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