Physical AI involves putting intelligence on physical machines like cars, trucks, and robots, with the goal of making autonomous systems accessible to everyone, including high school students. Unlike digital AI which operates in software, physical AI faces unique challenges including real-world safety constraints, hardware limitations, and the need for proprietary data collection. The key to democratizing physical AI is through platforms like Dana that provide tools for designing and developing autonomous systems, lowering the barrier to entry and enabling broader innovation across industries from agriculture to defense.
Install our extension to search inside any video instantly.
Why Physical AI Is the Next Frontier | The a16z Show
Added:Our mission is to put intelligence on a billion machines and we think that can have a profound impact on society.
Applied intuition is a physical AI company. We put intelligence on machines, cars, trucks, tanks, drones.
It's a physical moving thing. We make it intelligent.
>> Digital AI, of course, is building software and optimizing ads and creating videos. That's all interesting and good, but really where you talk about the global economy, that's physical AI. In this intelligence revolution, the companies that impact the physical world might actually be bigger than the companies that impact the digital world.
>> How many things are there where the idea of physical AI, physical intelligence are going to matter?
>> There's no reason autonomy should be this obscure, difficult technology. Our vision for that is a high school kid that can make iPhone apps should be able to make autonomous systems. That platform for designing and developing is what we're launching. It's called Dana.
Everything that we've built and developed over the past nearly a decade that's available in Dana.
>> Which will we get first? A perfectly simulated real world environment for training autonomous devices or Grand Theft Auto 6.
>> Kazer Peter welcome to the Z podcast.
>> Well, thanks for having us. Your name is >> just one of many.
>> I feel like we've I think we've all each known each other for >> too long. more than I'd like to admit.
>> Yeah. Long time. We're lucky to both be the first investors or among the first investor in the first round. Of course, different check sizes, but >> and I was an investor for you even before then.
>> Exactly.
>> Um, so let's do that as a segue. We have a lot to talk about today. We've, you know, the biggest launch in company history to talk about today, but first, why don't we just give an update or a status what does applied intuition do for those who >> Yeah, for the for the people who don't know, applied intuition is a physical AI company. We put intelligence on machines. That's the the simple way of uh describing it. um and all types of machines. So cars, trucks, tanks, drones, you name it. It's a physical moving thing. We we make an intelligent and uh the history of the company is we originally started by making the tools that would make the int then we got into the actual intelligence itself. Um in a very like in some ways like a very uh boring AI company in the sense of you know 83% of the company is engineering.
We win by making really great products.
It's not like a good sales or something like that. I don't think we're good enough for for a for a sales enabled company but uh yeah over a thousand engineers um and based in Silicon Valley but we have offices globally 18 offices and uh you know our mission is to to put intelligence on a billion machines and uh we think that can have a profound impact on society both in the kind of pissy things everyone talks about safety you know if you really talk to somebody who's been in a car accident or in a mining accident or you know uh in a farming accident. Those are real gnarly situations. Beyond just fixing that, you just if you can unlock productivity, I think, you know, we've seen the unlock in the digital world and everyone's super excited about it and you have trillion dollar companies emerging. I'm a pretty strong believer that I think when we look back 25 years, if you look back at the internet now, you know, people actually, if you look at the original internet companies that are doing, you know, serving or they're they're doing some analytics and those are interesting, but really when you look back 25 years from now, the big monolithic companies are Amazon that delivers you stuff, you know, uh Apple, these are the true kind of companies that come of age. And I think when we look back 25 years in this intelligence revolution, the companies that impact the physical world, you know, might actually be bigger than the companies that impact the digital world.
>> I would love for you to talk about the following, which is when you first started the company, you know, the knock on the company, I think, was oh well, it's it's like, you know, car it's making making cars autonomous, right?
Self-driving cars. But it's kind of like, okay, there's like whatever, >> you know, there's there Tesla and way more building their own self-driving cars, and then and then there's like six or eight other car companies that matter, and then the company just could never get that big because there just aren't that many customers.
>> Yeah.
>> So, What's the like how should people think about like how many things are there that are things that move where the idea of physical AI physical intelligence are going to matter?
>> Yeah, I mean even today even if you put that you know uh let's say uh uh view on us the automotive is like 30% of our business. So 70% already is non-automotive and I think if you fast forward another 10 20 years uh even the manufacturers themselves as a customer base will be a small amount. I think our that mission just keep thinking a uh you know a billion machines becoming intelligent and you think about all the types of machines that exist. Automotive is just an easy one. I think it sticks in people's heads because we all drive cars and it's a big market. Um but I think it's it'll be a minority of of the business. I mean it is a minority business. I think it'll be increasingly a minority of the business but that doesn't necessarily mean it'll be small, right?
>> Automotive is still huge just as a part of the globe's GDP. Automotive is something like 3% of all GDP. Um I think the way we always think about it like as you try to get to your mission initially the manufacturers were the distribution to that intelligence to consumers but then you start working in defense and you start working in construction and mining and agriculture and suddenly the manufacturers are important but maybe the mining operator is actually really important or the department of war is really important and suddenly they become customers and all of those are customers of ours as well.
>> Yeah. I think if you split AI into digital AI and physical AI, right?
Digital AI of course is building software and optimizing ads and creating videos, that sort of thing. Uh that's all interesting and good, but really where you talk about the global economy, that's physical AI. And then we're talking about manufacturing and uh and mining and logistics and transportation, all of these things that >> supply chains.
>> Yeah. Supply chain. Exactly. Well, I mean, let's build on that though for a second, which is like so things that move today or you know, historically things that move are things that have human beings at the wheel or at the controls in some form, right? Um, and so and you know, airplanes have had to get designed around a human in the cockpit.
Uh, boats have had to, you know, get designed around a human, you know, steering things.
>> Um, like in a world where in a world of autonomy like do we already know what the things are that move or are we going to discover that there are a lot of new things that are going to get built uh when you don't need a human in the in the driver's seat? I think both. Uh be well the the the the thing that you have to remember is like you take like a a holage system that's on a in a in a port um like a catap kamasu the dirt mover uh in a mine those are made for 20 25 years.
>> So the buyers of that of those products they're they might not have gotten their full cycle you know uh uh ROI on them.
So they're not immediately going to buy something new no matter how much better it is. So one part of our strategy is you got to make those things intelligence because they're not going anywhere. The second is what you're talking about which is well that depends on a human in a cab. If you don't have a human in cab you can run the the machine can be smaller. It can be shaped in very different ways. You talk about mining underground, >> right?
>> The constraint actually is the human because the human needs to breathe and it's very dangerous and so you can build a very very different machine. We're doing both of those things. And um and then the thing that we're not talking about is we're all talking about intelligence almost like within a system but the system level intelligence is where the unlock is and we're already doing work like that where you say hey let's take an entire port let's take an entire mind let's take an entire query and this heterogeneous mix of machines they're all can talk to each other and they can optimize and be efficient when one machine goes down or one machine has an issue the rest of the mind doesn't have to stop when it's human-driven.
event we don't even know the machine's going to go down because there's no analysis the human is not plugged into the core uh systems of the machine. So like a simple thing like knowing when a break system is going to break >> is actually huge because you can start preparing for it in advance. Oh this wear and tear is higher than in other mines. Just using an example but like the the other macro point is if you look at agriculture as an example you know average American farmers 58 years old the there's that the number something like under 35 it's like less than 10% of farmers are that young. So what's going to happen? the need for food growth is continuing to grow. The need for uh rare earth materials is contin so these these demands are only growing but the humans who are the bottleneck are decreasing.
>> Trucking is the same way. Um and so you can you can really just unlock a lot more efficiency. So I mean one way to think maybe think about this is like imagine if the cost for food decreases because it's way way more efficient.
What what's the downstream impact? the imagine for goods being transported let's say you know instead of a few dollars a mile it's 20 cents a mile and suddenly it's it's I think that the unlock is very very very big um and that isn't ne I think doesn't necessarily need for all the machines to be redesigned from the ground up >> right right got it makes sense and then maybe just one one more question would be just give give it give us a sense of parameterized like the scope and scale of a company today >> yeah uh uh north of a thousand engineers and those engineers are you know obviously the classic you know uh uh uh software and AI uh engineering teams, but we also have engineers who really know safety systems. Uh we also have engineers who really know hardware because the important thing that we we're kind of just uh tipping around stepping around is all this stuff is hard because it ultimately has to meet the real world and the real world is has way more complexity and and and has a lot more issues and we have engineering teams that can I mean we've deployed our our our models onto like 50ome platforms like even that sounds trivial because when you think about mostly when you think about models you think about deploying them through a browser or on a phone and everything's abstracted away because you of iOS and you have Android and you have Windows and you have Linux and you have all these systems that have already taken care in the real world you don't have that and so we've we have engineering teams that can do that as well our kind of a you know claim to fame as we've uh raised over about a billion dollars in the company's history all that is sitting in the bank and uh I think uh that's I always say that with an astrog doesn't mean we're not going to spend it you know next uh next month >> good news bad news >> yeah good news bad news uh it's uh but it's the the uh and I Like we're at that phase, you know, you you talk about scale, we're at that phase where these giant markets are around us and we can make the decision how aggressive do we want to pursue those because decade of frankly execution and deployment into production. I think the the hallmark of our engineering team is is putting you know products into production that that that really is is a big I don't know how do you think about scale? Yeah, I think that that's roughly I mean the mission of bringing intelligence to a billion machines that is how we think about it and then thinking about well what are the types of machines that we'll have the most impact on and focusing on those areas first but uh we'll get there >> right good I uh let's go deeper into the differences between digital and physical AI and more so into where where are we today what progress has has made what are some of the main major bottlenecks in physical AI when you unpack some of that >> yeah I mean I think a lot of times people think about the progress in physical AI is limited to basically two case use cases and they're just cuz they're obvious and interesting which is robo taxes and humanoids. Um they're very visceral. They're they're they excite you and they're kind of sci-fi.
Um I think they're those are very interesting. Uh and they there is real work being done by us and other people in in those domains. I think all the other uh all the other domains I think are going to be just as important. I mean, you just you just think about what happens on a port. That's there's a huge unlock there. Um, and that's I think that's the that's the area we're we're we're really focused on. It's like all the other nooks and crannies. If you look at like we we've talked before about uh the rise of Cisco and how you know networking kind of went from you know first individual machines and companies would get network and then entire countries are getting network.
There's a similar thing happening with AI. AI's getting to that level of kind of sovereign AI is now a discussion.
Sovereign AI really is about physical AI because that's where you're talking about AI in defense. You're talking about AI in the physical machines that are moving around. If you look just at the example of of uh Whimo from America and Pony from China trying to deploy in let's say the other countries, so not America, not Europe, not China. every one of those uh spaces, they're way more uh they're way more hesitant of saying, "Yeah, thumbs up. Your robo taxis can run unfettered on our in our uh country." And so, if you look back just at kind of this arc of the internet, you know, when the first internet companies come, nobody's really think about sovereignty at all. It's like the browser goes everywhere, the internet goes everywhere. That's almost the power of it. Then when social media emerges there's a bit more of hey actually not every social media and then you have China not allowing Facebook to come in and you have some then you get into the next level of like the online offline stuff there's more resistance the Ubers the door dashes suddenly there's local players who are being favored very aggressively when we get to physical AI I think there's going to be huge and also there's like a larger ge you know geopolitical theme of of kind of uh more fracturing than globalization you're going to have this demand and for this AI should somehow be localized and I think that has to play into our strategy as well. We're a technology provider so we can provide that technology across the globe. Um and I think that's that's that's uh that's that's that's something that's understated in this in this uh conversation.
>> Yeah. A few other things on on digital versus physical AI. So in digital AI, the state-of-the-art is you you can train models effectively on the entirety of the internet and then maybe augment that with additional data that's been collected and refined with some hired experts. Right? This is sort of a hot field right now. Uh but generally you're talking about a foundation model that's built on internet data. In physical AI, the internet data is is useful too.
However, to actually build a foundation model in physical AI, there's also a lot of private data collection. you when we're talking about mines or or logistics or any of these other uh other fields, the data that's useful for training models there is not necessarily available. So we have to do a lot of work ourselves actually going going out and collecting that data. And then the other key factor is safety, right? Uh if you're talking about uh building a smartphone app, you don't necessarily care about uh is is a safety critical application. But when you're talking about moving a a machine that weighs many tons, uh or think of a humanoid which could fall over on your children, uh you care a lot about safety and and the evaluation of that safety and and that is really sort of getting to the state of the art of physical AI and and really proving out uh the safety case around some of these state-of-the-art models.
>> Yeah. And I think like you know you talk about like humanoid data collection has been its own you know uh little uh area of interest but when you talk about collecting data like in places like Korea where they have north South Korea where you have North Korea they don't allow mapping companies let alone allowing a you know a an American company to come in and data collect.
We've figured out over the years whether it's the Middle East, whether it's Latim, how to get into these countries, work with the governments and get the thumbs up to collect proprietary data.
And so in in the way that it's it is similar to you know other uh digital AI systems, your proprietary data sets, scaling laws, all that stuff is the same. It's just applied in a very very different way. And uh it's almost like the way to think about it is like the diffusion of these models is very different because you can't it's not everyone can just access them through a phone. you're and so so that ironically is actually plays in our favor because once we have a massive proprietary data set where we've been building we already have hundreds of pabytes of data um and then we have our own tools which are like you know synthetic data tools uh neural sim we can use our own tools with our own proprietary data and that allows us to build some of the best systems in the business is there's kind of a chicken and egg thing which is like in order to build an autonomous physical thing you need a lot of data to gather that data, you need a lot of physical autonomous things running around collecting the data. So, it's like once you have a giant network of physical things running around, you have the data that makes them all work like is is there is there a flywheel aspect of that? And is there is there is there like what what's the level of difficulty involved in kind of booting up that flywheel?
>> Uh it's it's it's difficult, but it's also not difficult. I I mean I think we we have one of the largest data collection fleets on the planet, frankly speaking. Um so that's how you bootstrap your way into it. That's just money and resources and technical knowledge, but it's not like there's probably more than five companies that have that technical knowledge. So, it's not extremely obscure. I think what is more difficult is then how do you actually have that model which is going to work on lots of different hardware and is, you know, is is tested appropriately because the you saw it, you know, in Cruz, right? Cruz was this company that did amazing self-driving work and then one accident, General Motors owns them and they get super scared and they pull back. So, it's like just getting these things into production is actually more difficult than than uh than it seems. Um I think like we believed synthetic data was going to be important. So, we started our synthetic data team like 5 years ago now plus yeah more than that at this point. Uh and >> like and we're a strong believer that synthetic data can accelerate autonomy development. We've just seen that and then there are like lots of other secondary and tertiary like kind of technical innovations that happen.
Obviously the transformer revolution hitting self-driving massive basically everything done in self-driving pre2122 relevant but you're almost like that's kind of the starting point but it's also different than today being the starting point like there those four five years are actually there has been a lot of work done you can see it most clearly with Tesla but there's other folks uh in that process the actual techniques historically and I'm just simplifying here imitation learning was the way the way of the which was collect a bunch of data and then the models would basically imitate what human drivers do.
>> The real state-of-the-art right now is endtoend reinforcement learning in in a closed loop in in your tools. And so >> it's a little simplified to say the system learns itself. It identifies where the issues in the self-driving system are and essentially you then find data like that or you synthetically create data like that and then you you know close that loop and you see are you performing in those same scenarios better and better. I think if you fast forward some years that will be a completely closed loop like with no humans intervening. Right now you still have like like what's the the fog error that we saw um we still see errors in the real world that impact self-driving.
>> Oh yeah. So it's like well what what are the bottlenecks right? and and the the bottlenecks. Uh there's plenty of them, but uh whenever you're dealing with physical systems, you inevitably you hit a lot of gnarly hardware problems. And uh it could be anything from overheating to uh sensor being slightly miscalibrated or a funny funny issue we saw yesterday was uh basically a fogging sensor like fog impacting a sensor. And uh but these are the things that you actually have to solve for this stuff to work very reliably in the real world.
>> Yeah. So I want to ask you a thread a question and you can we can decide whether you guys want to engage on it or not. It might be an opportunity or might hate the question which is >> were you surprised? So Cruz was a super high-flying Silicon Valley autonomy startup that was kind of running neck and neck with Tesla early on and so forth and you know very top- end team and then they famously got bought by General Motors and >> one of my first distributions personally. So I >> There we go. Y Cominator Y Com Y Cominator company um >> and um and you know a top end team and they they were you know by all accounts making excellent progress. They got bought by General Motors. They became the GM autonomy program. GM got a lot of praise at least, you know, at least in the in the in the tech in tech circles for being like, okay, being like the the legacy automaker with the biggest investment.
>> I I called Peter when before before it was announced on that. And I said, hey, Cruz just got bought, you know, he's also GM family. We're both GM families.
And uh Peter Guest was he said Nvidia? I said no. He said I said go fish is Apple. I said no. I said General explicitive motors.
>> Right. So, so that's surprising to people who are from GM >> that they were willing to buy that they did okay that they did it and then by all accounts they were I mean as far as like as far as I ever heard like they were making excellent progress uh and then they had this there was a there was an accident there was a was that was a injury or fatality or >> uh it wasn't a fatality it was a serious injury somebody was dragged for 20 ft >> yeah serious injury bad press and then and then they put a bullet they the GM CEO on board put a bullet in the cruise project and I know the that at least some of the senior cruise people were extremely upset you know the by by the aftermath of that was it surprising that they reacted the way that they did?
>> Uh so you know full disclosure General Motors is a customer and I went to the General Motors Institute so we have a lot of love for the company. Uh but incidentally and ironically I'm reading uh coincidentally I should say I'm reading this very famous book which I had actually never read before called on a clear day you can see General Motors right >> and Delorean's book. Have you read it?
>> As one does. As one does.
>> Have you read that book?
>> So I have years ago I have. It's one of the great alltime book titles. And we should just pause to say John Delorean was like what he was like the super genius of the car industry.
>> He was going to be the next president of General Motors >> of General Motors and then later on he started his own car company which like Back to the Future and then that whole thing collapsed for variety of reasons.
>> Uh but yeah he he was like a legend. He was like one of the main principal drivers of innovation of the current Bob Lutz this category and you got to remember this is this right >> but sorry repeat the title. Repeat the title of the book.
>> On a clear day you can see General Motors >> and why why was that the title of the book? because uh there's a lot of >> It's a very large complex.
>> Yeah. Complex. Yeah.
>> It's like it's like a nation state.
>> Yeah. It I mean really I mean it is like I think we say that like sometimes almost like flippantly but these companies are like extension like Hyundai is an extension of the state.
Toyota is an extension of the state.
Volkswagen is literally Volkswagen board members are members of the government.
So these are extensions of the state and and almost every uh and there used to be an old saying what's good for General Motors is good for America and you cannot underate how important General Motors is the history of the American corporation Sloan's uh my years at General Motors and Adventures white collar man if you run a large engineering organization you should read that that is the this like you this >> this thing that that we talk as a modern corporation didn't just emerge Sloan and Ketaring create ketaring is the head of engineering created this this AR, you know, with levels and vice presidents and how do you do functional and matrix organizations? It's there really is like the source code comes along John uh uh you know comes on Delorean and he says um he writes he's going to be president and he's so fed up with a company but what was controversial was GM was doing really well at the time. GM >> was like a when we say like GM was number one in the Fortune 100 it was like number one two and three. It was everything and it was seen as the best company in America. So somebody to openly criticize the company and so he has a whole he writes this book as he quits out of out of uh how annoyed he was general is being read led uh he writes this book and then after he like so up he's like I don't want that book published and so he fights for years for his co-author not to publish the book.
The co-author still publishes it. So it's a real true insight into a large corporation. I'm incidentally just reading it out even though I've you know worked at GM 20 some years ago and had you know know a lot about the company and what's shocking is it's not only about GM most of the major manufacturers actually still operate that way uh on the inside and so the question isn't the the point I think for everyone to take away isn't that these people who run these companies are stupid they're not stupid they're it's kind of like you know when you're selling to the department of war and people say well why are you doing that it's like well the distribution defines the business >> that Like so like the distribution is this is a consumer product of of this this stat might be outdated but when I worked in safety systems 20 years ago I remember GM used to pound into your head of the top five consumer lawsuits in in American history three are automotive.
We got the majority right so it's like you have to be extremely careful. But we had these like weird things like inside the company you couldn't it wasn't red, yellow, green. It was like purple or like you'd always have to decoder because you know why? Cuz when they go to lawsuits they're like you let a safety system that was marked red go to production. I was like no it was marked magenta.
Like so like can you imagine how infuriating that is? Every time you're like >> what does orange mean? Does this mean I have to like So fast forward to you're meeting that system. Well, for then the the Ford slogan for a very long time was that it was quality is job one, right?
Safety is job.
>> Yes. Yeah. Exactly. And and that's the one two punch of automotive. It's quality and safety. Quality and safety and quality really becomes because the Japanese really reset that that stage and and because that's a whole separate automotive history. We could talk about automotive history for for an hour. But the punch line is you have the Silicon Valley company meeting this immovable object.
>> There is a parallel universe that cruises out there right now. even as a part of General Motors. So, I think you always have to take it into the context of where the company is, where union negotiations are happening literally that year. And if you're the union, you're like, you can't make a billion dollars for us, but you're funding this thing that's killing people and it's sloppy. And so, I'm not saying precisely that's what happened to be very clear, but the it's a multivariate problem. My other hot take is, you know, I worked at both companies, right? Uh Google Google Motors. Those companies are way more similar than they're different.
>> Way way more similar. Literally, people don't need to know this. The Google leveling system is the same as the General Motors leveling system. And I used to say I saying, you know, this inside of Google meetings is like, hey, actually some of the engineers I knew at General Motors are better than the engineers here. And people would look at me like I'm saying there's no God in church. It's like they're like, how dare you, you metal bending monkey from Detroit.
It's like, no, actually like making a modern combustion engine is extremely complex. It's not just like, you know, it's it's it's not simple stuff. And so the the the macro point I think is it's a bunch of things. I think safety is always at the top of their top of their list. I do think, you know, we've hired lots of cruise people. I think the way they dealt with that specific issue with a government, you you got to you got to dance a particular way when that happens and they just didn't dance exactly right and that just gives government bureaucrats >> more ammo to go after. and you're a big target like General Motors, you got to you know it it reminds you guys ever see that movie like uh uh Good Fellas uh you know uh there the one of the last scenes the house of the rising sun you know all the old bosses go in the back of the courtroom and they're like and you know that's what happened. And they're like the the board was like, "What are we going to do about cruise?" Like, "What can we do?"
It's like, "Kyle's a good guy." But and then it's like, "Que the rising sun."
People running through a San Francisco office. Just kidding.
Don't Don't make that an AI video.
Gonna get a mean text from Kyle.
>> So, I think I think there is a universe that would have survived, but it's it's it's it's tough. So then a lot a lot of what applied intuition does is kind of as you said like that dance it's like how to how to be a great partner to these companies.
>> Exactly. Bearing in mind their own re very real issues and constraints. I >> I think general also had the topic of business model right. So you have uh cruise was going after the robo taxi concept but GM makes its profits from personal car ownership >> and those things can be a bit odd. So I think that was also a bit of the bit of the equation.
>> Okay.
>> Yeah. And I think it wasn't clear. I mean the by the way you know um you actually all people you spoke at YC at 20 in 2013. I was in the audience. I was a a partner at the time and um you said something which I think is it's it's very like uh recursive here. We're we're feeding each other device. It's uh the key thing in in new techn in the new technology business actually everyone kind of figures out the technology though that's still hard. It's still hard sometimes to build really complex things. It's when and how you deploy them into the market. the when becomes really important. You're two years early and you're doomed. You're two years late, there's too many competitors. You have to like hit it at the right spot.
And I think >> it's like like I mean a controversial thing to say is like I actually think Cruz, you know, they were certainly moving at a much faster pace than Whimo.
They started way behind and you're talking about neck andneck when you know ultimately the the the plug was pulled.
So who knows what happens in the long term. Our hypothesis in that same equation is actually the distribution.
You let the manufacturers do that. Like we run self-driving trucks right now in Japan. They carry commercial loads.
They're they're, you know, they're safety drivers there, but they're autonomously running. And but you won't know that because the brand is Isuzu.
>> Mhm.
>> That's the customer. And and why it's so good for us to partner with Isuzu in that case is that company's been around for almost a hundred years, right? Uh if if I'm not mistaken, pre preWorld War II company. and they are uh you know they know the government they have test tracks they know safe they know their own trucks very well so when we go and provide them with the intelligence and the integration into their physical machinery that's a fantastic onetwo punch I think today the world is ready to consume AI in the real world and that's a lot because of chat GPT and anthropic and all these you know everything that's happened so people are no longer like what's a self-driving car and it's because of whimo and Tesla >> so the market is ready to consume and And I think you just have to meet the market in the way that the best way possible. And our view of that has always been you go through some of the people who run the economy right now.
Whether it's a mining operator, whether it's a department of war, whether it's uh uh the manufacturers and we work with, you know, within each vertical uh with with the right partner. But that's a fundamentally different view than a Tesla or a Whimo which are going to be vertical >> where we're really playing the horizont.
And I think the way we can always think we think about that our companies, we're kind of like a like a chipmaker, >> you know. We we we we actually look and talk and walk a lot like a silicon company except we obviously we don't make chips but you know we're are we have design wins and then we have really large long-term relationships and then once we're in we're in it's really hard to you know take take us out. So you need deep trust our our partners have really a a lot of deep trust and we know their markets really really well. The things that Jensen knows is he knows his customers.
>> That's why Nvidia does well beyond the fact obviously they make a very complex technology. Mhm. So, how are these legacy uh car companies preparing for the future? Are they making or more acquisitions? They're going to are they building, you know, partnering with you?
Are they how are they going to compete with, you know, tech technative companies?
>> It's like saying like how are governments dealing with AI? It's such it's such a broad topic. Um and each manufacturer like even like you take Honda, Nissan, Toyota, three Japanese manufacturers with, you know, long legacies, they all approach it very differently. They're roughly in a spectrum of we're going to build to we're going to buy. Um and other both extremes more than ever we're going to buy is the common answer because they've been trying and we've been there the whole time. Uh for the folks that are going to build, we provide them tools uh and you we talk a little bit about our our our new uh product that we're we're announcing here. And then uh on the ones that just want to buy, we we sell them the actual intelligence that goes on the machines. And so we meet the customer wherever they're ready in their in their journey. Um the more nuanced version of that is you know uh the reality is like every every product is a different product and so the amount of silicon and amount of dollars you can put towards it uh towards sensors what the customer is willing to pay all that depends on what actually gets in the long horizon. All these things will be fully autonomous, but the intermittent steps are very much what we saw in the PC where you you have this slow step up to one day that'll be like now nobody really looks at laptop specs and even maybe frankly your phone specs but that's not the case from basically 85 to 2002 2005 where finally people stop actually specking at all and and and then they're really moving to laptops. uh but there's a similar kind of 20-y year I think uh horizon there >> broadly when we talk about machines and machines becoming intelligent right fundamentally a machine is it's a collection of these different components that are integrated right and and whoever does that final integration is often times the company that puts their their badge on it the brand name but many many companies are building technology that goes into that machines and so we now have a bunch of technology components and platforms that can go into these machines but we also sell the core technology that can be used to develop help them as well. And if you look, by the way, under the hood of a of a dirt mover or like combine or diesel truck, they'll have Cumins engines in them. But nobody says, "Well, because all these guys buy Cumins, this this means that they're, you know, uh whatever Caterpillar is not a good company." It's like, no, that's just a component that they buy. Uh they they they have a different role. So, if you when you look in any of these verticals, there's it's just a complex web of folks. That's why I always say like the the chip kind of analogy actually works quite effectively because some none of those companies make chips but they all buy chips. Uh and so I think that that's it's a good way to think about it.
>> So self-driving cars. So you know we've all been talking about self-driving cars for like I think the whole thing started like around what 2005 or something with the DARPA Grand Challenge originally and so and then Google engaged in the program shortly after that.
>> Yeah. Late double O's. Yeah.
>> Late double O's. Um uh so almost basically around a little less than 20 years maybe. Um, and there have been lots of predictions over the last 20 years of like self-driving cars are imminent at any moment. Um, so I guess the the bad news is we're sitting here today and most cars are not self-driving. The good news is there are now self-driving cars.
>> Yeah.
>> And so the way most cars are driving all over, you know, in the places they're deployed, it's become, you know, like people in San Francisco, I think, treat it now as routine that they get into.
>> And I think you can call I think Tesla, it's kind of like the AGI thing. It's like, you know, if we're talking 20 years ago, everything we're seeing right now is like mind-blowingly AGI. The post keeps moving. The Tesla stuff's amazing.
You can look at a bunch of manufacturers. Blue Cruise, Super Cute Cruise, BMW, Volvo's Pilot, they're all quite impressive systems. They're not full self-driving, right?
>> But yeah, >> well, it's it's full self-driving X, whatever remote monitoring is happening.
Um, uh, the the Tesla we have we have a home in in Los Angeles, and you guys may recall there was a there was a large fire in Los Angeles. Then the power then and then the California power grid was buckling even before that. And so, um, it actually turns out among the things Cybertrs are good at is they're they're very good, uh, batteries. Yeah. Yeah.
>> And so literally we have Cyber Trucks as our backup battery for the house. And as of last year, whatever the FSD released, I forget the exact one, but there was one where it like at least a lot of people thought it like really turned the corner >> 14. Yeah.
>> And like that thing drives you. I talked to somebody yesterday uh talked to somebody yesterday who has a Model Y who let the let the thing let the thing do the full route all the way up Highway One through Big Su.
>> Yeah. I think mean disengage like uh the uh uh meanantime and uh like uh miles per disengagement are are really high. I think miles is like in the thousands.
>> Yeah. which is like very impressive.
>> Yeah. The big for people who haven't driven the big highway one big su like that's a that's a that's a stressfilled drive. He said it was great the whole Anyway, so and I wouldn't have been talking to him had it not been would have gone right off right off >> because he unbolted the steering wheel.
So >> right right off the cliff. Um so um uh and then you know Tesla's rolling out their robo taxi you know is is starting starting to show up in in the wild. So, so on the one hand those exist. On the other hand, you know, 99.9999999% of cars are are still not self-driving.
And then I would say maybe just one other would be the self-driving trucks.
There's been this recurring kind of panic in the press of like the trucks become self-driving and the employment, you know, all these truck drivers be out of a job. And sitting here today, I don't think I don't know. Is there are there any trucks on the road that are self-driving that don't have at least a safety driver in the truck? And I think the answer is probably still >> Yeah, silver, if you so let's let's split let's split the there's multiple points that we brought up here. is on the let's say personally owned vehicles and why are they not more ubiquitous. Uh the part of that is the manufacturers are not good at deploying technology.
Part of that is they're they're they want to be safety conscious but most of it is cost cost. Um the the what you're seeing in China which is a chi China is kind of a different EV ecosystem mainly because they don't care about profits.
when you're talking about business that doesn't care about profits, it changes the entire calculus of the entire industry that doesn't care. Uh but what you're seeing is you're seeing L2++ systems. So we can simplify the entire self-driving conversation to is there a driver behind the steering wheel, >> right?
>> So this is a driver behind the steering wheel still there, but generally like Tesla drives everywhere. Uh they're like sub $1,000, right?
>> There's an aggressive that's that's chip, sensors, the package, the software, everything. uh we you know anticipate that there there's a very aggress once you get to like 500 the automotive OEMs will actually subsidize it for free they'll just give it to you this happened in nav systems if you guys remember nav systems used to be a big thing for you pay 4 grand 3,500 to get a nav system and then suddenly it became free and it just became default I think that'll happen the there's a a weird thing which is like actually getting into a subset of your cars costs x dollars and to get into all the cars costs x plus just a small incremental amount because it's just a fixed cost and the way that how many vehicles and the way the assembly line comes and the way you have homologation all these testing regimes all this stuff so I think you'll have wait and then a lot >> every single OEM without exception even the lowest dollar OEMs are working on an FSD competitor >> so it it'll it'll come but it is just like uh you know with uh a good analogy to think about self-driving in the personally owned ecosystem is mobile phones >> we had we had the the satellite phones then we had the Qualcomm you brick phones. Then we had the Motorola Razors and from you know the late '9s to the late two you know double zeros everybody was like when's mobile going to come there was a huge like and then it it comes and by by 07 from the iPhone launch like it's like four years when you get Uber, Instagram, WhatsApp, Snapchat, right?
>> Those are the killer applications. So I think there's a very very similar kind of wait and then it's just basically ubiquitous in every vehicle. Um, if you had to ask me for what that number is, 28 SOP, 29, start of production, 29, 30, and then by the early 30s, it'll start becoming very cheap to free.
>> Routinely by by the early 30s, you would just you buy a car and you just assume self-driving.
>> Exactly. Or it has the driver in seat L2++ system being very specific >> like Cybertruck or Tesla what Tesla people have what Tesla owners have today will become common.
>> Yeah. Will be default. So then the question then uh the the the you know the the other side of this is why don't we have a bunch of Whimos everywhere >> specifically Whimo has a different technology without getting into the nuances uh here but Tesla and many of the Chinese and applied were very much in this uh endto-end model architecture.
Um this is a new way of doing self-driving. Whimo for the lack of a better word is not that. It's not doesn't mean they're not learned. It does it just there's not one endto-end system. It's not one monolithic model.
Um, one of the proclivities of that their approach is it does depend on HD maps. Therefore, there is a geo fencing concept. I think Whimo is trying hard to remove that bottleneck so they can expand geographically faster. But the reality of today isn't there. The other thing is when you have researchers, which they which really was uh coming out of uh an alphabet research uh organization, they didn't put commercial constraints. So, the sensors are bespoke and expensive. The cars and the than the compute that are in there, they're they're just not economically feasible.
And they've tried a lot to get that down. But, it's kind of like it's a lot easier to go from something that's really cheap and make it more uh, you know, more featureful than something that's overbuilt and then trying to trim and make it really, really cheap. And that's the big debate. Who's going to get there first? Tesla with full self-driving or Whimo with cost and geographic ubiquity. Uh, but you know what we're not debating about? Is it gonna happen, >> right?
>> You know what we're not debating about?
Like is is is there a big technical breakthrough that needs to happen? None of those things. So now we're clearly in the engineering side of self-driving, which is just this grind down to like dollar per mile efficiency. And the moment that it's cheap, guess what? All the OEMs are smart. They'll just they just adopt it. It's not the OEMs are resistant because they don't think consumers want it or they don't understand the technology. It's because they want a price envelope which allows them to keep their thin razor thin margins and at a scale right >> which is deployed across 100 plus countries in V1 right >> and so if you're just doing a small deployment it's very different and I think and and that was the last thing I would say is the buyer of a Subaru >> or a buyer of a Suzuki have very different brand expectations that a buyer of a Tesla >> and so including the age of the consumer and what they think will happen and won't happen. Um, so that that also the reason. So if you're a Suzuki, you're like, well, my buyer is like not doesn't want this stuff, so I'm not going to jam it into the car. It's not because they're not like technically competent.
It's just a different area.
>> When do you think when do you think it'll be routine? Let's say the 200 biggest American cities like um would it be routine to walk outside and you just you just take it for granted that a robot taxi can come pick you up?
>> Uh it's 26 now. Um I mean certainly by 30. All right. Okay.
>> Yeah. Certainly by 30. And I I would say the big like variable there really is like >> because you what Whimo will say is that the dollars and cents per city already work and it's like well a company that has basically unlimited capital. Why are they not already in 200 cities? But then you see their launch schedule is pretty aggressive and you're like that that can that can get there. So >> maybe if I was being aggressive I would say 28.
>> Yeah. Okay.
>> Like I would say I would say available in 30 but routine and maybe like 32.
Yeah, 33.
>> Sure. Because there's a scale up.
There's a volume.
>> And also if you you you know you live in LA and so like five years ago I'd go to LA, people would be like, "What's applied intuition? I don't know what self-driving cars are." Uh in the last couple years now, they all know self-driving. And some of them even know applied intuition because they know from the other manufacturers. Uh I think you fast forward another two to four years, everybody knows it. Now >> does that mean everyone's taking Whimos exclusively?
>> Right?
>> That answer is no actually. Now there is a huge huge if you look at the numbers if you're Uber you got to be scared. I mean they're just eating into into ride sharing.
>> Um >> yeah but to get 100% ubiquity I mean that's that's that's another it has to be extremely cheap.
>> And what about long haul trucking?
>> So long so that's what we that's the passenger side. The long haul trucking completely different economics completely different um business model.
Um, there are many companies right now, I would say probably north of five that are running long haul trucks with drivers carrying loads between America and China. If you had China, it's probably getting into double digits. So, it's there, but the reason uh you don't know it and the reason it's not top of mind is it's not a consumer product. And unlike uh on the Whimo and Tesla side where investors are willing to essentially give you uh you know some uh market cap uh you know uh adjustment for the potential of the they say the the trucking business is like you know made you buy a car with your heartstrings.
You buy a truck with a calculator.
>> Yeah. It's a calculator business >> and they and so it's like pure dollars and cents and so I think >> um you as the provider of self-driving trucks if you're doing the whole thing like some of the companies are which we're not >> you have to show every mile I'm going to I'm going to save you this many dollars and it's like for sure for sure for sure cuz the buyer's unsophisticated and they're just like well I already got a staff it can drive and it's like and they're like they're just not inclined now where we're playing in Japan it's not random that we're doing trucking demand. There's a massive labor shortage today and there's an imploding demographic uh uh you know situation and so there's a demand from almost every sector and that's why we've we've picked that market to to to really grow but I think like you can take like even more obscure like when will all uh queries you know literally like uh where you you know you you're moving cement you're moving you're moving dirt uh not queries u e a r u a r quaries. Quarries rock stone >> rock stone cement. Uh when when are those uh I can tell you the people who own those things and run those things want it today, >> right?
>> So it's literally then you don't you don't have a which literally we can't make the stuff fast enough, >> right?
>> Um the macro point though that people uh don't talk about I I think all the stuff's going to happen, >> right?
>> But that happened pretty soon >> and happened fairly soon. The macro point that in in legislation and and and kind of in in the in the kind of uh economics the political economy of of this conversation is AI is really you see you have this big push back in digital AI because accountants are like I don't know this is going to happen to my job and VCs I'm sure all of your associates are very scared but like in our univer they're debating whether they need us.
>> Yeah. Yeah. Yeah. In our in our universe it's the other way around. And it's like you literally I'll I'll I'll meet these you know operators and they're like we'll give you everything. Like if you can do this we'll give you everything.
So then it's just up to us to like get there as as you know aggressively.
>> Well, you know the fear for a long time has been for trucking for some reason triggers the at least the press's imagination on like you know sort of apocalyptic levels of job loss. Like will there >> but that's it's so wrong. Go ahead.
>> There's not there's not enough truck drivers and guess what? Nobody wants to freaking be a truck driver.
>> Why is that? Explain that.
>> Because it's a terrible job. It's like you're you're like you're like >> by the way I grew up the main feature of the town where I grew up was a truck stop. So I know the answer but why is truck why is truck driving now >> it's like it's like you're asking me it's you know what this is this is like a you know talking to my kid who's like well why can't I put my hand on the stove? It's like because it's going to burn your hand.
It's like but why? It's like after the third why it's like come on buddy let's do this >> the hard way.
>> Yeah let's So what's hard?
>> I'm kidding. Just wanted to make sure everybody knows I did not do that with the first time.
>> Yes. Yes. So what's hard? What why why is being a truck driver a difficult job or why would kids not want to do it when they grow up?
>> So what let me use a parallel analogy which is very clear and then you can why you know people will say like nobody wants to work anymore and they say well you know McDonald's has all these job openings. No, no. Actually, what it is is those people that used to work at McDonald's now Door Dash and Uber >> because it's better for them because they can open they can start their hours and end their hours and they don't have to there's no boss and they don't have to like stand on their feet and they can surf their phone in between, you know, orders and they don't like that's the reason it's not random. The market is efficient. And so in the truck driving example, why does somebody not want to be away from their family for 4 to 8 days in a row doing long haul trucking?
The more sharp example is in Australia, why don't people want to go literally buy a plane to go to a mine and work on or you go offro uh offshore oil rigs.
Those jobs exist. If you want a job that pays six figures, they exist. even with such lucrative pay packages, it's not enough because people are like, you know what, >> I like kind of being around my family and I'm willing to take an incremental decrease in cost and and how much money I make.
>> And and then also like I think today more than ever things like >> back pain and like being exposed to the sun and cancer and people that care about that's now a part of the >> this is the thing.
Tell me if I I have this right, but I believe it's because I think commercial long-term drivers die life expectancy 10 years less than their peers. And I think it's a con people say it's a consequence of several things. So one is some combination of nutrition and sleep. It's you know it's a basically you know >> it's yeah it's very difficult it's very difficult to eat well and exercise and >> what's your sleep score if you're a long haul trucker? Let me guess eight sleep on that.
>> Exactly. And so like obesity and then heart disease, hypertension so forth are all very high. Um one and then two is I think the vibration uh is very difficult. stress in the body. And then the third is you mentioned cancer, but I think it's the I think they uh truck drivers have like a much higher rate of melanoma on their left arm.
>> Exactly. Yeah. There's photos of like a truck driver who's been driving for 30 years, one half their face, the other half's face cuz they're exposed to the sun. Right.
>> A more interesting or even more uh stark stat. Mining is 1% of the labor pool globally, 8% of work related fatalities.
Do you think people are rushing to work in mines when they hear stats like this?
Most major mines have a fatality regularly, which means once, twice a year, three times a year. And if you ever visit a mine, you'll see that everything is based around safety because once you experience one of your co-workers dying, then you're like, >> "What am I doing here?"
>> Yeah.
>> Like there's other jobs I can take. And so it's uh I I I understand you're trying to enumerate for the audience like, "But these are not good jobs."
>> Yeah.
>> And and and and the best evidence is this is not a mining podcast. Isn't that a podcast about hey long haul trucking is so great? They're just not attractive jobs.
>> Yeah. And even trucker even truckers don't want their kids to become truckers like it's it's for that reason why they you know they want their kids to be in a safe at the very least like safe safer safer line of work.
>> But um do the do the notwithstanding all that do they how long will there be do you think there will be safety drivers in long haul trucks that are that are self-driving or or or let's say other even just somebody in the cab to deal with what happens when they arrive?
>> We we know multiple companies that have driver goals right now. Okay. So like they're they are working to get drivers out right now. Um >> you know without going into our own details >> to be honest it's not long. We're talking talking a few years and uh >> I think on the long end.
>> Yeah. On the long end. And the thing is it's it's a there there's a software technology thing which is one part of the problem. But the other part is it's the redundancies that you need in hardware and the validation necessary for those redundancies. And in many cases that can actually be a long pull.
It's like, oh, they're productionizing a fully redundant steering system, fully redundant braking system, that's that's not in high that's not in high value production yet. And once you get that in high value production, now you got the quality up and then that's validated and now you can actually do these >> near the price downs. Exactly.
>> Do do you guys do do you look like it's the you know these little delivery robots? Like is that do you see a world where there's a billion of those running around?
>> Yeah, I think so. I mean the uh the the the product that we're announcing I think it's probably come out around with this time is called Dana. So there's you can just simplify everything that applied intuition does into two buckets which is the we've been talking mostly about the models that go on the machines then this is we say onboard software or onboard AI then there's offboard AI this is the tools to design and develop these same systems the models that actually go on the machines our uh you know vision for that is and the the delivery robot is a great example is like a high school kid or a middle schooler they can make iPhone apps they should be able to autonomous systems. So why can't they just ask that's a very simple question.
Why can't a >> ninth grader make a delivery robot in their in their home? Well, they don't have the the actual environment that they would first develop the scenarios in. They would define the requirements.
Hey, I want this robot to go on my high school campus around these let's say four buildings.
>> Uh then how okay now that you define the the requirements then you have the scenarios get made. where are all the scenarios that can that can uh that can be made by using let's say a satellite image of the high school. Uh then now you have to train the robot. So you need some data. Where do you get that data?
There's maybe enough publicly available data that can actually train a fairly rudimentary robot. Okay. Now you got that data from online maybe YouTube videos couple other places. Suddenly the robot's not doing now you need to deploy it onto the actual machine. So then you deploy it onto the machine and then the robot runs into the wall. Okay. What happened there? The loop closes. That platform for designing and developing is what we're launching. It's called Dana.
Uh which is the street that applied intuition is headquartered on. Uh and uh and our our our you know this comes from our tooling background. And if you look at like how tooling has changed in the digital AI world, if you look at like what Claude did to all we alo remember like you know from mix panel to you know uh gitlab github all these now everything has moved into a very different almost IDE frankly speaking we think the same thing is going to happen in the physical world and so uh that that's yeah that's what we're that's what we're building that's what we built and that's what we're launching and we already use it inhouse uh to develop our autonomy system which is you know and we're working on the most kind of scaled complex systems uh on the planet in all these different verticals. So we're pretty confident that it's actually quite useful. We've seen massive productivity gains. Uh but also uh you know we think like uh other companies will use this to build their own systems because it gets to that mission that building intelligent machines.
Fundamentally or dana is our agentic platform for physical AI and and everything that we've built and developed over the past nearly a decade.
every every tool, every technique that's available in Dana and it's very actually easy to use uh with the aenic interface and so workflows that used to maybe take days or weeks to run, >> you can now run those in in minutes in many cases, right?
>> And uh and this just lowers the barrier to entry to building these systems and just lowering the bar of like >> you know what it means to develop an autonomous system. Autonomy is still actually quite in the scope of software is quite exotic. It's not because of the things that we've talked about and we've just brought that down very very aggressively and it's kind of like you know the old adage of like how do you make a great product in software? It's like you either increase safety, convenience or cost and we want to try to do all three of those things with with Dana. Um and our hope is just like you said like you know >> uh kids can develop robots for their uh for their own use and and that extends to humanoids. So we're not just talking about like landbased systems or or you know ones where that are that are uh so you humanoids you can do drones.
>> The fact that right now writing drone software and deploying it at the time it's quite obscure and almost hobbyist >> uh we want to just make that absolutely like you know maybe not child's play but like teenager play. So this points to a world of like just like a lot more experimentation and entrepreneurship and like agriculture bots and like basically every domain construction >> defense you just all of a sudden have a much larger number of people who are applying creativity and coming up with ideas and making things that move.
>> Yeah. And have you seen like with Claude it's like it's one thing just to make the engineer more efficient or bring more people into engineering but then when these agents really run you're getting into you it's just like the iPhone example of you couldn't imagine Instagram >> Yeah.
>> before >> like the iPhone it's like imagine 2005 on laptops you're like in >> Yeah.
>> 10 years there's going to be this this app.
>> Yeah.
>> Where and you can put photos. They're like well the phones don't have have cameras. like, yeah, but it's going to be like social, like what the hell?
Like, so like Facebook, it's like, see, it's just hard to hard. And so, we think >> by lowering that barrier, you're going to get way way more creative uh autonomy products, >> right?
>> Yeah.
>> I will definitely decide whether to include this or not. So, my kid is building autonomous bots in um uh Factorio.
>> Oh, nice.
>> Is one of his projects. And so, and but he's he's you know, he's had rolling because the the tool kit's not available yet. So, he's actually training and he's actually he's actually training models.
>> Yeah.
>> Uh he's gathering data in the game. Um and uh actually has like a whole army of like bots that he's developed that go.
>> Yeah. So like >> and then his mother is like why are you playing that game so much? And he explains of course it's a purely educational process and experience but but it's you know it's the kind of thing. It's like yeah it's like there you know >> like there's no reason autonomy should be this like >> uh you know uh obscure difficult you know alchemistic you know uh uh uh technology. And um and I think not only does that have a huge impact on on society, it also allows people to understand that these systems are not like >> you know magic.
>> It's like if I can develop a >> a Roomba for myself in my house on a weekend using Dana, >> right?
>> Then why then it's not suddenly so scary. And I think that's like that's that's important.
>> And we can it can it can support it can support people in all kinds of ways that we haven't even imagined yet cuz Yeah.
Absolutely.
>> Yeah. Exactly.
>> I mean you think about like uh you know folks with disabilities you know we always think about humanoids as like this very important task of folding laundry which seems to be.
>> So we focus on you know the important task but the when you allow these tools to exist. I mean I I you know we started a tooling company. I mean I feel so importantly that tools are like what separates actually advanced civilizations from you know less advanced civil civilizations. And our our first uh uh mark for the company was a monkeykey's head and then we got a designer who said what this is stupid.
>> I was like I thought it was pretty good.
>> So you were talking earlier about how when you know um the technology got got so good in mobile that there was a wave of these companies you know Uber, WhatsApp, Snap you know Airbnb etc. that emerged in quick succession. And so now that the technology is getting there for the infrastructure for physical AI, what are some use cases or companies that you can obviously it's hard to predict the future, but where where are you most excited for? Like what what could we be talking about the equivalent here of in quick succession?
>> I mean I think uh you know midterm we want Dana if not the short term to really you know make humanoids way more real. Uh there's I mean how many it's like a thousand core tasks in a home from uh from humanoids and these companies it's like such I mean I'm if you talk to people who work in these companies it's everything is difficult every step of the way is difficult collecting data is difficult uh you know cleaning that data is difficult training those models or deploying the model is difficult and the bar being I want a high school kid to make a humanoid so that that's our our our path and we think there there could be a lot there but that's like these the obvious stuff I think the true nonobvious stuff is going to we will we'll look back will be will be way way more interesting >> and and there's there's some core ingredients that we're bringing together in data, right? We're making it way easier to to actually get imitation learning to work, way easier to make reinforcement learning work in combination with that. Um we're we have pre-trained models that can be used as a baseline for a lot of things. Um world models, advanced simulation tech, all of these things come together and then you're sort of limited by your c creativity like well what what do I want to do? And if you think about any kind of physical AI task as it's it's a you are understanding the world and you're manipulating something and and we can build that that can be built now much more easily in this in this tool and I think sometimes people ask like us being a tooling company and like you take self-driving trucks we deploy self-driving trucks and many of the self-driving trucking companies use our tools I I I think some sometimes people ask oh look you know with Dana are you going to like enable all these competitors that's great >> right >> that's absolutely completely fine if If you look at Google and what Google did to web applications, there was a massive internet. Uh Google still succeeded through, you know, search and YouTube and and and other web apps and other folks learned and used open source products and then ultimately closed source products and ultimately ventureback products. And we think we think this the the same thing could happen here. I was at a robotics startup uh a while back that you you guys know well. Um and they had they were training you know they were doing go through a training process training their one of their arms to do particularly a killer app that I thought was very appealing which was >> picking up dog poop >> literally you know training over and over again the difference was for the >> and so you know I don't know why not right >> why not have the little why not have the little robot follow you around when you walk the dog pick up the >> poop. Yeah. And I think like like you >> I know somebody who built I forget who it was. Somebody who built a a little lawn robot that would go around and individ pick up individual leaves.
>> Yeah.
>> Because you got that problem right.
Okay. You you rake you you rake your yard. It's completely clean and then like two hours later there's like 14 leaves and you're like >> Yeah. Yeah.
>> Send up the little bot to pick up the leaves.
>> It's like if development costs are zero then people will do that. I mean you guys remember like the early iPhone apps the hits were like the beer one or the fart app. If you imagine that in like Yeah. If you imagine that in 98 with a, you know, with the Symbian mobile, you know, whatever the OS from, I think it was Ericson or somebody, that'd be impossible. You need a team of like 50 people to to develop like the a beer thing for the Blackberry. So, I think there's a similar type of thing that's happening. We're, you know, we really want to be a part of that and we're going to enable that. And if it like makes making like I think it still be a while before like making a robo taxi is like super super easy.
>> Yeah.
>> But that'll happen. But there's I mean the number of bots that could be the number of kinds of bots that could be deployed in healthare is almost healthare alone is is home care >> um and then um in um construction um you know all the physical traits >> it's like us sitting in 2007 and saying let's uh we should have an app store what types of apps and we would come up with like a list of eight and then like there'll be a messaging one and then there'll be a camera one and it's like now you look at the app store and it's like you know there's an app for like the hotel you go to and it's like you know to to order, you know, food off the menu, >> right?
>> Yeah.
>> Yeah, makes sense.
>> Yeah. The we're talking, you know, earlier about the the differences between digital AI and physical AI. We were sort of hinting at LLMs, but world models are, you know, in vogue right now. Why don't you talk about sort of the the state of them as it relates to physical AI and how we should think about them?
>> So, so first off, world models means about 100 different things and uh we had a team at CVPR recently and and I was joking with them about just how many different ways you can define what what a world model is. But uh when we're thinking about a world model way, we're we're typically thinking about it in the context of of a simulation, right?
Something that is effectively >> started as a sim company. Yeah.
>> Yeah. Something that is like sufficiently able to represent the real world and and is reactive in a sense where you can actually have let's say a uh an autonomous agent that's acting in this world and the world model is is behaving appropriately in response to that autonomous agent.
>> Maybe uh Peter, I think it's worth being super explicit here. We just go just one level lower. the the you know determinism in simulators kind of the sim tore gap physics based you know rendering all the way to like this generated world.
>> Yeah.
>> Where do we fit on it or or where you know Yeah. This describe the landscape I think maybe.
>> Yeah. Yeah. So, so this is like let's say simulation broadly, right? There's there's so many different ways of doing simulation and so the the more classical approaches of simulation very physics based and and you can decompose physics in all different ways and all different levels of abstraction and you can simulate with sensors or without sensors and and is it is just a body simulation or are we actually simulating for example the light in in the environment or the almost think about like the way CGI is done. If we l literally had technical artists and we have technical artists who would create assets which would go in the simulator which would mimic real like road signs and have you know reflectivity and material properties that you would see in the real world but as you guys know Hollywood is going through its own fundamental change and now you've generated uh techn the same thing is happening in our universe as well.
>> Yeah. Yeah. So, so that's sort of on the that's at the far end of physics- based simulation and and then the opposite end is is purely neural simulation. But within that spectrum, there's there's many different things you can do that are each useful in their own right. And so, one of those things is a Gaussianbased simulation, right? Where you have uh a a a effectively a representation of the real world that uh that has a a a 3D representation. um and that 3D representation is consistent meaning that if if you let's say have some reference point let's say a camera and that camera moves within that 3D world because the Gaussian uh is actually representing the 3D geometry of that world you'll actually get very high quality output from that there's a lot of value in that uh and and that's let's say one type of of world model but when you go further on that uh on that spectrum into really into neural simulation then you get into these uh where you're actually generating the video feeds you can think a neural network that's actually outputting uh a video as what's actually coming out of the neurons of that and and that can be reactive uh which gives you some very interesting properties.
>> Reactive is in the ego does something in the environment and the other agents respond to the ego.
>> Exactly. Uh however, you're not guaranteed in that reactivity that it's it's accurate. Right. and and and now it's a question of well how can I align this this simulation this world model with the real world and the way that the real world would actually react and if you have perfect alignment between the real world and and the world model I think you've just sort of solved the universe roughly right that's impossibly difficult problem um but but as we make progress towards that uh it makes uh training physical AI models much easier because you can do more of that in simulation but the hardest part though is we're always talking about performance, right? So the I like to say the the the labs they they have it easy because they they can they can make models that are trillions of parameters and those models can be super slow and that's fine but we don't have that luxury in physical AI, right? We deal we we deal in real time like the actual actual clock real time and so we we have so many milliseconds before we have to do something and and those performance constraints they actually constrain the problem in a lot of ways. So we can have very large models and we do have very large models that are used in the offboard environment but when once you go onboard all of those constraints are very real and now we need to train a much smaller model that has these safety constraints these determinism constraints and uh and that's the hard part about physical it's also the moat right it's it's it is what makes our tooling and and our our competencies uh valuable uh because it's just really hard to to meet all of these constraints in in a physical system. When will you which will we get first a perfect a perfectly simulated real world environment for training autonomy autonomous devices or Grand Theft Auto 6?
>> You know I as long as they keep putting out great trailers. I mean I I watch I feel like I'm getting entertained without paying a dollar and reintroduced to Tom Petty because of >> Will you give us some timelines?
>> So this for a second.
>> Yeah. Go for it. Go for it. Well, no, look, I mean, so the whole thing was the whole thing with Grand Theft Auto is the big innovation was open open world open world sandbox gaming. So, it's a it's a sim it's a simulated city and at least in theory >> on that spectrum. I mean, we hire so many people out of the video game world on that spectrum. It's absolutely real.
>> Well, tell tell us about that. Yeah.
What's the spectum? So I here I this is this is speculation but I think Grand Theft Grand Theft Auto 6 will be perhaps the last major uh realworld video game that's still really developed let's say in that legacy era of traditional computer graphics tooling >> technical artists and yeah >> like I think that the Grand Theft Auto 7 will much more likely be like a world model based video game um and where you you could imagine as as AI tech evolves here you you have like this this concept of this video game world model And there's there's like some sort of baseline let's say data store uh that represents the real world in somehow and then and then you have a some translation layer that's actually turning that data store into something that you can see and and run around in like it's >> the game as a consequence could be the real world right as said you could have a complete recreation of the real world in the game this has kind of happened with flight simulators hasn't it isn't the most recent flight simulators are literally it's the entire planet rendered accurately is my understanding at least from the air. Is that right?
>> Yeah. Yeah. And I mean you're really that's where our our bread and butters when we started uh started the business we hired so many people out of the Microsoft flight sim organized it and uh you know >> but when you fly over you you know whatever New Yorker or when you fly over to Duth in the flight simulator now it is the real city right >> exactly but there there's some tricks that they play there and a lot of that is fidelity you know the real world the more you zoom in >> it stays a certain level of fidelity and uh and so the the tricks that you play there is you you basically are uh downsampling very very aggressively and then as you get closer you know then it becomes more more high fidelity where the real world isn't like that. the if you were to try to rebuild the world with this level of fidelity, it would, you know, would take the all the energy of the universe, right? It's it's it's it's quite complex. And that's probably, by the way, the best argument against us being living in a simulation is among the But of course, and you would say, well, the simulator we're in doesn't follow the laws of physics that we're at that point, how how do we know that the simulator that we're in is rendering all the stuff that we can't see?
>> Yeah. Yeah. Yeah.
>> As far as far as I know, everything happening outside this room doesn't even exist.
>> Yeah. I mean you you know like Buddhism believes this is a different type of podcast. It's like you know when you open your eyes the world is rendered and then you close your eyes the world that's literally religious.
>> I don't see why I don't see why it's necessary for it to keep rendering if I'm not there.
>> Buddhism from first principles.
>> Yeah. Exactly. That's what you should That'll get a lot of clicks. That's what you need to call this.
>> Yeah. Well, just go on the timeline topic. You know, we gave us timelines on self-driving cars. What timelines do you want to give us if any on sort of uh you know other interesting things that were worth tracking like perhaps when we'll get laundry folded or uh other you know things that emerge because of humanoid >> and I think also maybe just touching a little bit on world models where we see world models going because I think it's it's fundamental to what the work we do.
>> True.
>> Yeah. Yeah. So, so to answer the the first question, um, so laundry folding, it's uh it's it's not terribly far from being solved to to be clear. Uh, and there is a lot of interesting research being >> and then humanity can rejoice. That's in Proverbs 4:16, I think.
>> Well, here I I I I do think I do think housekeeping is is a killer use case for physically eye, right?
>> Peter thinks two that he always talks about in the company. One is housekeeping and it's entertainment.
Peter's long on humanoid entertainment.
>> Is it like what kind of entertainment I would I would 100% agree with. I think entertainment is robotics killer ass. I don't think anybody knows. What do you >> I mean like some of these Midwest white guys are really into >> I'm just saying what I think I think >> I just want to know when I go west world. That's all I want.
>> No, I actually have an entertaining I have a little I have a little a tiny little Chinese robot dog that's just like literally it's just like a little it's just a little and it just like roams around and it just like does >> Would you pay to see Circus LA with robots? Yes. Yes. I I want to see I want to see kung fu trapeze swinging >> spoken by like a compiler's guy here.
>> I know. But >> I want Westworld. I want Westworld.
>> I mean the the funny thing is I was I was just saying like well people in like the suburbs of Detroit actually that passes the test. I bet you people in Sterling Heights would actually pay to see that. It's actually true.
>> I stand corrected. I stand corrected.
>> But uh but back on laundry folding for a moment. Um It's it's actually not far from being folded if you remove the time constraint. And so the the trick that's played and if you look at the latest research videos is they'll say like play it at 8x real or whatever, right? And and that's for you to make it watchable.
>> So so the question is when can you actually reach human parody of performance? That's that's further off >> when you decouple uh models from just the hardware. The hardware can do it now. That used to be a constraint. So the hardware is very fast and accurate now which was actually >> there's still overheating issues that are still being dealt with but it's it's not it's not terribly far off like these are sol >> I mean it's far off from like you know when I was a mechie that was like fantasy like there's like nothing can >> What's the movie that has the most realistic future vision of robots >> oh man bysentennial man >> is it okay >> yeah he's realistic why that one I actually haven't seen >> I like I like that scene I think it's iRoot when Will Smith jumps in the car and uh his you know whatever his like accomp's in the car and he's like puts the car in manual She's like, "What are you going to drive this thing yourself?"
Like out of like, you know, she's Yeah.
She's like, "Are you crazy? What are you going to drive this thing yourself?"
Like, that's what we're That's applied to intuition's, you know, like goal.
Yeah.
>> Well, get by the way, I haven't seen this movie in a long time. Probably since it came out. So, my recollection of probably a bit incorrect.
>> Don't worry, the internet will correct you.
>> But I I I think Bsentennial Man has uh fully self-driving cars. Uh and it also has the housekeeping robot, okay? which which is played by Robin Williams, and that's it's sort of like the the friendly guy that will uh the friendly robot that will clean up and also babysit your kids and stuff like that.
And it seems like it's it's in the not terribly distant future.
>> I got a different answer. You guys ever see that movie uh Sam Rockwell Moon?
>> Oh, yeah.
>> Yeah. The the setup. I don't want to It's a great movie. Don't watch the trailer. Just watch the movie. Uh it's the the premise is the tagline of the movie is 250,000 miles from home. You find who you are. And it's one guy who works in an energy harvesting base run by applied intuition uh run by lunar technologies.
I I I don't want to be whale I don't want to be whailing utani. I don't want to be you know that's from uh from the alien franchises and then tar corporation from bladeunner. No no I want to be lunar technologies in the moon franchise one. uh to this one guy who works on this and the base basically runs by itself and he's just there to kind of mind it when things kind of some you know error signal.
>> Yeah. The the reason why it's it's I think so accurate is because the state-of-the-art for AI systems is like these systems they just need the occasional grounding. They'll just go off and do something crazy and then you say no no stop doing that.
>> LMS are like that too. I mean that's that's what coding bots are like right >> and and the reason other reasons I think it's quite accurate they it's uh maybe uncou now but Kevin Spy is the the AI you know smiley face and he's he's just there to kind of plate the human to to assist but to also like he's like oh you're you seem like you're sad Sam and like you know like that's the the the the but really it's the one running the base and uh hopefully I mean I shouldn't say we want to be lunar technologies because I don't know that they're quite a positive force in nature in that in that. But I think massive energy farm that's completely autonomous, >> that's going to be the future. And I and I think everyone like everyone reacts to things like that with like fear. And it's like, >> guys, that's amazing. That means energy costs go way down. Like that's an incredible positive thing. I think I think the you know uh I just did this commencement speech at my uh my undergrad.
>> Did you get uh destroyed?
>> No. You know what? I I >> unlike Eric Schmidt.
>> Yeah. Yeah. Yeah. Yeah. Listen, listen listen. My my This is the true story. My My wife started watching. She said, "I feel like you're yelling at me. I can't watch this."
>> So, I I I basically I mean I I don't I'm not like I don't I don't I'm not going to say which tech leaders who just basically avoid it by like punting and saying I'm not going to talk about it. I I talk about this stuff. And partly it's the General Motors Institute. No one's like these these are I I don't want to I don't I don't I don't want to throw judgment on you know the people we recruit out of out of MIT and Stanford but I say GMI people are a little different and they're like you know they're like >> pragmatic people they under they you know you don't go to a place like GMI if you believe a superficial view of what corporations do. Corporations are just people working on projects together and by the way people working on projects together in government people working projects together in nonprofits they all screw up. It's and so it's it's too simple to say AI corporations are are terrible. This also you can't say the other side which is like it'll all be great. So you have a role to play.
That's basically what you know what my what my message is and it's that's that's the case. I think if you feel like if if it if the the of the the I think the obvious uh you know uh abundance that comes from self-driving truck self-driving cars and the fact that people don't die which is amazing but then you also get this efficiency of cheaper energy etc. If all those things don't still satisfy your your fear you you as a person it's up to your responsibility to really learn about that technology. You you you can't just say, "Well, I'm afraid of it and my reaction is shut it down." That's not that's simply it's the uh and I don't say this just to say that we're competing with the Chinese, but there's a confusion.
The saying by confusion is no hand can block the sun.
>> Mhm.
>> And the sun is technological progress.
And if we as a society don't embrace technological progress, we will be left behind.
>> Somebody else is going to do it.
>> Somebody else is going to do it. And it might it's not the Chinese. Who knows?
Maybe it's a Usuzbck or you know it's another country that is is recognizing hey my citizens are suffering and I'm going to use this technology to remove them. It is honestly it's because we live in such a great society that we can have these like >> I would say stupid conversations like there still are people who don't can get food. Yeah.
>> And and and and someone will immediately quip if they were debating me, they would say, "Well, there's plenty of food. It's it's the capitalist system that doesn't No, no, no. Let's let's be very specific. There's plenty of food, but getting that food to those people is difficult."
>> So, that means we should let robots get that food to them faster.
>> That's just that's just how it is. So, I think I'm I'm and I think like we we we as like technologists, I think sometimes we, you know, it's I think an inclination just to say leave these people behind. I think you have to bring them along. You have to explain it to them. But we also have to treat folks like adults and say if you don't get it after I explain it a couple times then you just don't get it. So there's like a middle ground. It's not everyone's an idiot and every or we should just be technology will just be perfect perfect.
There's a middle ground. Let's have that conversation to a point and then we just move forward and we make society better and then the results show it. I mean there's people who still shockingly believe communism is the right answer. I mean I just want to say why and I'm I'm you know I I am a capitalist. I I cannot admit that. But there's 70 years of history there. Like we're that's not even a debate anymore. I mean that I think it could be a debate. If we're if we're sitting here in 1965 and having a debate, you say, "Okay, maybe centrally controlled systems work better." There's no debate anymore, folks. You know, systems where individuals make decisions on their own interest actually work better for society. And so that doesn't mean everything is perfect and you can't extrapolate that same thing with, you know, AI. Doesn't mean everything is going to be perfect, but net net it's definitely going to be better. That's roughly what my commencement speech was without the booze. They Mark and these guys were booing so they just cut it out.
Eric was booing. He's throwing stuff and they just edited it out.
>> You mentioned uh you the Japan market earlier. Why don't you talk briefly about sort of the global ambitions and how these technologies you know interplay and and what we're doing here?
So I think uh America particularly is still uh the most advanced in terms of when you take account the business model. The second thing for a company like applied intuition, we're an extremely global company. We work with everybody uh uh uh uh minus we don't have an office in China but really everyone else on the globe. Uh and we're a horizontal company. We're a technology provider. And I think we I think more Silicon Valley companies I think can employ a little bit of what we do which is work very I would say collaboratively with the local economies as sovereign AI becomes more of a real thing. We have to uh you know build businesses that take that into account. By the way, we're not the first ones to do this. If you look at the history of America, you read the history of Standard Oil, you'll see that this is this is that was the history of companies. You'd work internationally. A RAMCO is not a random company, right?
you you build based on the real geopolitical realities of the time. And uh so I think you know we've I think navigated it quite well. I' I've lived you know in Japan, I lived in Germany, I lived in Dubai. So also being Pakistani by birth, I think that's also influenced our company. Peter's only lived in Michigan and here but he is uh a German.
So, so but so I think innately we're more we think about the globe more and I think when I was at both at Google and at YC I was always surprised at how uh kind of almost myopic the companies are just always looking at the market that's just like within the 30 you know between San Jose and San Francisco it's like actually the market is really big I think physical AI the nature of it being physical I think we we have to be a very international company and I think we've had a lot of success being a very very you know being international. Yeah.
Cool. I think it's a good place to wrap.
>> Okay. Peter Casser, thanks so much for coming on the podcast and congrats on big launch with Dana.
>> Yeah, thanks for having us.
>> Awesome. Great to see you.
>> Great.
Related Videos
1:08:04
Robotics
Setting up a curved screen with Immersive Calibration Pro 4 and multiple cameras (P3D v4)
FlyerOneZero
23K views•2019-07-21
1:01:27
Robotics
Robot Learning with Sparsity and Scarcity
allenai
379 views•2025-10-14
1:01:42
Robotics
Jorge Mendez-Mendez: Unlocking Lifelong Robot Learning With Modularity (2023-10-05)
umassmlfl
237 views•2024-01-06
2:16
Robotics
Northwestern’s MS in Robotics: Student Robotics Projects, 2023
NorthwesternEngineering
1K views•2024-05-31
11:14
Robotics
"Perfect" Turns: Turning by the Gyro - FIRST LEGO League (FLL) SPIKE Prime + EV3 RePlay Programming
ZacharyTrautwein
94K views•2020-10-02
1:01
Robotics
Gorkem Secer: TSLIP-based Deadbeat Running Control of Bipedal Robot ATRIAS
DynamicWalking-wv6qm
298 views•2018-06-22
4:10
Robotics
Self-Driving Cars Need Lessons On Human Drivers | Maddie About Science
skunkbear
26K views•2018-08-21
3:12
Robotics
Milrem Robotics’ THeMIS UGVs used in a live-fire manned-unmanned teaming exercise
MilremRobotics
99K views•2021-05-20
Trending
5:07
Philosophy
One Must Imagine Sisyphus Happy
vlogbrothers
61K views•2026-07-21
25:15
Business
Future of Taylor Farms
maighstirtarot5385
11K views•2026-07-21
12:30
Business
The Downfall of OnePlus!
techwiser
65K views•2026-07-21
23:12
Engineering
My Friend Locked Up The Engine On His K-Swapped Bug...
boostedboiz
128K views•2026-07-21