Training Robots in Worlds That Don't Exist | World Labs with a16z

We are building the next frontier We are building the next frontier [music] of AI which is what we call [music] of AI which is what we call [music] of AI which is what we call spatial intelligence. spatial intelligence. spatial intelligence. >> As Cynics, we are developing what we >> As Cynics, we are developing what we >> As Cynics, we are developing what we call a real to see to real pipeline. We call a real to see to real pipeline. We call a real to see to real pipeline. We can replace all the data, all the can replace all the data, all the can replace all the data, all the evaluation we need in the real evaluation we need in the real evaluation we need in the real environment by using the data that can environment by using the data that can environment by using the data that can generate at a scalable way in our generate at a scalable way in our generate at a scalable way in our [music] digital world. [music] digital world. [music] digital world. >> Think about human intelligence. We do a >> Think about human intelligence. We do a >> Think about human intelligence. We do a lot of simulation in our head. Why? lot of simulation in our head. Why? lot of simulation in our head. Why? [music] [music] [music] There's a very important role simulation There's a very important role simulation There's a very important role simulation plays that real world data doesn't play plays that real world data doesn't play plays that real world data doesn't play which is counterfactual. Reason which is counterfactual. Reason which is counterfactual. Reason >> what we are building is a consistent >> what we are building is a consistent >> what we are building is a consistent world consistent both over space over world consistent both over space over world consistent both over space over time over different [music] viewpoints time over different [music] viewpoints time over different [music] viewpoints and over different type of interactions. and over different type of interactions. and over different type of interactions. My north star is I want the robot to My north star is I want the robot to My north star is I want the robot to work. work. work. >> The world we live in can be multiverse >> The world we live in can be multiverse >> The world we live in can be multiverse that we create technology to allow that we create technology to allow that we create technology to allow people builders developers to act within people builders developers to act within people builders developers to act within different [music] spaces. Do you believe different [music] spaces. Do you believe different [music] spaces. Do you believe we'll ever be able to build robots that we'll ever be able to build robots that we'll ever be able to build robots that have the power efficiency of a human have the power efficiency of a human have the power efficiency of a human being? How far away are we from this? Is being? How far away are we from this? Is being? How far away are we from this? Is this like 5 years or this is like never? this like 5 years or this is like never? this like 5 years or this is like never? >> The TLDDR is >> All right. Well, it's great to have you >> All right. Well, it's great to have you both here. So, Feay, for the listeners both here. So, Feay, for the listeners both here. So, Feay, for the listeners that may not have the background, maybe that may not have the background, maybe that may not have the background, maybe you can give an overview of what World you can give an overview of what World you can give an overview of what World Labs does. Labs does. Labs does. >> Yeah. Well, World Lab has is a >> Yeah. Well, World Lab has is a >> Yeah. Well, World Lab has is a two-year-old startup. I I I think we two-year-old startup. I I I think we two-year-old startup. I I I think we should just recognize it's a frontier should just recognize it's a frontier should just recognize it's a frontier model lab. It's uh we we are building model lab. It's uh we we are building model lab. It's uh we we are building the next frontier of AI which is what we the next frontier of AI which is what we the next frontier of AI which is what we call spatial intelligence. And uh call spatial intelligence. And uh call spatial intelligence. And uh spatial intelligence is about spatial intelligence is about spatial intelligence is about um creating AI that has the ability to um creating AI that has the ability to um creating AI that has the ability to generate generate generate uh understand uh understand uh understand reason with and interact with spaces reason with and interact with spaces reason with and interact with spaces whether it's physical or virtual. And of whether it's physical or virtual. And of whether it's physical or virtual. And of course uh a means to an end towards course uh a means to an end towards course uh a means to an end towards spatial intelligence is building large spatial intelligence is building large spatial intelligence is building large world models. And that's what uh World world models. And that's what uh World world models. And that's what uh World Labs is mostly focused on. Labs is mostly focused on. Labs is mostly focused on. >> Yeah. So you've been saying this since >> Yeah. So you've been saying this since >> Yeah. So you've been saying this since the very beginning which is um you know the very beginning which is um you know the very beginning which is um you know the machine's ability to perceive and the machine's ability to perceive and the machine's ability to perceive and reason about spaces and act on spaces reason about spaces and act on spaces reason about spaces and act on spaces but I always had the assumption that the but I always had the assumption that the but I always had the assumption that the acting on spaces was some like longist acting on spaces was some like longist acting on spaces was some like longist future thing but now you're acquiring a future thing but now you're acquiring a future thing but now you're acquiring a robotics company and so maybe talk a robotics company and so maybe talk a robotics company and so maybe talk a little bit about the timeliness of this little bit about the timeliness of this little bit about the timeliness of this and the intentions. and the intentions. and the intentions. >> Yeah. So first of all, it doesn't just >> Yeah. So first of all, it doesn't just >> Yeah. So first of all, it doesn't just take robotics to act within spaces or to take robotics to act within spaces or to take robotics to act within spaces or to interact, right? I mean, look at the interact, right? I mean, look at the interact, right? I mean, look at the creative field, whether it's VFX or creative field, whether it's VFX or creative field, whether it's VFX or gaming and or design. Many use cases, gaming and or design. Many use cases, gaming and or design. Many use cases, you can create and act within virtual you can create and act within virtual you can create and act within virtual spaces. World Warlab's thesis has always spaces. World Warlab's thesis has always spaces. World Warlab's thesis has always been that um the the the the world we been that um the the the the world we been that um the the the the world we live in can be multiverse that we create live in can be multiverse that we create live in can be multiverse that we create technology to allow people builders technology to allow people builders technology to allow people builders developers to act within different developers to act within different developers to act within different spaces. spaces. spaces. >> Having said that the ability to act >> Having said that the ability to act >> Having said that the ability to act within the physical space is one of the within the physical space is one of the within the physical space is one of the most exciting and most profoundly most exciting and most profoundly most exciting and most profoundly important capability of the future AI important capability of the future AI important capability of the future AI world. So robotics is very much that. So world. So robotics is very much that. So world. So robotics is very much that. So worldlap has always believed that worldlap has always believed that worldlap has always believed that robotics is a important robotics is a important robotics is a important application as well as use case of uh application as well as use case of uh application as well as use case of uh spatial intelligence and world modeling. spatial intelligence and world modeling. spatial intelligence and world modeling. So by joining force with uh inviting So by joining force with uh inviting So by joining force with uh inviting scen scen scen team to world labs is part of our team to world labs is part of our team to world labs is part of our long-term vision and uh mission. We long-term vision and uh mission. We long-term vision and uh mission. We we've always committed to that. Amazing. we've always committed to that. Amazing. we've always committed to that. Amazing. So, Yunu, you're the co-founder of So, Yunu, you're the co-founder of So, Yunu, you're the co-founder of Scenix. So, maybe provide everyone with Scenix. So, maybe provide everyone with Scenix. So, maybe provide everyone with a quick um overview of your background a quick um overview of your background a quick um overview of your background and what Scenix does. and what Scenix does. and what Scenix does. >> Yeah. So, I'm Yunu. So, I'm currently >> Yeah. So, I'm Yunu. So, I'm currently >> Yeah. So, I'm Yunu. So, I'm currently co-founder of Cynics and also assistant co-founder of Cynics and also assistant co-founder of Cynics and also assistant professor at Columbia University. Wow. professor at Columbia University. Wow. professor at Columbia University. Wow. >> So, my research started from my PhD at >> So, my research started from my PhD at >> So, my research started from my PhD at MIT and then postto with Fay really at MIT and then postto with Fay really at MIT and then postto with Fay really at the Stanford [laughter] University. the Stanford [laughter] University. the Stanford [laughter] University. >> That's great. >> That's great. >> That's great. >> The world is small. >> The world is small. >> The world is small. >> The world is small. >> The world is small. >> The world is small. >> It is. >> It is. >> It is. >> Throughout my career, my goal has been >> Throughout my career, my goal has been >> Throughout my career, my goal has been very simple. trying to help the robots very simple. trying to help the robots very simple. trying to help the robots better perceive and interact with the better perceive and interact with the better perceive and interact with the physical world. So I'm a very practical physical world. So I'm a very practical physical world. So I'm a very practical person. I want my robot to work in the person. I want my robot to work in the person. I want my robot to work in the real physical environments. real physical environments. real physical environments. >> So for cynics the unique opportunity we >> So for cynics the unique opportunity we >> So for cynics the unique opportunity we see is that there has been a lot of like see is that there has been a lot of like see is that there has been a lot of like a bottlenecks. Right now we see faced by a bottlenecks. Right now we see faced by a bottlenecks. Right now we see faced by the developments of general purpose the developments of general purpose the developments of general purpose robots especially around training and robots especially around training and robots especially around training and also around evaluations. So as we are also around evaluations. So as we are also around evaluations. So as we are developing what we call a real to sim to developing what we call a real to sim to developing what we call a real to sim to real pipeline. Okay. We want to map the real pipeline. Okay. We want to map the real pipeline. Okay. We want to map the real environments into the digital world real environments into the digital world real environments into the digital world that has the best alignments with the that has the best alignments with the that has the best alignments with the real environments. By alignments we mean real environments. By alignments we mean real environments. By alignments we mean that whatever happens in the digital that whatever happens in the digital that whatever happens in the digital world is also going to happen in the world is also going to happen in the world is also going to happen in the real environments real environments real environments >> such that we can replace all the data >> such that we can replace all the data >> such that we can replace all the data all the evaluation we need in the real all the evaluation we need in the real all the evaluation we need in the real environment by using the data that can environment by using the data that can environment by using the data that can generate at a scalable way in our generate at a scalable way in our generate at a scalable way in our digital world. So that is how everything digital world. So that is how everything digital world. So that is how everything started in Synex. We put together a very started in Synex. We put together a very started in Synex. We put together a very very strong and best teams around very strong and best teams around very strong and best teams around robotics, robot learning and also robotics, robot learning and also robotics, robot learning and also simulation and rendering trying to build simulation and rendering trying to build simulation and rendering trying to build this realtom real stack to solve some of this realtom real stack to solve some of this realtom real stack to solve some of the key bottlenecks. the key bottlenecks. the key bottlenecks. >> It's amazing that you two work together. >> It's amazing that you two work together. >> It's amazing that you two work together. >> Yeah. And uh there is a funny story here >> Yeah. And uh there is a funny story here >> Yeah. And uh there is a funny story here because you would think because we because you would think because we because you would think because we worked together he was my amazing postto worked together he was my amazing postto worked together he was my amazing postto we've been talking about this and world we've been talking about this and world we've been talking about this and world lab um um integration for a long time. lab um um integration for a long time. lab um um integration for a long time. It's actually not true. They came into It's actually not true. They came into It's actually not true. They came into World Wars as a customer. World Wars as a customer. World Wars as a customer. >> Really? >> Really? >> Really? >> When we when we released the first >> When we when we released the first >> When we when we released the first version of our generative model called version of our generative model called version of our generative model called Marble last winter around uh November, Marble last winter around uh November, Marble last winter around uh November, December, December, December, >> Cynics just signed up. >> Cynics just signed up. >> Cynics just signed up. >> No kidding. AS A CUSTOMER. >> No kidding. AS A CUSTOMER. >> No kidding. AS A CUSTOMER. >> YES. [laughter] >> YES. [laughter] >> YES. [laughter] And I didn't even know what it was. And And I didn't even know what it was. And And I didn't even know what it was. And then I realized this is Vindrew's then I realized this is Vindrew's then I realized this is Vindrew's company. I called Vindro. I'm like, company. I called Vindro. I'm like, company. I called Vindro. I'm like, "Wow, this is your company." and you "Wow, this is your company." and you "Wow, this is your company." and you guys and and then we realized there's so guys and and then we realized there's so guys and and then we realized there's so much synergy. much synergy. much synergy. >> Maybe Faith could just quickly describe >> Maybe Faith could just quickly describe >> Maybe Faith could just quickly describe what Marble is. what Marble is. what Marble is. >> Yeah, Marble is the code name for the >> Yeah, Marble is the code name for the >> Yeah, Marble is the code name for the base model that World War has been base model that World War has been base model that World War has been training and iterating on that the training and iterating on that the training and iterating on that the fundamental capability right now of fundamental capability right now of fundamental capability right now of Marble that is publicly released is to Marble that is publicly released is to Marble that is publicly released is to take a a prompt. It can be an image, it take a a prompt. It can be an image, it take a a prompt. It can be an image, it can be a few images and uh or a text and can be a few images and uh or a text and can be a few images and uh or a text and turn that into a geometrically turn that into a geometrically turn that into a geometrically consistent world that can be represented consistent world that can be represented consistent world that can be represented in 3D geometry whether it's gausian in 3D geometry whether it's gausian in 3D geometry whether it's gausian splat or mesh. Really what Synix team is splat or mesh. Really what Synix team is splat or mesh. Really what Synix team is doing is trying to solve this extremely doing is trying to solve this extremely doing is trying to solve this extremely difficult problem in robotics which is difficult problem in robotics which is difficult problem in robotics which is the lack of data. M the lack of data. M the lack of data. M >> the lack of data in training, the lack >> the lack of data in training, the lack >> the lack of data in training, the lack of data in uh evaluation. This is very of data in uh evaluation. This is very of data in uh evaluation. This is very very different from language models very different from language models very different from language models where data is abundant on the internet. where data is abundant on the internet. where data is abundant on the internet. >> And we know that um in order for >> And we know that um in order for >> And we know that um in order for robotics to work, we have to somehow robotics to work, we have to somehow robotics to work, we have to somehow unlock the power of scaling law. But unlock the power of scaling law. But unlock the power of scaling law. But where does that come from? This is where does that come from? This is where does that come from? This is something that that is a profound something that that is a profound something that that is a profound problem that everybody's battling with problem that everybody's battling with problem that everybody's battling with in in robotics. It'd actually be great in in robotics. It'd actually be great in in robotics. It'd actually be great to talk about this energy like you have to talk about this energy like you have to talk about this energy like you have put together a very very talented team. put together a very very talented team. put together a very very talented team. You have put together a very talented You have put together a very talented You have put together a very talented team and so like to what extent is there team and so like to what extent is there team and so like to what extent is there overlap to what extent is this an overlap to what extent is this an overlap to what extent is this an extension? Maybe talk a little bit about extension? Maybe talk a little bit about extension? Maybe talk a little bit about that. Yeah, that's actually like how that. Yeah, that's actually like how that. Yeah, that's actually like how complimentary it is. complimentary it is. complimentary it is. >> It's it's actually the the TLDDR is is >> It's it's actually the the TLDDR is is >> It's it's actually the the TLDDR is is very complimentary and with the shared very complimentary and with the shared very complimentary and with the shared mission. So is one of the three uh mission. So is one of the three uh mission. So is one of the three uh technical co-founders. The other two are technical co-founders. The other two are technical co-founders. The other two are Changi Jan, another Colombia professor Changi Jan, another Colombia professor Changi Jan, another Colombia professor who has been a world-class technologist who has been a world-class technologist who has been a world-class technologist in simulation. Wow. And Changi has his in simulation. Wow. And Changi has his in simulation. Wow. And Changi has his background in also um VFX. He worked at background in also um VFX. He worked at background in also um VFX. He worked at Weta, he worked at Tencent, he's being Weta, he worked at Tencent, he's being Weta, he worked at Tencent, he's being an entrepreneur. Uh then there's Sunonni an entrepreneur. Uh then there's Sunonni an entrepreneur. Uh then there's Sunonni who is a phenomenal engineering leader who is a phenomenal engineering leader who is a phenomenal engineering leader who was also in a uh startup uh that was who was also in a uh startup uh that was who was also in a uh startup uh that was acquired by Amazon many years ago. So he acquired by Amazon many years ago. So he acquired by Amazon many years ago. So he worked in many different tech stacks in worked in many different tech stacks in worked in many different tech stacks in the computer vision field in uh in the computer vision field in uh in the computer vision field in uh in Amazon. So when when we started talking Amazon. So when when we started talking Amazon. So when when we started talking more seriously I recognized that uh a more seriously I recognized that uh a more seriously I recognized that uh a couple of things that Phoenix has from a couple of things that Phoenix has from a couple of things that Phoenix has from a talent point of view is extremely talent point of view is extremely talent point of view is extremely complimentary to to uh worldaps. one is complimentary to to uh worldaps. one is complimentary to to uh worldaps. one is obviously obviously obviously incredible um uh thought leadership and incredible um uh thought leadership and incredible um uh thought leadership and and just technical prowess in robotics and just technical prowess in robotics and just technical prowess in robotics right so uh from really from hardware right so uh from really from hardware right so uh from really from hardware full stack robotics and even when he was full stack robotics and even when he was full stack robotics and even when he was my posttock at Stanford at that time you my posttock at Stanford at that time you my posttock at Stanford at that time you already had your faculty offer so you already had your faculty offer so you already had your faculty offer so you were there only for one year I wanted were there only for one year I wanted were there only for one year I wanted you for more than one year but he had to you for more than one year but he had to you for more than one year but he had to go become a have the real job so uh he go become a have the real job so uh he go become a have the real job so uh he was a full stack researcher in in was a full stack researcher in in was a full stack researcher in in robotics from modeling to to hardware. robotics from modeling to to hardware. robotics from modeling to to hardware. Uh and of course uh VR and his student Uh and of course uh VR and his student Uh and of course uh VR and his student uh students at Synenix was that pool of uh students at Synenix was that pool of uh students at Synenix was that pool of talent world hasn't had yet. Then on the talent world hasn't had yet. Then on the talent world hasn't had yet. Then on the Changi side is just incredible Changi side is just incredible Changi side is just incredible simulation um um capability, right? he's simulation um um capability, right? he's simulation um um capability, right? he's such a senior um researcher and such a senior um researcher and such a senior um researcher and technologist in simulation and and uh technologist in simulation and and uh technologist in simulation and and uh what world labs is doing is very much um what world labs is doing is very much um what world labs is doing is very much um interfacing the world of simulation. So, interfacing the world of simulation. So, interfacing the world of simulation. So, so I think what they don't have um so I think what they don't have um so I think what they don't have um obviously is on the generative model obviously is on the generative model obviously is on the generative model side as well as the computer vision 3D side as well as the computer vision 3D side as well as the computer vision 3D reconstruction side we're also very reconstruction side we're also very reconstruction side we're also very strong at world labs. So that's a strong at world labs. So that's a strong at world labs. So that's a technology that scenics needs. So technology that scenics needs. So technology that scenics needs. So together these uh these two sides come together these uh these two sides come together these uh these two sides come together and make it much more complete. together and make it much more complete. together and make it much more complete. Fe's motivation in this as like this is Fe's motivation in this as like this is Fe's motivation in this as like this is an extension and a compliment to get an extension and a compliment to get an extension and a compliment to get into robotics you know having been in into robotics you know having been in into robotics you know having been in your situation which is deciding when to your situation which is deciding when to your situation which is deciding when to sell a company it would be great to hear sell a company it would be great to hear sell a company it would be great to hear from you on like how you think about from you on like how you think about from you on like how you think about joining world labs and kind of the fit joining world labs and kind of the fit joining world labs and kind of the fit there and like why you made the decision there and like why you made the decision there and like why you made the decision to do it. Yeah. So at the very beginning to do it. Yeah. So at the very beginning to do it. Yeah. So at the very beginning we were deciding okay we want to just we were deciding okay we want to just we were deciding okay we want to just keep going but after chatting with F keep going but after chatting with F keep going but after chatting with F after seeing all the synergies that after seeing all the synergies that after seeing all the synergies that happening in the middle it just makes happening in the middle it just makes happening in the middle it just makes perfect sense for the forces to join perfect sense for the forces to join perfect sense for the forces to join each other. So in any sense as cynics each other. So in any sense as cynics each other. So in any sense as cynics what we have been doing is real to seem what we have been doing is real to seem what we have been doing is real to seem to real is to do this reconstruction of to real is to do this reconstruction of to real is to do this reconstruction of the environment. So we capture the the environment. So we capture the the environment. So we capture the appearance of the environment geometry appearance of the environment geometry appearance of the environment geometry of the environments and also the of the environments and also the of the environments and also the dynamics of the environment meaning how dynamics of the environment meaning how dynamics of the environment meaning how the environment is going to change when the environment is going to change when the environment is going to change when you apply actions. So this tensor you apply actions. So this tensor you apply actions. So this tensor reconstruction right now is still a reconstruction right now is still a reconstruction right now is still a little bit on the heavier side and what little bit on the heavier side and what little bit on the heavier side and what world labs right now has been doing world labs right now has been doing world labs right now has been doing involves a lot of profound capabilities involves a lot of profound capabilities involves a lot of profound capabilities around sparse reconstruction and around sparse reconstruction and around sparse reconstruction and generations. So we see a lot of generations. So we see a lot of generations. So we see a lot of opportunities of leveraging like a opportunities of leveraging like a opportunities of leveraging like a marble and other like capabilities as marble and other like capabilities as marble and other like capabilities as world labs in order to do very efficient world labs in order to do very efficient world labs in order to do very efficient reconstructions and modeling of the reconstructions and modeling of the reconstructions and modeling of the environments. environments. environments. >> So can we expect a foundation model for >> So can we expect a foundation model for >> So can we expect a foundation model for robotics from world labs? robotics from world labs? robotics from world labs? World Lab is building a foundation World Lab is building a foundation World Lab is building a foundation model. As you know, Martin, we're model. As you know, Martin, we're model. As you know, Martin, we're building a base model and uh as the building a base model and uh as the building a base model and uh as the technology has been evolving, technology has been evolving, technology has been evolving, some of the most exciting base models some of the most exciting base models some of the most exciting base models are omniodels, right? They take they are omniodels, right? They take they are omniodels, right? They take they take multimodal input, they have take multimodal input, they have take multimodal input, they have multimodal outputs. And uh what is a multimodal outputs. And uh what is a multimodal outputs. And uh what is a foundation model for robotics? Uh it's foundation model for robotics? Uh it's foundation model for robotics? Uh it's very likely going to involve actions. M very likely going to involve actions. M very likely going to involve actions. M >> it's very likely going to involve the >> it's very likely going to involve the >> it's very likely going to involve the output of actions in addition to the output of actions in addition to the output of actions in addition to the state of the world and we're definitely state of the world and we're definitely state of the world and we're definitely not ruling this out. not ruling this out. not ruling this out. >> Yeah. Great. >> Yeah. Great. >> Yeah. Great. >> So for example for the foundation models >> So for example for the foundation models >> So for example for the foundation models it's essentially needs to be a it's essentially needs to be a it's essentially needs to be a multimodal model. So it has to take into multimodal model. So it has to take into multimodal model. So it has to take into account frame text image deps and account frame text image deps and account frame text image deps and different kind of modalities and action different kind of modalities and action different kind of modalities and action is a very very important parts of that is a very very important parts of that is a very very important parts of that modalities. So if you think about frame modalities. So if you think about frame modalities. So if you think about frame actions as inputs that is essentially a actions as inputs that is essentially a actions as inputs that is essentially a forward simulator that is going to forward simulator that is going to forward simulator that is going to predict how the environment is going to predict how the environment is going to predict how the environment is going to change when you apply a specific action. change when you apply a specific action. change when you apply a specific action. When the action is output this is When the action is output this is When the action is output this is essentially a policy model that is essentially a policy model that is essentially a policy model that is trying to predict given a specific goal trying to predict given a specific goal trying to predict given a specific goal like what should be the action you take like what should be the action you take like what should be the action you take in the real environment to get you in the real environment to get you in the real environment to get you closer to that goal. So this kind of closer to that goal. So this kind of closer to that goal. So this kind of omni models actually can benefit a lot omni models actually can benefit a lot omni models actually can benefit a lot and actually provide huge amount of and actually provide huge amount of and actually provide huge amount of values for the robotics communities in values for the robotics communities in values for the robotics communities in trying to understand how to model the trying to understand how to model the trying to understand how to model the environments and at the same time how to environments and at the same time how to environments and at the same time how to act in the environments and this can act in the environments and this can act in the environments and this can also acts as a backbone for you to also acts as a backbone for you to also acts as a backbone for you to fine-tune into specific robotic fine-tune into specific robotic fine-tune into specific robotic applications to making sure it's really applications to making sure it's really applications to making sure it's really leave up to the reliability and leave up to the reliability and leave up to the reliability and efficiency that expected by the clients. efficiency that expected by the clients. efficiency that expected by the clients. you know if uh Yun if you don't uh if you know if uh Yun if you don't uh if you know if uh Yun if you don't uh if you don't mind a kind of a lay investor you don't mind a kind of a lay investor you don't mind a kind of a lay investor question I see a lot of robotics question I see a lot of robotics question I see a lot of robotics companies and a very popular approach companies and a very popular approach companies and a very popular approach right now for the robotics companies right now for the robotics companies right now for the robotics companies that come in is like we'll use a video that come in is like we'll use a video that come in is like we'll use a video model you know and like you know that's model you know and like you know that's model you know and like you know that's the the predominant method where this is the the predominant method where this is the the predominant method where this is you know 3D and simulation it's a very you know 3D and simulation it's a very you know 3D and simulation it's a very different approach and so maybe you different approach and so maybe you different approach and so maybe you could contrast could contrast could contrast you know this popular approach of just you know this popular approach of just you know this popular approach of just using video only versus kind of what the using video only versus kind of what the using video only versus kind of what the ambition here is ambition here is ambition here is >> yeah so in order to create words where >> yeah so in order to create words where >> yeah so in order to create words where the robot can learn the words as I the robot can learn the words as I the robot can learn the words as I mentioned is to capture the essential mentioned is to capture the essential mentioned is to capture the essential structure of the problem and one of the structure of the problem and one of the structure of the problem and one of the very important necessary like very important necessary like very important necessary like requirements for those words will be requirements for those words will be requirements for those words will be consistency so that is where I actually consistency so that is where I actually consistency so that is where I actually see there's very very strong synergies see there's very very strong synergies see there's very very strong synergies with marble because what we are building with marble because what we are building with marble because what we are building is a consistent world consistence both is a consistent world consistence both is a consistent world consistence both over space over time over different over space over time over different over space over time over different viewpoints and over different type of viewpoints and over different type of viewpoints and over different type of interactions and marble the generated interactions and marble the generated interactions and marble the generated words from marble is also provides an words from marble is also provides an words from marble is also provides an infrastructure a component of that infrastructure a component of that infrastructure a component of that entire words that we believe is entire words that we believe is entire words that we believe is necessary for the robot to learn. necessary for the robot to learn. necessary for the robot to learn. Imagine if a robot push an object Imagine if a robot push an object Imagine if a robot push an object forwards. The object just magically forwards. The object just magically forwards. The object just magically disappear which has been a problem of disappear which has been a problem of disappear which has been a problem of many of the existing like video many of the existing like video many of the existing like video prediction models. This won't provides prediction models. This won't provides prediction models. This won't provides good enough signal for the robot to know good enough signal for the robot to know good enough signal for the robot to know like what is the right thing to do. But like what is the right thing to do. But like what is the right thing to do. But obviously right now there has been a lot obviously right now there has been a lot obviously right now there has been a lot of investigation on building better and of investigation on building better and of investigation on building better and better and stronger and stronger like a better and stronger and stronger like a better and stronger and stronger like a video models. So we actually see a way video models. So we actually see a way video models. So we actually see a way where some of the infrastructure we where some of the infrastructure we where some of the infrastructure we build can provide as a initial momentums build can provide as a initial momentums build can provide as a initial momentums and to going through this data flywheel and to going through this data flywheel and to going through this data flywheel of going from this like a more of going from this like a more of going from this like a more simulationdriven models into like a simulationdriven models into like a simulationdriven models into like a robot policy models which going to do robot policy models which going to do robot policy models which going to do the execution in the real environment the execution in the real environment the execution in the real environment collecting new data the data will come collecting new data the data will come collecting new data the data will come back in where the model doesn't back in where the model doesn't back in where the model doesn't necessarily have to be physics only or necessarily have to be physics only or necessarily have to be physics only or learning only but somewhere in the learning only but somewhere in the learning only but somewhere in the middle which be able to capture the middle which be able to capture the middle which be able to capture the essential structure of the problem but essential structure of the problem but essential structure of the problem but at the same times be able to scale and at the same times be able to scale and at the same times be able to scale and become better and better as you become better and better as you become better and better as you accumulate more data. accumulate more data. accumulate more data. >> You know, um I've worked now fay very >> You know, um I've worked now fay very >> You know, um I've worked now fay very closely for a while and and and closely for a while and and and closely for a while and and and you've always had this north star which you've always had this north star which you've always had this north star which has driven this and you know you you've has driven this and you know you you've has driven this and you know you you've articulated variously as kind of 3D and articulated variously as kind of 3D and articulated variously as kind of 3D and and in a number of other ways. And I'm and in a number of other ways. And I'm and in a number of other ways. And I'm just wondering for you is there also a just wondering for you is there also a just wondering for you is there also a similar philosophical northstar or similar philosophical northstar or similar philosophical northstar or you're more the pragmatic you're more the pragmatic you're more the pragmatic like I am like build the system like do like I am like build the system like do like I am like build the system like do the thing. the thing. the thing. >> My northstar is to make robots work. >> My northstar is to make robots work. >> My northstar is to make robots work. >> Amazing. Yeah. in the real environment. >> Amazing. Yeah. in the real environment. >> Amazing. Yeah. in the real environment. I'm a very practical person. I want the I'm a very practical person. I want the I'm a very practical person. I want the robot to work. One interesting thing robot to work. One interesting thing robot to work. One interesting thing that's actually coming from my that's actually coming from my that's actually coming from my collaborations with FIP during my collaborations with FIP during my collaborations with FIP during my postto, we are building this kind of postto, we are building this kind of postto, we are building this kind of benchmark. We actually send out surveys benchmark. We actually send out surveys benchmark. We actually send out surveys asking the general public what they want asking the general public what they want asking the general public what they want their robots to do for them. their robots to do for them. their robots to do for them. >> Yeah. >> Yeah. >> Yeah. >> Among the thousand tasks we collected, >> Among the thousand tasks we collected, >> Among the thousand tasks we collected, onethird of the tasks are about clean. onethird of the tasks are about clean. onethird of the tasks are about clean. >> People just don't like to do those like >> People just don't like to do those like >> People just don't like to do those like a d and dirty tasks. And those are the a d and dirty tasks. And those are the a d and dirty tasks. And those are the scenarios we really want to making sure scenarios we really want to making sure scenarios we really want to making sure we have robotic solutions to deal with. we have robotic solutions to deal with. we have robotic solutions to deal with. >> One thing I really like about Synenix uh >> One thing I really like about Synenix uh >> One thing I really like about Synenix uh Martin especially uh uh continuing your Martin especially uh uh continuing your Martin especially uh uh continuing your question there's a lot of robotics question there's a lot of robotics question there's a lot of robotics companies building models and all that. companies building models and all that. companies building models and all that. One thing I truly like about Synenix is One thing I truly like about Synenix is One thing I truly like about Synenix is is Renu and his co-founders have such an is Renu and his co-founders have such an is Renu and his co-founders have such an incredibly pragmatic approach to incredibly pragmatic approach to incredibly pragmatic approach to robotics. They especially they come from robotics. They especially they come from robotics. They especially they come from academia right Sunny doesn't but Vindra academia right Sunny doesn't but Vindra academia right Sunny doesn't but Vindra and Chani come from academia but their and Chani come from academia but their and Chani come from academia but their first instinct is work with design first instinct is work with design first instinct is work with design partners and customers in real industry partners and customers in real industry partners and customers in real industry whether it's is uh labs at um industry whether it's is uh labs at um industry whether it's is uh labs at um industry labs or or warehouses or um um labs or or warehouses or um um labs or or warehouses or um um >> electronics >> electronics >> electronics >> electronics >> electronics >> electronics you know uh assembly that is such a you know uh assembly that is such a you know uh assembly that is such a refreshing actually a refreshing way of refreshing actually a refreshing way of refreshing actually a refreshing way of approaching approaching approaching robotics and that that really tr made me robotics and that that really tr made me robotics and that that really tr made me very excited to work with them. very excited to work with them. very excited to work with them. >> Maybe this is for you and but I'll just >> Maybe this is for you and but I'll just >> Maybe this is for you and but I'll just be this is this is personal curiosity be this is this is personal curiosity be this is this is personal curiosity which is it seems to me that for which is it seems to me that for which is it seems to me that for robotics you have to be pretty exact. I robotics you have to be pretty exact. I robotics you have to be pretty exact. I mean not perfect but pretty close. But mean not perfect but pretty close. But mean not perfect but pretty close. But for the creative use cases which for the creative use cases which for the creative use cases which worldaps has done a lot of you kind of worldaps has done a lot of you kind of worldaps has done a lot of you kind of don't need to because you know I mean don't need to because you know I mean don't need to because you know I mean you know even sometimes like being wrong you know even sometimes like being wrong you know even sometimes like being wrong is stylistic or intentional or whatever. is stylistic or intentional or whatever. is stylistic or intentional or whatever. And so from a technical perspective what And so from a technical perspective what And so from a technical perspective what is the challenge here for reconciling is the challenge here for reconciling is the challenge here for reconciling these two things or do they never get these two things or do they never get these two things or do they never get reconciled like there always be two reconciled like there always be two reconciled like there always be two points in the design space? So they will points in the design space? So they will points in the design space? So they will be like reconciled in the in the long be like reconciled in the in the long be like reconciled in the in the long terms of course and um modeling of the terms of course and um modeling of the terms of course and um modeling of the environments um doesn't have to be environments um doesn't have to be environments um doesn't have to be perfect. The model doesn't have to be perfect. The model doesn't have to be perfect. The model doesn't have to be perfect in robotics. perfect in robotics. perfect in robotics. >> And by the way is there again this is >> And by the way is there again this is >> And by the way is there again this is pure curiosity but is there like um a pure curiosity but is there like um a pure curiosity but is there like um a bit more formal way to say that like bit more formal way to say that like bit more formal way to say that like what does that mean not to be perfect? what does that mean not to be perfect? what does that mean not to be perfect? It has to be pretty close. It has to be pretty close. It has to be pretty close. >> So so let me putting this way for >> So so let me putting this way for >> So so let me putting this way for example models over the developments of example models over the developments of example models over the developments of all different kind of robotic all different kind of robotic all different kind of robotic applications has been a very important applications has been a very important applications has been a very important cornerstones. Yeah. If you look at all cornerstones. Yeah. If you look at all cornerstones. Yeah. If you look at all the existing robotic applications like the existing robotic applications like the existing robotic applications like plane, drones, Roomba or even for plane, drones, Roomba or even for plane, drones, Roomba or even for quadripad robots, bipedal robots, model quadripad robots, bipedal robots, model quadripad robots, bipedal robots, model has been the way has been the way has been the way >> for them to actually work and be able to >> for them to actually work and be able to >> for them to actually work and be able to transfer from simulation to the real transfer from simulation to the real transfer from simulation to the real environment. environment. environment. >> I see. >> I see. >> I see. >> But if you look at those locomotion >> But if you look at those locomotion >> But if you look at those locomotion robots like quadriped robots, bipedal robots like quadriped robots, bipedal robots like quadriped robots, bipedal robots, they can walking on snows, they robots, they can walking on snows, they robots, they can walking on snows, they can walking on bushes. But you don't can walking on bushes. But you don't can walking on bushes. But you don't need to have a simulator. They can need to have a simulator. They can need to have a simulator. They can simulate all the bushes and snows like simulate all the bushes and snows like simulate all the bushes and snows like very precisely. you needs to have a very precisely. you needs to have a very precisely. you needs to have a simulation that capture the essential simulation that capture the essential simulation that capture the essential structure of the problem and do a whole structure of the problem and do a whole structure of the problem and do a whole different kind of randomizations inside different kind of randomizations inside different kind of randomizations inside the digital environment. So that is what the digital environment. So that is what the digital environment. So that is what we're aiming for. So basically with we're aiming for. So basically with we're aiming for. So basically with Synenix and together with word labaps Synenix and together with word labaps Synenix and together with word labaps we're trying to investigate like what is we're trying to investigate like what is we're trying to investigate like what is the level of fidelity we needs to model the level of fidelity we needs to model the level of fidelity we needs to model the mass massive worlds besides the the mass massive worlds besides the the mass massive worlds besides the robots I said we'll be able to transfer robots I said we'll be able to transfer robots I said we'll be able to transfer the robotic systems training the the robotic systems training the the robotic systems training the simulated environment digital worlds simulated environment digital worlds simulated environment digital worlds back into the real scenarios. As an back into the real scenarios. As an back into the real scenarios. As an investor, I've heard other researchers investor, I've heard other researchers investor, I've heard other researchers say like Sergey Lavine say simulation say like Sergey Lavine say simulation say like Sergey Lavine say simulation will always eventually deviate from the will always eventually deviate from the will always eventually deviate from the physical world and real world data physical world and real world data physical world and real world data collection is absolutely critical. And collection is absolutely critical. And collection is absolutely critical. And so maybe talk a little bit about like so maybe talk a little bit about like so maybe talk a little bit about like the viability of this approach where the viability of this approach where the viability of this approach where simulation is a cornerstone as opposed simulation is a cornerstone as opposed simulation is a cornerstone as opposed to some other approach. to some other approach. to some other approach. >> So they don't contradict with each >> So they don't contradict with each >> So they don't contradict with each other. So if you think about the other. So if you think about the other. So if you think about the simulation, simulation is essentially simulation, simulation is essentially simulation, simulation is essentially trying to predict how the environment is trying to predict how the environment is trying to predict how the environment is going going to change when you apply the going going to change when you apply the going going to change when you apply the actions and this is essentially a model actions and this is essentially a model actions and this is essentially a model of the worlds. It doesn't necessarily of the worlds. It doesn't necessarily of the worlds. It doesn't necessarily have to be pure physics. It can be a have to be pure physics. It can be a have to be pure physics. It can be a combination between both physics and combination between both physics and combination between both physics and also learning. We are collecting real also learning. We are collecting real also learning. We are collecting real world data. We will be using those real world data. We will be using those real world data. We will be using those real world data. It just at different stages world data. It just at different stages world data. It just at different stages of this like a data fly. Maybe at the of this like a data fly. Maybe at the of this like a data fly. Maybe at the very beginning we have stronger very beginning we have stronger very beginning we have stronger emphasize on we have more physics to emphasize on we have more physics to emphasize on we have more physics to making sure we have the right making sure we have the right making sure we have the right consistency and right structure for us consistency and right structure for us consistency and right structure for us to learn the uh the the worlds for us to to learn the uh the the worlds for us to to learn the uh the the worlds for us to train the robot policies. But as we train the robot policies. But as we train the robot policies. But as we accumulate more and more data both accumulate more and more data both accumulate more and more data both through data collection and also through through data collection and also through through data collection and also through the collaboration with our clients we'll the collaboration with our clients we'll the collaboration with our clients we'll have the data that will be moving have the data that will be moving have the data that will be moving towards more towards more learning based towards more towards more learning based towards more towards more learning based like modeling of the environments. So like modeling of the environments. So like modeling of the environments. So this kind of transition and also this this kind of transition and also this this kind of transition and also this kind of data flow is really an enabling kind of data flow is really an enabling kind of data flow is really an enabling factors of both getting the best of both factors of both getting the best of both factors of both getting the best of both physics and the and and geometry and physics and the and and geometry and physics and the and and geometry and consistency as well as all the power and consistency as well as all the power and consistency as well as all the power and magics from the data and compute. magics from the data and compute. magics from the data and compute. >> I want to add to this and be slightly >> I want to add to this and be slightly >> I want to add to this and be slightly philosophical here is there isn't a a philosophical here is there isn't a a philosophical here is there isn't a a binary choice between simulation or no binary choice between simulation or no binary choice between simulation or no simulation. All this come um in together simulation. All this come um in together simulation. All this come um in together um to to make robotics work. Think about um to to make robotics work. Think about um to to make robotics work. Think about human intelligence. We do a lot of human intelligence. We do a lot of human intelligence. We do a lot of simulation in our head. You know why? simulation in our head. You know why? simulation in our head. You know why? There's a very important role simulation There's a very important role simulation There's a very important role simulation plays that real world data doesn't play plays that real world data doesn't play plays that real world data doesn't play which is counterfactual reasoning is which is counterfactual reasoning is which is counterfactual reasoning is that you play out events that c hasn't that you play out events that c hasn't that you play out events that c hasn't happened or cannot happen or you don't happened or cannot happen or you don't happened or cannot happen or you don't have enough data to make it happen in have enough data to make it happen in have enough data to make it happen in real world. And while you play it out, real world. And while you play it out, real world. And while you play it out, you learn how to act in it. Humans do you learn how to act in it. Humans do you learn how to act in it. Humans do this all the time. We probably don't, this all the time. We probably don't, this all the time. We probably don't, you know, we just I know you were at you know, we just I know you were at you know, we just I know you were at World Cups. World Cups. World Cups. >> I was at the World. >> I was at the World. >> I was at the World. >> Congratulations to Spain winning. I'm >> Congratulations to Spain winning. I'm >> Congratulations to Spain winning. I'm sure in the planning of every game there sure in the planning of every game there sure in the planning of every game there is simulation, whether it's digital or is simulation, whether it's digital or is simulation, whether it's digital or on the on the whiteboard or whatever, on the on the whiteboard or whatever, on the on the whiteboard or whatever, that simulation, the role simulation that simulation, the role simulation that simulation, the role simulation plays is counterfactual reasoning. And plays is counterfactual reasoning. And plays is counterfactual reasoning. And that's really important in robotics that's really important in robotics that's really important in robotics because we just do not have cannot because we just do not have cannot because we just do not have cannot possibly have enough real world data for possibly have enough real world data for possibly have enough real world data for that. Here's a real life uh example the that. Here's a real life uh example the that. Here's a real life uh example the the industry of self-driving cars. Whimo the industry of self-driving cars. Whimo the industry of self-driving cars. Whimo has officially said they use billions of has officially said they use billions of has officially said they use billions of hours of simulation hours of simulation hours of simulation >> and and actually Whimo is more >> and and actually Whimo is more >> and and actually Whimo is more simulationheavy than just real world simulationheavy than just real world simulationheavy than just real world data heavy. So these are real examples data heavy. So these are real examples data heavy. So these are real examples and and as you know Martin Andrew too and and as you know Martin Andrew too and and as you know Martin Andrew too cars are the simplest kind of robots. cars are the simplest kind of robots. cars are the simplest kind of robots. >> Yeah. >> Yeah. >> Yeah. >> Yeah. So so clearly simulation plays a >> Yeah. So so clearly simulation plays a >> Yeah. So so clearly simulation plays a huge role in robotic learning. huge role in robotic learning. huge role in robotic learning. >> I also want to add to that. So like >> I also want to add to that. So like >> I also want to add to that. So like there are if you put things more there are if you put things more there are if you put things more specific simulation can provide two specific simulation can provide two specific simulation can provide two levels of benefits. The first one is levels of benefits. The first one is levels of benefits. The first one is reliability and the second one is reliability and the second one is reliability and the second one is efficiency. So for reliability, if efficiency. So for reliability, if efficiency. So for reliability, if you're thinking about a robotic system you're thinking about a robotic system you're thinking about a robotic system working reliable in the real working reliable in the real working reliable in the real environments, you need data to provide environments, you need data to provide environments, you need data to provide systematic coverage of all the state systematic coverage of all the state systematic coverage of all the state space and the variations that robots space and the variations that robots space and the variations that robots might encounter. That's how you can might encounter. That's how you can might encounter. That's how you can learn of how that is robust. So with learn of how that is robust. So with learn of how that is robust. So with simulation, you can do systematic simulation, you can do systematic simulation, you can do systematic randomizations and control and the randomizations and control and the randomizations and control and the variations of lighting, frictions, variations of lighting, frictions, variations of lighting, frictions, geometries, object types and also all geometries, object types and also all geometries, object types and also all different kind of physical parameters to different kind of physical parameters to different kind of physical parameters to making sure you have sufficient coverage making sure you have sufficient coverage making sure you have sufficient coverage of the state space. So this is what can of the state space. So this is what can of the state space. So this is what can give the robotic systems reliability and give the robotic systems reliability and give the robotic systems reliability and second is about efficiency. So right now second is about efficiency. So right now second is about efficiency. So right now many people are doing tele operation and many people are doing tele operation and many people are doing tele operation and if you look at many of the television if you look at many of the television if you look at many of the television device imagining all the actual device imagining all the actual device imagining all the actual skeletons you are using you're actually skeletons you are using you're actually skeletons you are using you're actually collecting the data at a speed that is collecting the data at a speed that is collecting the data at a speed that is actually slower than human actually actually slower than human actually actually slower than human actually doing the task. doing the task. doing the task. >> But for many of our clients human speed >> But for many of our clients human speed >> But for many of our clients human speed to them is not good enough. They want to them is not good enough. They want to them is not good enough. They want faster than human speeds. Yeah. faster than human speeds. Yeah. faster than human speeds. Yeah. >> So for the robot to like move faster, >> So for the robot to like move faster, >> So for the robot to like move faster, it's not as simple as just drive the it's not as simple as just drive the it's not as simple as just drive the robot faster because the gravity doesn't robot faster because the gravity doesn't robot faster because the gravity doesn't change. change. change. >> But in simulation, you can do systematic >> But in simulation, you can do systematic >> But in simulation, you can do systematic speed up of the robots behaviors to speed up of the robots behaviors to speed up of the robots behaviors to train the robots such that it considers train the robots such that it considers train the robots such that it considers all the dynamics changes of the all the dynamics changes of the all the dynamics changes of the environments. So this is what's can give environments. So this is what's can give environments. So this is what's can give like our clients for them efficiency. So like our clients for them efficiency. So like our clients for them efficiency. So both for the reliability and efficiency both for the reliability and efficiency both for the reliability and efficiency though there are some kind of like a though there are some kind of like a though there are some kind of like a very unique like values where simulation very unique like values where simulation very unique like values where simulation can provide. You've talked about the can provide. You've talked about the can provide. You've talked about the technology and the platform, what it technology and the platform, what it technology and the platform, what it does. Maybe talk about the specific use does. Maybe talk about the specific use does. Maybe talk about the specific use cases people use it for. cases people use it for. cases people use it for. >> There are essential like two specific >> There are essential like two specific >> There are essential like two specific use cases especially around both use cases especially around both use cases especially around both training and also around evaluations. training and also around evaluations. training and also around evaluations. >> Okay. >> Okay. >> Okay. >> Um starting from the evaluations. >> Um starting from the evaluations. >> Um starting from the evaluations. >> So evaluation is something like people >> So evaluation is something like people >> So evaluation is something like people often overlooked in the robotics often overlooked in the robotics often overlooked in the robotics >> but if you are tuning like robotic >> but if you are tuning like robotic >> but if you are tuning like robotic models you have to know how well it models you have to know how well it models you have to know how well it works and that is the only source of works and that is the only source of works and that is the only source of information for you to iterate. information for you to iterate. information for you to iterate. >> Yeah. By the way, a lot of a lot every >> Yeah. By the way, a lot of a lot every >> Yeah. By the way, a lot of a lot every every AI person really understands what every AI person really understands what every AI person really understands what eval are and uses it all the time. NonAI eval are and uses it all the time. NonAI eval are and uses it all the time. NonAI people, it often means something a people, it often means something a people, it often means something a little different. So maybe it's even little different. So maybe it's even little different. So maybe it's even worth just describing specifically what worth just describing specifically what worth just describing specifically what you mean by evaluation. you mean by evaluation. you mean by evaluation. >> Okay. So what I mean by evaluation is >> Okay. So what I mean by evaluation is >> Okay. So what I mean by evaluation is you'll be able to understand for this you'll be able to understand for this you'll be able to understand for this specific checkpoints how well does it specific checkpoints how well does it specific checkpoints how well does it perform? Does it perform for example 95% perform? Does it perform for example 95% perform? Does it perform for example 95% of the time or 99.9% of the time or 99.9% of the time or 99.9% of the time? And the key criteria people of the time? And the key criteria people of the time? And the key criteria people use in industry is how long does it use in industry is how long does it use in industry is how long does it take? How long in walk work walk time take? How long in walk work walk time take? How long in walk work walk time does it take for you to distinguish does it take for you to distinguish does it take for you to distinguish between a checkpoints that is 90% from a between a checkpoints that is 90% from a between a checkpoints that is 90% from a checkpoint that is 90 92 points. And if checkpoint that is 90 92 points. And if checkpoint that is 90 92 points. And if you only do that in the real environment you only do that in the real environment you only do that in the real environment that's just takes so long for you to do that's just takes so long for you to do that's just takes so long for you to do the distinguishment. And if you really the distinguishment. And if you really the distinguishment. And if you really think about also the robotic evaluations think about also the robotic evaluations think about also the robotic evaluations right now people are doing in the real right now people are doing in the real right now people are doing in the real environments the iteration speeds is environments the iteration speeds is environments the iteration speeds is multiple orders of magnitude slower than multiple orders of magnitude slower than multiple orders of magnitude slower than iterations of those language models. iterations of those language models. iterations of those language models. >> Yeah. >> Yeah. >> Yeah. >> So not only is like the robotic tasks >> So not only is like the robotic tasks >> So not only is like the robotic tasks very varied very diverse. very varied very diverse. very varied very diverse. >> Oh yeah because like you actually have >> Oh yeah because like you actually have >> Oh yeah because like you actually have to do the thing right to do the thing right to do the thing right atoms have to move through space. atoms have to move through space. atoms have to move through space. [laughter] [laughter] [laughter] >> Exactly. But only there are >> Exactly. But only there are >> Exactly. But only there are >> the laws of physics have to do. >> the laws of physics have to do. >> the laws of physics have to do. >> Have you watched those robotics videos? >> Have you watched those robotics videos? >> Have you watched those robotics videos? Every video has like 10 x 8x because it Every video has like 10 x 8x because it Every video has like 10 x 8x because it moves so slowly. moves so slowly. moves so slowly. >> Exactly. So not only it's slow, it's >> Exactly. So not only it's slow, it's >> Exactly. So not only it's slow, it's dangerous, it's costly, but at the same dangerous, it's costly, but at the same dangerous, it's costly, but at the same time the speeds is also like multiple time the speeds is also like multiple time the speeds is also like multiple orders of magnitudes like slower. So orders of magnitudes like slower. So orders of magnitudes like slower. So some of our clients actually needs this some of our clients actually needs this some of our clients actually needs this digital environment that's can be used digital environment that's can be used digital environment that's can be used to evaluate their robotic like systems to evaluate their robotic like systems to evaluate their robotic like systems and because our digital environment has and because our digital environment has and because our digital environment has proven alignments with the real world. proven alignments with the real world. proven alignments with the real world. So meaning whatever happens in the sim So meaning whatever happens in the sim So meaning whatever happens in the sim is also likely to happen in the real is also likely to happen in the real is also likely to happen in the real environment. If a checkpoint is working environment. If a checkpoint is working environment. If a checkpoint is working better in the simulation is also highly better in the simulation is also highly better in the simulation is also highly likely to also work better in the real likely to also work better in the real likely to also work better in the real environments as we have also been environments as we have also been environments as we have also been discussed in the blog post. So that discussed in the blog post. So that discussed in the blog post. So that actually give our clients very strong actually give our clients very strong actually give our clients very strong confidence in actually using the data confidence in actually using the data confidence in actually using the data using the signal from the digital using the signal from the digital using the signal from the digital environment to do scalable safe and much environment to do scalable safe and much environment to do scalable safe and much faster evaluations of their robotic faster evaluations of their robotic faster evaluations of their robotic systems. Great. So that is on the systems. Great. So that is on the systems. Great. So that is on the evaluation then on the training. So on evaluation then on the training. So on evaluation then on the training. So on the training sides, so basically like I the training sides, so basically like I the training sides, so basically like I also mentioned it's about also mentioned it's about also mentioned it's about controllability. So you want to control controllability. So you want to control controllability. So you want to control all the different possible variations of all the different possible variations of all the different possible variations of states, parameters, lighting, frictions, states, parameters, lighting, frictions, states, parameters, lighting, frictions, physical parameters like even object physical parameters like even object physical parameters like even object geometry, object types. So you want to geometry, object types. So you want to geometry, object types. So you want to making sure you have sufficient coverage making sure you have sufficient coverage making sure you have sufficient coverage of all different kind of scenarios such of all different kind of scenarios such of all different kind of scenarios such that you'll be able to generate like a that you'll be able to generate like a that you'll be able to generate like a informative data for your robots to be informative data for your robots to be informative data for your robots to be robust. And this is just going to be so robust. And this is just going to be so robust. And this is just going to be so hard to do just in the real environments hard to do just in the real environments hard to do just in the real environments like we discussed. If you do tele like we discussed. If you do tele like we discussed. If you do tele operation, the speed at which you're operation, the speed at which you're operation, the speed at which you're collecting data is slow. You're also collecting data is slow. You're also collecting data is slow. You're also limited by like how many robots you limited by like how many robots you limited by like how many robots you have, how many tele operation device you have, how many tele operation device you have, how many tele operation device you have. There's like a whole different have. There's like a whole different have. There's like a whole different kind of like challenges around all the kind of like challenges around all the kind of like challenges around all the data operations around it. But in data operations around it. But in data operations around it. But in simulation everything can be simulation everything can be simulation everything can be controllable, everything can be controllable, everything can be controllable, everything can be systematic and everything can be systematic and everything can be systematic and everything can be understand at a level where you know understand at a level where you know understand at a level where you know exactly and making claims about exactly exactly and making claims about exactly exactly and making claims about exactly what distribution you have covered to what distribution you have covered to what distribution you have covered to develop confidence about within that develop confidence about within that develop confidence about within that distribution. we know the robot will distribution. we know the robot will distribution. we know the robot will work. So those kind of confidence and work. So those kind of confidence and work. So those kind of confidence and efficiency and scalability is something efficiency and scalability is something efficiency and scalability is something that our clients also value to use our that our clients also value to use our that our clients also value to use our like a digital words for the training of like a digital words for the training of like a digital words for the training of robotic systems. robotic systems. robotic systems. >> Here's the crazy thing even before uh >> Here's the crazy thing even before uh >> Here's the crazy thing even before uh Synix and we are talking our inbound Synix and we are talking our inbound Synix and we are talking our inbound customers for Marble were already seeing customers for Marble were already seeing customers for Marble were already seeing this kind of demands. we just cannot this kind of demands. we just cannot this kind of demands. we just cannot serve these customers. But we are serve these customers. But we are serve these customers. But we are already getting a lot of phone calls already getting a lot of phone calls already getting a lot of phone calls from from robotics early stage robotics from from robotics early stage robotics from from robotics early stage robotics companies who are developing their companies who are developing their companies who are developing their models all the way to downstream very models all the way to downstream very models all the way to downstream very pragmatic use cases and uh we're seeing pragmatic use cases and uh we're seeing pragmatic use cases and uh we're seeing these uh these needs. When people hear these uh these needs. When people hear these uh these needs. When people hear you're going into robotics, what they're you're going into robotics, what they're you're going into robotics, what they're going to envision is you're pulling out going to envision is you're pulling out going to envision is you're pulling out a 3D printer and you're going to be a 3D printer and you're going to be a 3D printer and you're going to be making hardware and then you're going to making hardware and then you're going to making hardware and then you're going to be programming the brain of a robot and be programming the brain of a robot and be programming the brain of a robot and sticking it in the robot and then you've sticking it in the robot and then you've sticking it in the robot and then you've got a robot. And I don't think that's got a robot. And I don't think that's got a robot. And I don't think that's what you guys are talking about here. So what you guys are talking about here. So what you guys are talking about here. So maybe talk about what you know where maybe talk about what you know where maybe talk about what you know where this fits in the life cycle of creating this fits in the life cycle of creating this fits in the life cycle of creating a robot and like where you will end and a robot and like where you will end and a robot and like where you will end and where the rest of the ecosystem will where the rest of the ecosystem will where the rest of the ecosystem will will begin. So what we have been will begin. So what we have been will begin. So what we have been building you can imagine is a building you can imagine is a building you can imagine is a infrastructure like with the softwares infrastructure like with the softwares infrastructure like with the softwares around this infrastructures for people around this infrastructures for people around this infrastructures for people to for them build words such that robot to for them build words such that robot to for them build words such that robot can learn and evaluate and this can learn and evaluate and this can learn and evaluate and this infrastructures is naturally model infrastructures is naturally model infrastructures is naturally model agnostic and embodiment agnostic. So I agnostic and embodiment agnostic. So I agnostic and embodiment agnostic. So I just want to be very clear just because just want to be very clear just because just want to be very clear just because this is actually a very subtle for you this is actually a very subtle for you this is actually a very subtle for you it's obvious but it's a very subtle it's obvious but it's a very subtle it's obvious but it's a very subtle point which is um point which is um point which is um from what you said that's not building a from what you said that's not building a from what you said that's not building a robot it's building an environment which robot it's building an environment which robot it's building an environment which another company can place their robot another company can place their robot another company can place their robot brain brain brain >> exactly >> exactly >> exactly >> to navigate and to learn. >> to navigate and to learn. >> to navigate and to learn. >> Yeah. So for our customers right now >> Yeah. So for our customers right now >> Yeah. So for our customers right now they have all different kind of robots. they have all different kind of robots. they have all different kind of robots. Some are using for them single robot arm Some are using for them single robot arm Some are using for them single robot arm some are using bio some are using a some are using bio some are using a some are using bio some are using a fixed arm. Some are using like a mobile fixed arm. Some are using like a mobile fixed arm. Some are using like a mobile manipulators. Some using grippers, some manipulators. Some using grippers, some manipulators. Some using grippers, some are using some more elaborate versions are using some more elaborate versions are using some more elaborate versions of the end factors. So our platform of the end factors. So our platform of the end factors. So our platform right now is just naturally embodiment right now is just naturally embodiment right now is just naturally embodiment agnostic. We can very easily integrate agnostic. We can very easily integrate agnostic. We can very easily integrate different kind of robotic embodiment be different kind of robotic embodiment be different kind of robotic embodiment be able to put them into the works we able to put them into the works we able to put them into the works we generated with digitalized such as we generated with digitalized such as we generated with digitalized such as we will be able to give those individual will be able to give those individual will be able to give those individual robots capabilities of doing the right robots capabilities of doing the right robots capabilities of doing the right tasks and at the right levels of tasks and at the right levels of tasks and at the right levels of reliability and efficiency in the real reliability and efficiency in the real reliability and efficiency in the real environments. environments. environments. >> Yeah. >> Yeah. >> Yeah. >> And we are also for example model >> And we are also for example model >> And we are also for example model agnostic. So we can just using the data agnostic. So we can just using the data agnostic. So we can just using the data generated by our words to train generated by our words to train generated by our words to train different models either from scratch or different models either from scratch or different models either from scratch or doing post training of existing doing post training of existing doing post training of existing foundation models like vision language foundation models like vision language foundation models like vision language action models or word action models. So action models or word action models. So action models or word action models. So to us it doesn't matter we just want to to us it doesn't matter we just want to to us it doesn't matter we just want to making sure we have the infrastructure making sure we have the infrastructure making sure we have the infrastructure we have all the words such as the robot we have all the words such as the robot we have all the words such as the robot can work reliably in the real can work reliably in the real can work reliably in the real environment. you know, you you you have environment. you know, you you you have environment. you know, you you you have uh told me um that you think uh a lot of uh told me um that you think uh a lot of uh told me um that you think uh a lot of the predictions around humanoids were a the predictions around humanoids were a the predictions around humanoids were a little bit aggressive and we're likely little bit aggressive and we're likely little bit aggressive and we're likely to see more constrained rollouts like to see more constrained rollouts like to see more constrained rollouts like warehouses or whatever. Can you talk a warehouses or whatever. Can you talk a warehouses or whatever. Can you talk a little bit about that and like how that little bit about that and like how that little bit about that and like how that impacts what you're going to be tackling impacts what you're going to be tackling impacts what you're going to be tackling here at uh the like world labs? here at uh the like world labs? here at uh the like world labs? >> So that's a very good question. So if >> So that's a very good question. So if >> So that's a very good question. So if you look at for example all the uh you look at for example all the uh you look at for example all the uh progressions of robotic applications in progressions of robotic applications in progressions of robotic applications in the real environments it has always the real environments it has always the real environments it has always followed the trend from going from fully followed the trend from going from fully followed the trend from going from fully structured environments into structured environments into structured environments into semiructured environments and then into semiructured environments and then into semiructured environments and then into unstructured environments. unstructured environments. unstructured environments. >> For fully structured environments what >> For fully structured environments what >> For fully structured environments what do we mean that you have knowledge and do we mean that you have knowledge and do we mean that you have knowledge and control over all the configurations control over all the configurations control over all the configurations within the environments within the environments within the environments >> like factories >> like factories >> like factories >> like factories or for car manufacturing >> like factories or for car manufacturing >> like factories or for car manufacturing mice those has been automated for mice those has been automated for mice those has been automated for decades. decades. decades. >> Yeah. Yeah. Yeah. And then you have for >> Yeah. Yeah. Yeah. And then you have for >> Yeah. Yeah. Yeah. And then you have for example semiructured environments which example semiructured environments which example semiructured environments which you have certain controls over the you have certain controls over the you have certain controls over the environments for example like the Amazon environments for example like the Amazon environments for example like the Amazon warehouses or for example like warehouses or for example like warehouses or for example like restaurants hotels where you have restaurants hotels where you have restaurants hotels where you have certain control over the environment to certain control over the environment to certain control over the environment to just make the task easier for your just make the task easier for your just make the task easier for your robots but there are obviously many robots but there are obviously many robots but there are obviously many other like objects or for example other like objects or for example other like objects or for example clothes those are the object you don't clothes those are the object you don't clothes those are the object you don't have control and then for the have control and then for the have control and then for the unstructured environments it's like your unstructured environments it's like your unstructured environments it's like your home and my host home and my host home and my host >> those is I would say the the grand >> those is I would say the the grand >> those is I would say the the grand challenge Especially Especially my challenge Especially Especially my challenge Especially Especially my house, trust me. Three dogs, [laughter] house, trust me. Three dogs, [laughter] house, trust me. Three dogs, [laughter] >> five-year-old. >> five-year-old. >> five-year-old. >> Yes, dogs. >> Yes, dogs. >> Yes, dogs. >> Exactly. >> Exactly. >> Exactly. >> If you're thinking about where does the >> If you're thinking about where does the >> If you're thinking about where does the robustness coming from, robustness robustness coming from, robustness robustness coming from, robustness coming from a sufficient coverage of the coming from a sufficient coverage of the coming from a sufficient coverage of the scenarios that robots might encounter. scenarios that robots might encounter. scenarios that robots might encounter. So, it's so much easier and more So, it's so much easier and more So, it's so much easier and more approachable at least like right now to approachable at least like right now to approachable at least like right now to focus more on the semiructured focus more on the semiructured focus more on the semiructured environments before we move on to fully environments before we move on to fully environments before we move on to fully unstructured environments. So, we will unstructured environments. So, we will unstructured environments. So, we will move into that direction. It's just we move into that direction. It's just we move into that direction. It's just we want to uh take a more sustainable and want to uh take a more sustainable and want to uh take a more sustainable and more realistic approach towards it. more realistic approach towards it. more realistic approach towards it. >> I think your point here is that >> I think your point here is that >> I think your point here is that humanoids mimics human body and humanoids mimics human body and humanoids mimics human body and evolution has optimized human body for evolution has optimized human body for evolution has optimized human body for unstructured environment. unstructured environment. unstructured environment. And so our fingers, our legs are not the And so our fingers, our legs are not the And so our fingers, our legs are not the best apparatus to do one thing. For best apparatus to do one thing. For best apparatus to do one thing. For example, if if if our only goal as a example, if if if our only goal as a example, if if if our only goal as a species is to climb trees, we will not species is to climb trees, we will not species is to climb trees, we will not have this body necessarily, right? So, have this body necessarily, right? So, have this body necessarily, right? So, we'll have different kind of fingers. we'll have different kind of fingers. we'll have different kind of fingers. But what humans end up having are But what humans end up having are But what humans end up having are involve evolved into is this this body involve evolved into is this this body involve evolved into is this this body shape that can be very general but not shape that can be very general but not shape that can be very general but not necessarily best at everything. And that necessarily best at everything. And that necessarily best at everything. And that is for the survival of unstructured is for the survival of unstructured is for the survival of unstructured environment. But but but from a business environment. But but but from a business environment. But but but from a business point of view, from a pragmatic point of view, from a pragmatic point of view, from a pragmatic um technology point of view that this um technology point of view that this um technology point of view that this unstructured environment and a unstructured environment and a unstructured environment and a generalized body is actually the hardest generalized body is actually the hardest generalized body is actually the hardest problem to solve. It's not necessarily problem to solve. It's not necessarily problem to solve. It's not necessarily even the right way to solve the problem. even the right way to solve the problem. even the right way to solve the problem. It's we specialize. So we take more more It's we specialize. So we take more more It's we specialize. So we take more more specialized body to solve a narrower specialized body to solve a narrower specialized body to solve a narrower problem. But the challenge for scenics problem. But the challenge for scenics problem. But the challenge for scenics is is that to be more body agnostic so is is that to be more body agnostic so is is that to be more body agnostic so that their infrastructure can serve that their infrastructure can serve that their infrastructure can serve different bodies and different different bodies and different different bodies and different semistructured environments. semistructured environments. semistructured environments. >> You know a common lens to look at >> You know a common lens to look at >> You know a common lens to look at exactly this question is an economic exactly this question is an economic exactly this question is an economic lens right which is um lens right which is um lens right which is um >> well you compare it to like generative >> well you compare it to like generative >> well you compare it to like generative LLMs where they can create uh pros or LLMs where they can create uh pros or LLMs where they can create uh pros or code 10,000 times faster than a human code 10,000 times faster than a human code 10,000 times faster than a human being a bunch cheaper than a human being a bunch cheaper than a human being a bunch cheaper than a human being. So the economic case makes sense being. So the economic case makes sense being. So the economic case makes sense because our brains aren't very efficient because our brains aren't very efficient because our brains aren't very efficient at that. However, our brains and our at that. However, our brains and our at that. However, our brains and our bodies are very efficient at 3D bodies are very efficient at 3D bodies are very efficient at 3D navigation, right? You know, like moving navigation, right? You know, like moving navigation, right? You know, like moving through the world or picking things up. through the world or picking things up. through the world or picking things up. And so this is just a prediction And so this is just a prediction And so this is just a prediction question, but do you believe we'll ever question, but do you believe we'll ever question, but do you believe we'll ever be able to build robots, at least in the be able to build robots, at least in the be able to build robots, at least in the foreseeable future, that have the power foreseeable future, that have the power foreseeable future, that have the power efficiency of a human being when it efficiency of a human being when it efficiency of a human being when it comes to to menial tasks? So let's say comes to to menial tasks? So let's say comes to to menial tasks? So let's say just basically you know minimum wage or just basically you know minimum wage or just basically you know minimum wage or something like that like how far away something like that like how far away something like that like how far away are we from this? Is this like 5 years are we from this? Is this like 5 years are we from this? Is this like 5 years or this is like never I think it's going or this is like never I think it's going or this is like never I think it's going to take a very long time. So if you to take a very long time. So if you to take a very long time. So if you really think about like a robots in the really think about like a robots in the really think about like a robots in the real environment in the end it it will real environment in the end it it will real environment in the end it it will always be a system. So every working always be a system. So every working always be a system. So every working robot in the real environment is a robot in the real environment is a robot in the real environment is a system work. It's you need to be very system work. It's you need to be very system work. It's you need to be very mindful and thoughtful about how the mindful and thoughtful about how the mindful and thoughtful about how the system are coming together. the system are coming together. the system are coming together. the hardware, the software, the brain, even hardware, the software, the brain, even hardware, the software, the brain, even like to the details of for example like to the details of for example like to the details of for example what's the friction coefficients of your what's the friction coefficients of your what's the friction coefficients of your of your fingers. So there's a lot of of your fingers. So there's a lot of of your fingers. So there's a lot of things you have to consider to make things you have to consider to make things you have to consider to make these things uh a reality and it will these things uh a reality and it will these things uh a reality and it will take iterations. But what I am excited take iterations. But what I am excited take iterations. But what I am excited about is that I have always been at the about is that I have always been at the about is that I have always been at the state of the arts of robot learning and state of the arts of robot learning and state of the arts of robot learning and also trying to push the state of the art also trying to push the state of the art also trying to push the state of the art forward. Yeah. forward. Yeah. forward. Yeah. >> But the state-ofthe-arts always moving >> But the state-ofthe-arts always moving >> But the state-ofthe-arts always moving faster than I expected. So what's I'm faster than I expected. So what's I'm faster than I expected. So what's I'm focusing on and trying to investigate focusing on and trying to investigate focusing on and trying to investigate right now is very different from for right now is very different from for right now is very different from for them when I started my PhD. So this is a them when I started my PhD. So this is a them when I started my PhD. So this is a speak to how fast the whole ecosystem speak to how fast the whole ecosystem speak to how fast the whole ecosystem has been evolving and all the moving has been evolving and all the moving has been evolving and all the moving pieces started coming together or pieces started coming together or pieces started coming together or building this robotic systems but we building this robotic systems but we building this robotic systems but we also have to be like calibrated about also have to be like calibrated about also have to be like calibrated about our predictions. So we will see a lot of our predictions. So we will see a lot of our predictions. So we will see a lot of like a progress but to achieve for like a progress but to achieve for like a progress but to achieve for example human level efficiency and example human level efficiency and example human level efficiency and capabilities it will take uh longer. capabilities it will take uh longer. capabilities it will take uh longer. Martin, the hardest thing in today's AI Martin, the hardest thing in today's AI Martin, the hardest thing in today's AI is to have the right measured optimism, is to have the right measured optimism, is to have the right measured optimism, [laughter] [laughter] [laughter] >> right? >> right? >> right? >> That's right. It's totally true. Yeah. >> That's right. It's totally true. Yeah. >> That's right. It's totally true. Yeah. >> I mean, even LLM does not have human >> I mean, even LLM does not have human >> I mean, even LLM does not have human brain efficiency. Human brain operates brain efficiency. Human brain operates brain efficiency. Human brain operates on 30 watts. on 30 watts. on 30 watts. >> Yeah, that's true. >> Yeah, that's true. >> Yeah, that's true. >> That's so we are far from that. So, >> That's so we are far from that. So, >> That's so we are far from that. So, >> but that but I mean >> but that but I mean >> but that but I mean performance performance performance to power it may be close, right? in to power it may be close, right? in to power it may be close, right? in narrow task like software narrow task like software narrow task like software >> like generating an image or or software >> like generating an image or or software >> like generating an image or or software engineering than it is right. Yeah, I engineering than it is right. Yeah, I engineering than it is right. Yeah, I think so. I don't think we're anywhere think so. I don't think we're anywhere think so. I don't think we're anywhere close when it comes to robotics. Does close when it comes to robotics. Does close when it comes to robotics. Does this change how you think about your um this change how you think about your um this change how you think about your um like strategically the level of ambition like strategically the level of ambition like strategically the level of ambition that your team can go after? I mean does that your team can go after? I mean does that your team can go after? I mean does it is it changed that or is it still it is it changed that or is it still it is it changed that or is it still very much in line with what you expected very much in line with what you expected very much in line with what you expected to do when you started? to do when you started? to do when you started? >> It definitely changed the trajectories >> It definitely changed the trajectories >> It definitely changed the trajectories in a very profound manners. So we see a in a very profound manners. So we see a in a very profound manners. So we see a lot of unlock in be able to do this lot of unlock in be able to do this lot of unlock in be able to do this whole process do the modeling of the whole process do the modeling of the whole process do the modeling of the environments yeah much more efficient environments yeah much more efficient environments yeah much more efficient and much more scalable manners and much more scalable manners and much more scalable manners especially in partner together with word especially in partner together with word especially in partner together with word labaps and I also want to add to f if labaps and I also want to add to f if labaps and I also want to add to f if you think about for example the current you think about for example the current you think about for example the current states of the language models states of the language models states of the language models >> so those are models that's with >> so those are models that's with >> so those are models that's with incredible capabilities incredible capabilities incredible capabilities >> but still you don't just blind trust it >> but still you don't just blind trust it >> but still you don't just blind trust it to book your flight tickets or make your to book your flight tickets or make your to book your flight tickets or make your hotel reservations you still hopefully hotel reservations you still hopefully hotel reservations you still hopefully there's still a person who reading the there's still a person who reading the there's still a person who reading the output from those language models. output from those language models. output from those language models. >> Yeah. >> Yeah. >> Yeah. >> But that is very different from how >> But that is very different from how >> But that is very different from how people and we'll be using for them people and we'll be using for them people and we'll be using for them robotic models because for robotic robotic models because for robotic robotic models because for robotic models out of the box the robots has to models out of the box the robots has to models out of the box the robots has to work reliably in the real environment. work reliably in the real environment. work reliably in the real environment. >> So and we don't even have the data we >> So and we don't even have the data we >> So and we don't even have the data we don't even have all the necessary don't even have all the necessary don't even have all the necessary infrastructures around those for the infrastructures around those for the infrastructures around those for the robots to just out of box work reliable robots to just out of box work reliable robots to just out of box work reliable in the real environments. So for that in the real environments. So for that in the real environments. So for that reasons be able to create this digital reasons be able to create this digital reasons be able to create this digital worlds this scalable digital worlds worlds this scalable digital worlds worlds this scalable digital worlds where the robot can learn in and where the robot can learn in and where the robot can learn in and evaluates within that is going to unlock evaluates within that is going to unlock evaluates within that is going to unlock so much more potentials for be able to so much more potentials for be able to so much more potentials for be able to replace all the like a costly and unsafe replace all the like a costly and unsafe replace all the like a costly and unsafe data in the real environments with the data in the real environments with the data in the real environments with the data generated from the the the the data generated from the the the the data generated from the the the the words for the robots to be able to do words for the robots to be able to do words for the robots to be able to do scalable learning and evaluations. scalable learning and evaluations. scalable learning and evaluations. >> You know I've I've seen uh I've seen >> You know I've I've seen uh I've seen >> You know I've I've seen uh I've seen many of kind of these kind of many of kind of these kind of many of kind of these kind of integrations. they actually work very integrations. they actually work very integrations. they actually work very well at this stage when they have this well at this stage when they have this well at this stage when they have this much alignment which is great. Um but much alignment which is great. Um but much alignment which is great. Um but there's always like this question of do there's always like this question of do there's always like this question of do you integrate now into what's happening you integrate now into what's happening you integrate now into what's happening now or do you keep things quite separate now or do you keep things quite separate now or do you keep things quite separate and provide kind of like a long-term and provide kind of like a long-term and provide kind of like a long-term trajectory that will you know be trajectory that will you know be trajectory that will you know be realized you know in in in the year time realized you know in in in the year time realized you know in in in the year time frame. How are you thinking about this frame. How are you thinking about this frame. How are you thinking about this fee? Is this something that integrates fee? Is this something that integrates fee? Is this something that integrates right away or is this kind of a separate right away or is this kind of a separate right away or is this kind of a separate longer term? longer term? longer term? >> This is a great question. I think at >> This is a great question. I think at >> This is a great question. I think at this point you know Changi Sunonny this point you know Changi Sunonny this point you know Changi Sunonny Justin Ben and have been talking about Justin Ben and have been talking about Justin Ben and have been talking about this at this point we are going to take this at this point we are going to take this at this point we are going to take it thoughtfully we're not rushing to it thoughtfully we're not rushing to it thoughtfully we're not rushing to integrate integrate integrate everything from codebase to teams everything from codebase to teams everything from codebase to teams because I think Synex does have a very because I think Synex does have a very because I think Synex does have a very uh um wellthought and I wouldn't call it uh um wellthought and I wouldn't call it uh um wellthought and I wouldn't call it standalone completely but fairly uh standalone completely but fairly uh standalone completely but fairly uh contained contained contained um tech stack as well as their customers um tech stack as well as their customers um tech stack as well as their customers as well as the the the kind of products as well as the the the kind of products as well as the the the kind of products they're building. We're going to take they're building. We're going to take they're building. We're going to take time. We definitely will we already on time. We definitely will we already on time. We definitely will we already on the simulation side as well as the the the simulation side as well as the the the simulation side as well as the the potential base model um action condition potential base model um action condition potential base model um action condition model side. We already are starting to model side. We already are starting to model side. We already are starting to talk and also they are using Marvel as a talk and also they are using Marvel as a talk and also they are using Marvel as a internal customer. So we we will be internal customer. So we we will be internal customer. So we we will be integrating but uh we're not rushing to integrating but uh we're not rushing to integrating but uh we're not rushing to blend the team as like a full salad blend the team as like a full salad blend the team as like a full salad bowl. bowl. bowl. >> Yeah. >> Yeah. >> Yeah. >> How are you thinking about geographies >> How are you thinking about geographies >> How are you thinking about geographies with this with scenics move? Is it going with this with scenics move? Is it going with this with scenics move? Is it going to stay in the same place? to stay in the same place? to stay in the same place? >> We're going to Vundra is going to move. >> We're going to Vundra is going to move. >> We're going to Vundra is going to move. >> Oh well, welcome here. [laughter] Moving >> Oh well, welcome here. [laughter] Moving >> Oh well, welcome here. [laughter] Moving to San Francisco. Yeah, Florence and the to San Francisco. Yeah, Florence and the to San Francisco. Yeah, Florence and the Renaissance. Perfect. Renaissance. Perfect. Renaissance. Perfect. >> I think we our labs is officially >> I think we our labs is officially >> I think we our labs is officially becoming a by coastal company where the becoming a by coastal company where the becoming a by coastal company where the headquarter is in San Francisco. I've, headquarter is in San Francisco. I've, headquarter is in San Francisco. I've, you know, I live in Palo Alto. I feel you know, I live in Palo Alto. I feel you know, I live in Palo Alto. I feel like I'm in a [laughter] different like I'm in a [laughter] different like I'm in a [laughter] different state. We [gasps] we but uh we but um state. We [gasps] we but uh we but um state. We [gasps] we but uh we but um I'm actually excited that we're going to I'm actually excited that we're going to I'm actually excited that we're going to have an office in New York that can help have an office in New York that can help have an office in New York that can help us to attract talent on East Coast and us to attract talent on East Coast and us to attract talent on East Coast and also we have been talking about making also we have been talking about making also we have been talking about making sure that in both offices we set up the sure that in both offices we set up the sure that in both offices we set up the robots so that we get to robots so that we get to robots so that we get to basically test out and and and uh mature basically test out and and and uh mature basically test out and and and uh mature our engineering stack so that we can our engineering stack so that we can our engineering stack so that we can work with robots remotely because we work with robots remotely because we work with robots remotely because we have to do that for for our customers. have to do that for for our customers. have to do that for for our customers. Anyway, Anyway, Anyway, >> so may maybe just to be very concrete, >> so may maybe just to be very concrete, >> so may maybe just to be very concrete, Fay, maybe let's just pencil out like Fay, maybe let's just pencil out like Fay, maybe let's just pencil out like what is the the perfect success case in what is the the perfect success case in what is the the perfect success case in two years? Like what product do you two years? Like what product do you two years? Like what product do you have? Who's engaging with it? How do have? Who's engaging with it? How do have? Who's engaging with it? How do they use it? Just just the crisp like they use it? Just just the crisp like they use it? Just just the crisp like what? We're very happy that uh Synix what? We're very happy that uh Synix what? We're very happy that uh Synix team and World Apps team will have validated validated um um um customers in in a small number of customers in in a small number of customers in in a small number of important vertical use cases where our important vertical use cases where our important vertical use cases where our system, our infrastructure has proven to system, our infrastructure has proven to system, our infrastructure has proven to be truly beneficial to their automation be truly beneficial to their automation be truly beneficial to their automation needs And these customers became our needs And these customers became our needs And these customers became our lighthouse examples to scale our lighthouse examples to scale our lighthouse examples to scale our business. business. business. >> And how how early let's say someone >> And how how early let's say someone >> And how how early let's say someone listening to this listening to this listening to this uh is running a robotics company? How at uh is running a robotics company? How at uh is running a robotics company? How at what stage do they engage with world what stage do they engage with world what stage do they engage with world labs? Is it really early on? Is it labs? Is it really early on? Is it labs? Is it really early on? Is it somewhere in the middle? somewhere in the middle? somewhere in the middle? So right now um for our customers So right now um for our customers So right now um for our customers because we are building this kind of because we are building this kind of because we are building this kind of real to sim to real pipelines like where real to sim to real pipelines like where real to sim to real pipelines like where the simulation is essentially the words the simulation is essentially the words the simulation is essentially the words we're going to provide the training we're going to provide the training we're going to provide the training evaluation grounds. evaluation grounds. evaluation grounds. >> Some customers they needs only the real >> Some customers they needs only the real >> Some customers they needs only the real to sim part to sim part to sim part >> they want to digitalize the task they >> they want to digitalize the task they >> they want to digitalize the task they care about and be able to do the care about and be able to do the care about and be able to do the evaluations of their robotic systems. evaluations of their robotic systems. evaluations of their robotic systems. >> Some customers they need this real to >> Some customers they need this real to >> Some customers they need this real to sim to this entire pipeline such as they sim to this entire pipeline such as they sim to this entire pipeline such as they will be able to like have policies like will be able to like have policies like will be able to like have policies like running on their hardwares. So our running on their hardwares. So our running on their hardwares. So our platform is also designed in a way that platform is also designed in a way that platform is also designed in a way that is flexible depending on what our is flexible depending on what our is flexible depending on what our clients needs and at the same times uh clients needs and at the same times uh clients needs and at the same times uh the clients we are working with are the clients we are working with are the clients we are working with are actually pretty close to the deployments actually pretty close to the deployments actually pretty close to the deployments like a stage. So basically they are like a stage. So basically they are like a stage. So basically they are working on very very practical tasks. working on very very practical tasks. working on very very practical tasks. those tasks when replaced when we have a those tasks when replaced when we have a those tasks when replaced when we have a robotic solutions that are there can robotic solutions that are there can robotic solutions that are there can just create value immediately and they just create value immediately and they just create value immediately and they have like at least like uh tons or have like at least like uh tons or have like at least like uh tons or hundreds of like this kind of situations hundreds of like this kind of situations hundreds of like this kind of situations they are thinking about to do the they are thinking about to do the they are thinking about to do the automations for. So as like together automations for. So as like together automations for. So as like together with word labaps we'll be able to with word labaps we'll be able to with word labaps we'll be able to develop reliable solutions for those develop reliable solutions for those develop reliable solutions for those scenarios as we have already showed we scenarios as we have already showed we scenarios as we have already showed we have a number of scenarios already have a number of scenarios already have a number of scenarios already instantiated in our blog post and we'll instantiated in our blog post and we'll instantiated in our blog post and we'll be able to further our investigation to be able to further our investigation to be able to further our investigation to see how they can actually solve the key see how they can actually solve the key see how they can actually solve the key like like requirements and and and also like like requirements and and and also like like requirements and and and also constraints faced by the real world constraints faced by the real world constraints faced by the real world deployments. deployments. deployments. >> Great. I want to be very specific about >> Great. I want to be very specific about >> Great. I want to be very specific about this. Is it ever too late or too early this. Is it ever too late or too early this. Is it ever too late or too early to call World Labs if you're a robotics to call World Labs if you're a robotics to call World Labs if you're a robotics company? company? company? >> No. >> No. >> No. We we want everybody to call us. We want We we want everybody to call us. We want We we want everybody to call us. We want to learn about your use case. to learn about your use case. to learn about your use case. >> Wonderful. If you listen, if you're >> Wonderful. If you listen, if you're >> Wonderful. If you listen, if you're listening to this and you're anywhere listening to this and you're anywhere listening to this and you're anywhere close to a robotics project or robotics close to a robotics project or robotics close to a robotics project or robotics company, uh please track World Labs. company, uh please track World Labs. company, uh please track World Labs. >> Yes. Thank you. [laughter] >> Yes. Thank you. [laughter] >> Yes. Thank you. [laughter] >> Definitely open for business. >> Definitely open for business. >> Definitely open for business. >> We are open for business. All right. >> We are open for business. All right. >> We are open for business. All right. >> Not too early. >> Not too early. >> Not too early. >> All right. If you're doing robotics, >> All right. If you're doing robotics, >> All right. If you're doing robotics, call World Labs. Thank you both very call World Labs. Thank you both very call World Labs. Thank you both very much for coming. much for coming. much for coming. >> Thank you. >> Thank you. >> Thank you. [music]