Elon Musk Reveals Tesla Optimus Production and 2027 Delivery Plans
Publicado no IA Dots ·
AI Signaldot-artificial_intelligencedot-humanoid_robotsidioma-enInglêsRobótica e humanoidesyt-wrPgb5Ee_2E
Vídeo em Inglês, publicado pelo canal DPCcars.
About this content: The summary, key points and topics below are generated automatically from the video transcript. They reflect the content, opinions, claims and predictions presented in the video and do not represent a position, validation or fact-check by IA Dots.
Video summary
The video discusses the production, development timeline, and technological challenges of the Tesla Optimus humanoid robot. It covers manufacturing ramp-up in Fremont, AI development, self-play training, the integration of the upcoming AI5 chip, and expected commercial availability for industrial applications.
Key points
- Optimus production follows an S-curve growth pattern with an initial shallow and bumpy slope due to the complexity of the hardware and supply chain.
- Reaching a high-volume annualized rate of one million units per line is expected to take at least a year or slightly longer.
- Unlike cars, Optimus cannot rely on an existing vast fleet for training data, requiring bootstrapping through self-play and the Optimus Academy.
- The integration of the Tesla AI5 chip into Optimus is targeted for approximately April of next year, providing an order of magnitude more compute.
- Initial deliveries of Optimus to external customers are expected around the end of next year, focusing primarily on industrial, dangerous, or repetitive tasks.
Topics
Optimus humanoid robotTesla manufacturingAI training and self-playAI5 chip integrationIndustrial applicationsProduction timeline
Video transcript
Transcript from available captions; it may contain errors.
Regarding Optimus, we we've started like what I call early production on our high volume line in Fremont. Now, I want to elaborate on this point because the And I've said this before that the the growth of Optimus it's it's an S curve, but the initial part of the S curve is has a shallow slope and is somewhat bumpy because it is such a radically new product. So, um you know, the way to think about it is there's like 10,000 things that have to go right for production of the of the robot and it'll move as fast as the slowest least lucky of its of its 10,000. Um but we do have um the the you know, the the vast majority of a a million unit per line production line um landed in Fremont right now. Um but it it I would expect that it takes us probably at least a year, maybe a little longer to get to that annualized rate. Get to 20,000 units a week. Um so that this is just a fundamentally because there's you know, there's there's still parts of the design that are changing. Um like a lot of people ask me about Optimus. That's why I think I've got to kind of um sort of belabor this point. Um because uh peop- if if people aren't don't don't sort of understand like new technology that well or volume production of complex new hardware that well. It sort of seems like, "Well, can't I just make a whole bunch of these things immediately and and send them out and and it's like, "No, you've you've got to uh basically build up an entire gigantic global supply chain and uh have domestically um a a very very big factory and everything's got to work. Um and there's still some things we we need to tweak about the design because you you to deliver a robot that is that is safe and reliable and has long duration. So it can't like wear out after a month. Um and and there's still quite a lot of work to do with regard to the AI in the robot um for it to be truly useful for open-ended tasks. Um But if if but like I said, if if I I would I would I would expect that the the AI, the you know, the electromechanical design and the um reliability and and and and and uh duration usage duration, like the the usage life of the robot, I expect these I do expect these issues to be solved around the end of next year. Um so that that's the time to ex- to expect um when we're delivering Optimus uh uh to companies outside of Tesla. Um And and I I remain convinced that Optimus will be the biggest product on Earth. Um And and humanoids humanoid robots in general will be the biggest product on Earth. So so it's it's a Optimus is a very big deal, um but we should I don't want to set your expectations that we we still have a lot to get through and it's about a about a year or so that it'll take us to get through all that stuff. Um But I as as to the best of my knowledge, there is no one that is even close to Tesla with respect to a general purpose humanoid robot. >> What are your thoughts on crowd sourcing training data for Optimus versus your internal Optimus Academy approach? Is data currently the primary primary bottleneck? I think you covered this quite a bit, but any other color there? >> Mostly uh talked about this already, but um It it's it's not crowd sourcing is is the is is not the issue for Optimus. Um There's a lot of things we have to get right about the um AI architecture um uh that's what we're still working on. We're making good progress uh but but uh but that that needs to be sorted out um and really in order for Optimus to and get the training data we need, we actually need to make a lot of Optimi. So, we need we need to make I think on the order of 10 or 20,000 um bots and and then have them do self-play and and we're building up an Optimus Academy where Optimus can practice doing a a very wide range of tasks um and then we can use that data uh in in a in sort of a flywheel to make it better and better at these various tasks. Um the you know the a big difference with Optimus versus the cars is we don't we don't have um this vast fleet of cars from which we can gather training data. So, with Optimus we we we have to bootstrap it by by making the robots. Um and and having them do self-play, which is really how a human learns to navigate the world. You know, it's that's uh you can you can see this when you have a toddler or and they grow up and they they initially are sort of wobbly and they don't know how to do things and they eventually they um you know they learn to wipe their own ass, which is a big milestone. Um >> [laughter] >> as anyone who's had kids knows. Um and uh and then they learn to read and and do all sorts of things and and be able to to do do these things and they learn this by by playing um and we actually need Optimus to play at scale in order to uh have it be really good at at uh tasks in the real world. And I should ask you if you want to add to that. >> Yeah, um like I mentioned Elon, for the car we actually have the data from the car's embodiment itself, but for Optimus there's no exoskeleton for humans to get into and like make Optimus into a puppet or something like that. Like we have to learn uh Optimus has to learn by watching other people do tasks initially and then start practicing the task yourself. For example, you can For example, if you're watching a soccer game, just because you're watching it doesn't mean you'll be good at soccer yourself. You got to like start practicing and get good at the game and learn from the bot's own experience. And we are setting up this huge reinforcement learning cycle with initial base training done from humans to bootstrap and then put it on Optimus and do a lot of self-play both in simulation and in the real world to learn on the embodiment of Optimus itself. I think we're pretty close to having very good results in this big loop. >> Yeah, I think we we may be ready to release like an Optimus video um in a month or so. Um I want to make sure that like when we release this video that it's not like a canned demo that um you know, like when you see these robot humanoid robot uh demonstrations, some of them look very impressive, but they're they're they're not really using uh any kind of advanced AI or SI. I mean, they're they're like pre-programmed deterministic tasks. Um and if you if you actually try to use that humanoid robot in the real world, you would have a lot of trouble. It it wouldn't work, basically. So, I want to release a video that's where it's Optimus doing things that um that that are not pre-programmed. Um and I and I think that's that that that would be it's maybe a month or two for that um to to have a a demo that you can believe what you see type of thing. Um and and then uh there's there's another milestone that's I think important we want to get to, which is to integrate our the the Tesla AI 5 uh chip into Optimus, which if if things go according to plan, will be in approximately April of next year. Um and that that's a giant improvement in capability with an order of magnitude more um uh sort of more operations per second. Like it it's an order of magnitude more compute AI5 versus AI4. Um and that will that will help Optimus a lot because the the degrees of freedom that a humanoid robot has to deal with are far greater than what a car has to deal with. So, it is a bit of a struggle to squeeze uh the you know, everything needed to operate with um all the degrees of freedom of Optimus um and and and the wide range of tasks, you know, can you think of like a humanoid robot is going to do like a thousand times more things than a car would do. Uh or that's some very big number basically. So, so I think the additional compute of AI5 will will make a big difference um and that comes in in and and hopefully in volume around April of next year. Um and uh and that that's why I think like all the pieces like the the the refinements in the electromechanical dexterity, um making sure Optimus can operate for long periods of time without wearing out, um you know, getting the AI software right uh and and the AI5 chip integrated, all of these things kind of come together. Um that's why I think about a year from now is when it's something that would be that that that you could ship to customers. Um and it would also be reaching scale in production. So, that's that's that's the probably the right time to think about um And in terms of the the initial customers for Optimus, I would I think probably would be things that are industrial applications at first. Um things that that are like like uh dangerous repetitive tasks, um things that people really don't want to do. So, [snorts] and and and that that you want to done basically 24/7. So, like around the clock stuff. That's going to be the highest value for Optimus in the beginning. Um as we scale production, of course, it would be something that people can buy, um you know, for for their home. Uh but the initial high value use will be the highest value use, I think, will be uh kind of heavy industry, dangerous tasks, um or things that may cause repetitive stress injuries, um and and and that need you want done, um 24/7. So, that's uh uh yeah, the way to the way to think about it.
Assistir ao vídeo original no YouTube
Assuntos: Robótica e humanoides, Robótica e humanoides. Publicado no YouTube em 2026-10-22.
Seleção automática por assunto e idioma. O vídeo pertence ao canal de origem.
About Tesla Optimus
Optimus is Tesla's humanoid robot project. On its AI & Robotics page, Tesla describes the goal as a general-purpose, bipedal, autonomous humanoid robot capable of performing unsafe, repetitive or boring tasks, which requires software for balance, navigation, perception and interaction with the physical world.
According to the video's description, Elon Musk gave the update during Tesla's third-quarter 2026 results: initial production of Optimus is underway at the Fremont factory in California, customer deliveries are targeted for 2027, and the long-term plan is to reach about 20,000 robots per week after at least one year of scaling.
Fact sheet
- Robot: Tesla Optimus
- Company: Tesla
- Type: General-purpose bipedal humanoid robot
- Factory cited in the description: Fremont, California
- Delivery target cited in the description: 2027
- Official page: Tesla – AI & Robotics
Read also
- Elon Musk Just Revealed Why Tesla Optimus Was Delayed
- Elon Musk Reveals the BIG Problem Holding Back Tesla Optimus
- Google Gemini Robotics 2: AI Gets a Body — Here’s What It Can Do
Sources and references: Tesla – AI & Robotics · YouTube description of the video by the DPCcars channel.