It's been a pleasure to actually see the progress that's been made within World Labs in the past year as well. So, let's start by talking about what you're doing and why it's so important.
Yeah. Well, right now I wake up every day and think about just one thing and one thing only, which is spatial intelligence. That's the company that I co-founded about two years ago with a group of young technologists and it's called World Labs. So why is spatial intelligence so important and why do I say that it's the next frontier of AI? If you take a little bit of an evolutionary view of the development of intelligence, which starts more than half a billion years ago, the first thing that kicked off an evolutionary arms race of the development of nervous system is actually perception, not language. Language compared to perception is a very, very new form of intelligence. If you're generous and think about language, it's probably half a million years, probably not even that, but more than half a billion years ago, animals start to see light, start to touch the environment. So the tactile and visual sensing...
And do you put instinct in perception as well?
Instinct is a very interesting word. To me, instinct is a very fluffy word. So I'm not rejecting the word instinct, but physically, touch and seeing that kicked off the development of nervous system and that sets off the evolutionary arms race. So animals become more and more active interacting with its environment and more and more intelligent. All this is a long way of saying that the ability to understand, to reason, to interact with, and to navigate the real 3D, 4D physical world is as foundational as language intelligence, and the lynchpin technology is spatial intelligence. And that's why I believe this is the next generation or next frontier of AI. And this is what World Labs has been building towards.
So talk to us a little bit about Marble. What is Marble, and you launched that recently.
Right. So Marble is the name for our first generation of our spatial intelligence model. We tend to call it casually a world model, but the truth is that there's so many definitions of world model, it almost doesn't matter if you call it a world model or not. It's a spatial intelligence model, it's a frontier model. What Marble does is that it takes multimodal inputs. Whether it's just a sentence or it's a picture or it's a video or it's a few pictures or it's a simple 3D input, it doesn't matter. It's multimodal. And then it turns that prompt into a fully navigable, interactable world that is 3D, which is permanently consistent, which is very different from the video models you see these days, and it has the geometric structure to support whether you want to simulate robotic actions or code up a game and all that. So Marble is a frontier model. We released it about two months ago. It's still first generation. We're very excited to start this journey and right now it's still the state-of-the-art 3D generative world model.
And you know, there's a school of thought that says that without, if you just go with language as the model, we will not be able to get to AGI. Like you need to have an augmentation of the physicality of it. You of course believe that, but talk to us a little bit about that dimension. What's the big unlock that happens over time? The obvious use case is robotics, but even beyond that, like what are the things that we'll be able to do in five years with world models that we don't even think about today?
Well, first of all, you don't even have to wait for five years. We're already engaging with users and customers to play with the early days of our world model. We are seeing users using it to develop games. We're seeing VFX users and customers using it to work in virtual production settings, actually commercial virtual production settings. We are engaging with our robotics and simulation partners from big companies like Nvidia as well as startups and academic labs where they're using Marble as training environments for robots. We're seeing that architects and designers are using this for interior design and other kinds of design. We're also getting surprise use cases. For example, clinical researchers have come to us and expressed a lot of enthusiasm. It turned out a lot of psychiatric and mental health research as well as intervention requires immersive environments that's personalized in particular situations. For example, OCD patients. You know, some people are triggered by very specific situations. Personally, I'm triggered by dirty laundry, but I'm sure that people are triggered in different ways and these researchers are studying that. It's very hard for them to get their hands on these environments and Marble is just, within every time you prompt, within minutes you get many different kinds of environments that you can use. So this is actually a surprise use. We've also seen people with well-being and fitness training to personalize yoga environments or other environments. So, I think as Marble becomes better, we're still in the early days, we're going to see more and more horizontal use cases.
And so, now you've been in the AI industry for a very long time.
It's a nice way to talk about my age.
No, no, no. I mean like you're one of those people that hasn't just been like, you know, you're not just like, okay, I've looked at this for 3 years, like you've dedicated your life to AI. What surprised you the most as you built this company and as you were starting to think about building the spatial intelligence side?
That's a great question. I know G2, you're the same. We talk about, you know, even in our age, the most important thing is learning and this hunger for learning. Yes, I've been in this field for many, many years and one thing I would say is that I came to AI because of my very personal curiosity. I just love to ask that audacious question of what is intelligence and how to make machines intelligent, but it was very personal. Nobody cared about AI. The world didn't know how to spell AI and it was very fun. So, the past 10 years have completely changed. AI is not just personal for me. AI is civilizational. And that added a layer of sense of responsibility for me in terms of how I can contribute not only technologically and also as an educator, but also contribute to the healthy development of AI. What surprised me the most as an entrepreneur? A couple of things. One is the speed of development of AI this past few years is just breathtaking. I know everyone, whether we pretend or not, everyone deep in your heart is feeling that anxiety of there's just too much to read, too many blogs, too many news, too many model releases, and that sense of anxiety speaks of our time, is that this technology is just moving at a breathtaking speed. So that gives me a lot of excitement but also keeps me very grounded about how little I know. You know this famous saying, I don't know anything, and someone like me even feels that and I want you to at least hear this from me and recognize we all feel that and don't give up your learning and continue to be curious. The other thing that surprised me and frankly continues to worry me is the polarized rhetoric of AI, that too much cyberspace cycles are still spent in discussing AI and its implication at the extreme ends, whether it's total tech utopian as if technology is never a double-edged sword, or it's total doomsday.
As if humans are facing the existential crisis every minute. That's just not a responsible way of talking about a technology that is so profoundly important to our civilization. Technology is a double-edged sword. But the agency is within all of us. Every entrepreneur, every product manager, every business leader, every software engineer, every citizen, we still have to have that agency to know where to lead this technology. And I really, really hope that going into 2026 and forward, we don't stay at the two extreme ends of talking about AI. Let's be more nuanced. Let's be benevolent. Let's have the optimism of using technology for good but the sense of responsibility of using it responsibly and let's have that conversation and not just optimize for clickbaits.
So what does success look like in your mind for AI in the next few years? And I really like your framing of this kind of polarized view that we have. All jobs are going to be lost and we're just going to be staring at the ocean, or this thing is completely useless and it's not going to do anything, tends to be...
