
basically a walking cell colony of roughly 35 trillion depending
40:06
on your body mass and about three times that number in
40:09
bacteria. So anyway you you've got to
40:14
reorganize the the the company for a new phase of growth or it
40:19
will fail to achieve that growth.
40:23
Thank you. Let's go to Mark Delaney from
40:25
Goldman Sachs. Mark, please go ahead and
40:28
unmute. Yes, good afternoon.
40:31
Thanks very much for taking the question.
40:33
The company previously characterized potential FSD
40:35
licensing discussions as in the early phase and some OEMs had
40:39
not really been believing in it. Can you elaborate on how much
40:42
the licensing business opportunity you mentioned today
40:44
has progressed? And is there anything Tesla
40:46
needs to achieve with the technology in terms of product
40:48
milestones in order to be successful at reaching the
40:50
licensing agreement in your view?
40:53
But I think we just need to, it just needs to be obvious that
40:57
our approach is the right approach and I think it is.
41:01
I think we're now with 12.3. If you just have the car drive
41:05
you around, it is obvious that our solution with a relatively
41:10
low cost inference computer and standard cameras can achieve
41:17
self driving. No lidars, no radars, no
41:21
ultrasonics, nothing. Just no heavy integration work
41:24
for vehicle. Manufacturers, yeah, it's so it
41:27
really just be a case of, you know, having them use the same
41:33
cameras and adverse computer and licensing our software and but
41:40
but it's it's it's once it becomes obvious that if you
41:42
don't have this in a car, nobody wants your car.
41:46
Yeah, it's like it's a smart car.
41:48
And you know, I mean, I still remember like in the back when
41:52
Nokia was king of the hill. Yeah, cell phone, yeah.
41:56
Crushing and and and and I slowly come out with a
42:01
smartphone that was basically a brick with limited
42:04
functionality. And then, you know, the iPhone
42:09
and Android, but people still did not understand that all the
42:14
phones are going to be that way. There's not going to be any flip
42:16
phones if there'll be a niche product.
42:19
Or home phones. Yeah, not even exactly.
42:21
When was the last time you saw a home phone?
42:25
Is it like an acronym? Yeah, in a hotel.
42:28
Sometimes in a hotel. Yeah, the hotels have them,
42:30
Yeah. So if people don't understand
42:35
all cars will need to be smart cars or they or or you will not
42:39
sell this. The car will not, nobody will
42:41
buy it. Once that becomes obvious, I
42:45
think licensing becomes not optional, becomes a method of
42:51
survival. Yeah, it's license in order to
42:54
buy your car. I mean, one other thing which
42:58
I'll add is in the conversations which we've had with some of
43:02
these OEMs, I just want to also point out that they take a lot
43:05
of time in their product life cycle.
43:08
Yeah, they're talking about years before they will, you
43:12
know, put it in their product. They might, we might have a
43:15
licensing deal earlier than that, but it takes a while.
43:19
So this is where the big difference between us and them
43:22
is but. Yeah.
43:23
I mean really a deal sign now would result in it being in a
43:28
car in probably three years that.
43:31
Would be early. Yeah, that's like lightning
43:33
basically. So.
43:35
That's being an eager OEM. Yeah.
43:40
So I mean, I, I wouldn't be surprised if we do sanity.
43:44
I think we're good chance we do sign a deal this year, maybe
43:49
more than one. But yeah, it would be probably
43:52
three years before it's integrated with a car, even
43:54
though all you need is cameras and our parents computer.
43:59
So it's talking about a massive design change.
44:03
Yeah. And again, just to clarify, it's
44:05
not that work which we have to do, it's the work which they
44:08
have to do with the paper time. Yeah.
44:16
Yeah, very helpful. Thank you.
44:17
My follow up was to better understand Tesla's approach to
44:19
pricing going forward. Previously, the company had said
44:21
that the price reductions were driving incremental demand.
44:24
With how affordable the cars have become, especially for
44:27
vehicles that have access to IRA credits and some of the leasing
44:29
offers that Tesla has in place, do you still see meaningful
44:32
incremental price reductions as making sense from here for the
44:35
existing products? And can the company immediately
44:38
lower prices from here and also stay free cash flow positive on
44:41
an annual basis with the current product set?
44:42
Thanks. Yeah.
44:47
I, I think we, we can be free cash flow positive meaningfully,
44:55
yeah. I think Vibov said it in his
45:00
opening remarks like our cost down efforts, we basically we're
45:04
offsetting the price cut. So I was going to say, oh, like
45:06
we're trying to give it back to the customers.
45:10
Yeah. I mean, at the end of the day,
45:12
like for any given company, if you sell a great product at a
45:14
great price, the sale, if if you have a great product at a great
45:18
price, the sales will be excellent.
45:20
That's true of any arena. So over time, we do need to keep
45:25
making sure that we're that that's a great product at a
45:29
great price and moreover that that price is accessible to
45:31
people. So it's not you have to solve
45:33
both the value for money and the fundamental affordability
45:36
question. The fundamental affordability
45:38
question is sometimes overlooked.
45:42
If if somebody's earning 100 several $100,000 a year, they
45:45
they don't think of a car from a fundamental affordability
45:48
standpoint. But from vast majority of people
45:51
are living paycheck to paycheck. So it actually makes difference
45:55
if the if the cost per month for lease or financing is $10 one
45:59
way or the other. So it is important to keep
46:04
improving the affordability and to and to keep this is like
46:09
making the price more accessible.
46:12
Yeah, the the exactly make the price more accessible, the value
46:14
for money better and to keep keep improving that over time.
46:18
But also the main kick ass if people want to buy.
46:20
Yeah, it's got to be a great product at a great price.
46:23
And and the, the, the standards for what constitutes great
46:28
product at a great price keep keep increasing.
46:32
So there's like you can't just be static.
46:34
You have to keep saying keep making the car better, improving
46:37
the price but improving the cost of production and that's what
46:41
we're doing. Yeah.
46:42
And in fact, like I said in my opening remarks, also like the
46:46
revised, the updated Model 3 is a fantastic card.
46:50
I don't think people fully even understand the model of
46:53
engineering effort which has gone.
46:54
And Lars and team have actually put out videos explaining how
46:58
much the car is different when it looks and feels different.
47:02
Not only it looks and feels different, we've added so much
47:04
value to it. But you can lease it for like as
47:07
low as 299 a month, yeah. Without gas.
47:12
Yeah. All right.
47:15
The next question comes from George from Canaccord.
47:18
George, please go on on mute. Hi.
47:24
Thank you for taking my question.
47:26
First, could you please help? US understand some of the timing
47:30
of launching FSD and additional geographies including maybe
47:35
clarifying your recent comment about China.
47:38
Thank you. I mean like new markets, yeah,
47:43
we are. There are a bunch of marks where
47:45
we don't currently sell cars that we should be selling cars
47:51
in. We'll see some acceleration of
47:54
that. And FSD new.
47:58
Markets, yeah. So, so the thing about the the
48:02
the old, the the end to end neural net based autonomy is
48:07
that just like a human, it actually works pretty well
48:11
without modification in almost any market.
48:15
So we plan on with the approval of the regulators releasing it
48:22
as a supervised autonomy system in any market that work that
48:28
where we can get regulatory approval for that, which we
48:31
think includes China. So yeah, it's, it's a, you know,
48:39
just like a human, you can go rent a car in a foreign country
48:41
and you can drive pretty well. Obviously if you're, if you live
48:45
in that country, you'll drive better.
48:46
And so we, you know, we'll make the car drive better in in these
48:51
other countries with country specific training, but it can
48:55
drive quite well almost everywhere.
48:57
Yeah. The basics of driving are
48:59
basically the same everywhere, like in a car Is a car is a
49:02
traffic light or is there a nice?
49:03
Yeah, it understands that it shouldn't hit things no matter
49:06
where. It.
49:06
Is exactly there are some Rd. rules that you need to follow
49:10
and in China it's you shouldn't crossover a solid line to do a
49:12
lane change. US is a recommendation I think.
49:18
China, yeah, you get fine heavily if you do that.
49:21
We have to do some adoptions, but it's mostly smaller
49:24
adoptions, not like the entire change of stack or something
49:26
like that. Yeah.
49:32
Hey, George, do you have a follow up?
49:33
Yep. So my follow up has to do with
49:36
the first quarter deliveries. And I'm curious as to whether or
49:40
not you feel that supply constraints that you mentioned
49:43
throughout the release impacted the results.
49:45
And if you can, can you help us quantify that?
49:47
And is that why you have some confidence in unit growth in
49:51
2024? Yeah.
49:54
I think we did cover this a little bit in the opening
49:57
remarks to, you know, Q1 had a lot of different things which
50:00
were happening. You know, seasonality was a big
50:04
one, continued impression, you know, pressure from the
50:08
macroeconomic environment. We had attacks at our factory,
50:13
we had Red Sea attacks, we're ramping Model 3, we're ramping
50:17
model cyber truck. All these things are happening.
50:22
I mean, it almost feels like a culmination of all those
50:25
activities in a constraint period.
50:28
And that gives us that confidence that, hey, we don't
50:31
expect these things to recur. Yeah.
50:35
We think Q2 will be a lot better, yeah.
50:38
It's just one thing after another.
50:40
Our side of towns are crazy. Yeah, exactly.
50:43
It's just if you've got cars that are sitting on ships, they
50:46
obviously cannot be delivered to people.
50:49
And if you've got the, you know, excess demand for one for, you
50:55
know, Model 3 and Model Y in one market, but you don't have it
50:57
there, it's like it's, it's a quite a, it's, it's extremely
51:01
complex logistics situation. So you know that.
51:07
And I'd say also the we, we, we, we did over complicate the sales
51:12
process, which we've just in the past a week or so have greatly
51:18
simplified. So the IT just, it just, it
51:21
became far too complex to buy a Tesla, whereas it should just be
51:26
you can buy the card and under a minute.
51:31
So we're we're getting back to the you can buy a Tesla in under
51:34
a minute interface from what was quite complex.
51:42
OK, thank you. Let's go to Colin Rush from
51:44
Oppenheimer. Colin, go ahead and unmute,
51:48
please. Thanks so much, guys.
51:51
You know, given, you know, the pursuit of Tesla really is the
51:54
leader in AI for the physical world and your comments around
51:58
distributed inference, can you talk about what that approach
52:01
is, is unlocking beyond what's happening in the vehicle right
52:03
now? So do you want to say this one?
52:12
Yeah, I cannot mention like the car even when it's a full robot
52:16
taxi that's probably going to be used on 50 hours a week.
52:19
That's. That's my guess.
52:20
Like a third of the hours of the week, yeah.
52:22
It could be more or less, but then there's certainly going to
52:24
be some hours left for charging and cleaning and maintenance.
52:27
In that world, you can do a lot of other workloads.
52:31
Even right now we are seeing for example, these LLM companies
52:33
have this like batch workloads where they send a bunch of
52:35
documents and those are run through pretty large neural
52:39
networks, take a lot of, you know, compute to chunk through
52:44
those workloads. And now that we have already
52:48
paid for this compute in these cars, it might be wise to use
52:51
them and not let them be idle. Be like buying a lot of
52:54
expensive machinery and letting them be idle.
52:57
Like we don't want that. People want to use the computer
52:59
as much as possible and close to like basically 100% of the time
53:02
make effective use of it. That's right.
53:05
I think it's analogous to Amazon Web Services where, you know,
53:09
people didn't expect that AWS would be the most valuable part
53:13
of Amazon when it started out as a bookstore.
53:16
So that was on nobody's radar, but they they found that they
53:21
had excess compute because the compute needs would would spike
53:27
to extreme levels for brief periods of the year and then
53:29
they had idle compute for the rest of the year.
53:31
So then what should they do with pull that excess compute for the
53:34
rest of the year that's kind of monetize it.
53:38
Yeah, monetize it. So it seems like kind of a no
53:41
brainer to say, OK, if we've got millions and then 10s of
53:44
millions of vehicles out there where the computers are idle
53:50
most of the time, that we might as well have them do something
53:53
useful. Yeah, exactly.
53:56
And then I mean, if you get like to the 100 million vehicle
54:00
level, which I think we will at some point get to then and
54:04
you've got a kilowatt of usable compute and maybe you're on
54:08
hardware 6 or 7. By that time.
54:13
Then you really, I think you could have on the order of 100
54:16
gigawatts of usable compute, which which might be more than
54:20
anyone, more than any company, probably more than any company.
54:25
Yeah, probably because it takes a lot of intelligence to drive
54:27
the car anyway. And when when it's not driving
54:29
the car, you just put this intelligence to other users,
54:32
just solving the scientific problems.
54:34
Like a human. Or Anthony Dump.
54:36
Actually, someone else. We've already learned a lot
54:39
about deploying workloads to these nodes, yeah.
54:45
And unlike laptops and our cell phones, it is totally under test
54:48
of control, so it's easier to distribute the work product
54:52
across different nodes as opposed to, you know, asking
54:54
users for permission on their own cell phones would be very
54:56
tedious. Well, you, you just drain the
54:58
battery on the Yeah. Exactly.
54:59
The battery is also limited. So like technically I suppose
55:02
like Apple would have the most amount of distributed compute,
55:04
but but you can't use it because you can't get the, you can't
55:07
just run the phone at full power and drain the battery.
55:12
So the whereas for the car, even if you're a kilowatt level
55:19
reference computer, which is crazy power compared to a phone,
55:24
you know, if you've got a, if you're a 60 kWh pack, that's
55:28
it's still not not a big deal to run, right.
55:31
If you plugged in, whether you plugged in or not, you could be
55:33
plugged in or not plugged in. You know, you could run for 10
55:37
hours and use 10 kilowatt hours if you're kilowatt of compute
55:42
per hour. Yeah, we're together built in
55:44
liquid coal thermal management. Yeah, it's exactly.
55:47
That's for data centers. So it's already there in the
55:49
car. Exactly.
55:49
So it's distributed power generation.
55:52
It's just distributed access to power and distributed cooling,
55:56
and it's already paid for. Yeah, I mean that that
56:01
distributed power and cooling, people underestimate that costs
56:05
a lot of money, really big. Data.
56:06
It does, yeah. And the CapEx is shared by the
56:08
entire world, yes, sort of. Everyone owns a small chunk and
56:11
they get a small profit out of it maybe.
56:13
Yeah. Thanks so much guys.
56:16
And just my follow up is a little bit more mundane looking
56:18
at the the 4680 ramp. Can you talk about how?
56:20
Close you are. To target yields and when you
56:22
might start to accelerate incremental capacity expansions
56:26
on on that on that technology. You know, I, we, we're making
56:35
good progress on that, but I don't, I don't think it's super
56:40
important for at least the near term as, as, as Laura said, we,
56:48
we think it'll be, you know, exceed the competitiveness of
56:51
suppliers by the end of this year and then we'll continue to
56:56
improve. Yeah.
56:58
I mean, I think it's important to note also that like the ramp
57:00
right now is relevant to the Cybertruck ramp.
57:03
And so like we're not going to just randomly build 4680's once
57:07
we have a place to put them. And so we're going to, you know,
57:10
make sure we're prudent about that, but we also have a lot of
57:13
investments with all our cell suppliers and vendors.
57:15
They're great partners and they, you know, have done great
57:17
development work without us. And a lot of the advancements in
57:21
technology and chemistry we found 4680 they're also putting
57:24
into their cells. Yeah.
57:27
I mean, a big part of the 4680 it tells us during internal
57:30
cells was a hedge against what how what would happen without
57:33
suppliers because for a while there it was very difficult
57:37
because every big car maker put in massive battery orders.
57:42
And so the price is the price per kWh of of battery of lithium
57:45
ion batteries went to crazy numbers to crazy levels.
57:48
Bonkers. Yeah, just bonkers.
57:50
So like, OK, we've got to have some hedge here to deal with,
57:56
you know, cost per kilowatt hours numbers that were double
58:00
what we anticipated. If we have an internal cell
58:05
production, then we have that hedge against demand shocks.
58:12
There was too much demand. That's that's really the way to
58:17
think about it. It's not like we want to take on
58:19
a whole bunch of problems that just for the hell of it.
58:24
We we did this health program in order to address the the crazy
58:30
increase in cost per kWh from our suppliers due to gigantic
58:35
orders placed by every car maker on earth.
58:39
So. OK.
58:41
Thank you. And the last question comes from
58:43
Ben Kalo from Baird. Ben, go ahead and unmute and
58:54
you're still muted. Well, I once again would just
58:59
like to strongly recommend that anyone who is I guess thinking
59:04
about the Tesla stock should really drive FSD 12.3.
59:08
You'd really you, you can't. It's impossible to understand
59:12
the company if you do not do this.
59:17
All right, so since Ben is not unmuting, let's try Shreyas
59:23
Patel from Wolf Research. Final question.
59:32
Oh hey, thanks so much. Just.
59:37
You know, Elon, during the Investor day last year, you
59:40
mentioned that auto COGS per unit for the next Gen. vehicle
59:43
would decline by 50% versus the current three.
59:46
And Yi think that was implying something around $20,000 of
59:49
COGS, about 1/3 of that was coming from the unbox
59:52
manufacturing process. But I'm curious if you see an
59:55
opportunity that the other some of the other some of the other
59:59
drivers around powertrain cost reduction or material cost
1:00:01
savings, would those be largely transferable?
1:00:04
To some of the new products that you're that you're now talking
1:00:07
about, about introducing. Yeah, yeah, sure.
1:00:12
I mean, in short, yes, I mean, like, you know, unbox
1:00:14
manufacturing method is certainly great in
1:00:16
revolutionary, but with with it comes some risk because, you
1:00:18
know, it's new production lines, not but all the subsystems we
1:00:20
developed, whether it was powertrains, BAT, you know,
1:00:24
drive units, battery improvements in manufacturing
1:00:27
and automation, thermal systems, seating, integration of interior
1:00:31
components and reduction of LV controllers, all that's
1:00:34
transferable. And that's what we're doing, you
1:00:37
know, trying to get it into our products as fast as possible.
1:00:40
And so so I, yeah, I that engineering work, we're not
1:00:43
trying to just throw it away and, and put it, you know, in in
1:00:46
a coffin. We're going to take it and and
1:00:48
utilize it and utilize it to the best advantage of of the cars we
1:00:52
make and the future cars we make.
1:00:55
OK, great. And then just on, on that topic
1:00:58
of 4680 cells, I, I know you, you mentioned it, you really
1:01:02
thought of it more as like a hedge against, against, you
1:01:05
know, rising battery costs from other OEMs. But it seems, you
1:01:09
know, even today, you know, it, it, it seemed like you would
1:01:13
have a cost advantage against some of those other automakers.
1:01:16
And I'm wondering, you know, given the rationalizing of your
1:01:20
vehicle manufacturing plans that, that you're talking about
1:01:22
now, if there's an opportunity to maybe you know, convert the
1:01:26
4680 cells and, and, and maybe sell those to other automakers
1:01:31
and really generate an additional revenue stream.
1:01:33
I'm just curious if you have any thoughts about that.
1:01:36
Great. What what seems to be happening
1:01:37
is that the unless I'm missing something, the orders for
1:01:41
batteries from other water makers have declined
1:01:44
dramatically. So we're we're seeing much more
1:01:49
competitive prices for sales from our suppliers dramatically
1:01:56
more competitive than in the past.
1:01:58
It is clear that a lot of our suppliers have excess capacity.
1:02:04
Yeah. One, in addition to what Elon
1:02:06
said, this is currently all the way in addition to what Elon
1:02:08
said about 4680. What 4680 did for us from a
1:02:11
supply chain perspective was help us understand the supply
1:02:13
chain that's upstream of our cell suppliers.
1:02:15
So a lot of the deals that we had struck for 4680, we can also
1:02:19
supply those materials to our partners.
1:02:21
Help. Reducing the overall cost back
1:02:24
to Tesla. So we're we're basically
1:02:26
inserting ourselves in the upstream supply chain by doing
1:02:29
that. So that's also been beneficial
1:02:31
in reducing the overall pricing in addition to the excess
1:02:34
capacity that these suppliers. Have yeah, Now, I mean, this is
1:02:40
going to wax and wane, obviously.
1:02:41
So, you know, there's, there's going to be a boom and bust in,
1:02:46
in battery cell production, yeah, where production exceeds
1:02:50
supply and then supply exceeds production and back and forth,
1:02:54
kind of like on a DRAM or something.
1:02:59
But yeah, so, so it's like what what is true today will not be
1:03:03
true in the in the future. There's going to be somewhat of
1:03:06
a boom and bust cycle here. And then there are additional
1:03:10
complications by, you know, with government incentives like the
1:03:15
the Inflation Reduction Act, the IRA, Victoria's found like a
1:03:20
funny name for a found. My cool name.
1:03:22
Yeah, is it like the Irish Republican Army, the Internet
1:03:25
research agency from Russia independent retirement camp?
1:03:28
Yeah, exactly. Roth IRA and there's like 4
1:03:34
spider man situations which I which IRA wins.
1:03:41
So, but it does, it does complicate the incentive
1:03:45
structure so that there's, there, there is perhaps there's
1:03:52
the stronger demand for cells that are produced in the US than
1:03:55
outside the US, But then how long does that the IRA last, I
1:04:02
don't know. Which is why it's important that
1:04:05
we have both inland cells and other cells that hedge against
1:04:08
all of this. Yeah.
1:04:11
OK, thank you very much. That's all the time we have
1:04:13
today. But at the same time, I would
1:04:16
like to make a short announcement and I wanted to let
1:04:20
the investment community know that about a month ago, I met up
1:04:24
with Elon and Vaibhav and announced that I'll be moving on
1:04:28
from the world of investor relations.
1:04:31
I'll be hanging around for another couple of months or so,
1:04:33
so feel free to reach out anytime.
1:04:35
But after the seven-year Sprint, I'm going to be taking a break
1:04:39
and spending some good quality time with my family.
1:04:42
And I wanted to say that these seven years have been the
1:04:45
greatest privilege of my professional life.
1:04:47
I'll never forget the memories from I started literally at the
1:04:51
beginning of production hell and just watching the company from
1:04:54
the inside to see what it's become today.
1:04:58
And I'm especially super thankful to the people in this
1:05:01
room and dozens of people outside of this room that I've
1:05:05
worked for over the years. I think the the team strength
1:05:09
and teamwork at Tesla is unlike anything else I've seen in my
1:05:13
career. Elon, thank you very much for
1:05:15
this opportunity that I got back in 2017.
1:05:19
Thank you for seeking investor feedback and regularly and
1:05:22
debating it with me. Yeah, Well, I mean, the, the
1:05:25
reason I I reached out to you is because I thought your analysis
1:05:28
of Tesla was the best that I'd seen.
1:05:30
Thank you. So yeah, thank you for helping
1:05:33
Telsey get to where it is today. Over 7 years it's been a
1:05:36
pleasure working with you. Thank you so much.
1:05:38
And yeah, thank you for all the thousands of shareholders that
1:05:42
we've met over the the years and walked around factories and
1:05:46
loved all the interactions even with the even the tough ones.
1:05:49
And yeah, looking forward to the call the next three months, but
1:05:53
I'll be on the the other side listening in.
1:05:55
Thank you very much. Thanks.