00:07:42 · Josh Gruenstein, Martin Casado
[00:07:42] S0: No, I think it's a maturity in the market, and you can see similar things in like the early computing market. And so to us the question then becomes, okay, how do you build kind of the general platforms that let people build businesses that aren't robotics businesses? Right? We have all these incredible software businesses today, and the people who run them don't have to think about chip architectures. Right? How do we make that possible in the physical world with robotics, and how do go about building that platform? And I agree with you. People have tried to solve that problem by thinking about very narrow layers of the stack. And in our view, the reality is you need to build a full stack platform. And the only way to know that full stack platform really works is to actually have successful business units operating on top of it. [00:08:19] S3: I mean, are you open to, like, it being the case that actually the physical world is different than the digital world and the reason we horizontalize compute is because you have clean abstractions that are generally applied and that's just not the case for, like, navigating physical space? [00:08:36] S0: Think the beauty of something like a Humoid is it's the same... It's that clean abstraction. Right? You think about what is a CPU? And it's this general computing machine. And it's not perfectly optimized for every use cases, right? But it's general enough that you can run most computing
00:08:36 · Josh Gruenstein, Martin Casado, Raghu Raghuram
[00:08:36] S0: Think the beauty of something like a Humoid is it's the same... It's that clean abstraction. Right? You think about what is a CPU? And it's this general computing machine. And it's not perfectly optimized for every use cases, right? But it's general enough that you can run most computing [00:08:50] S3: workloads Oh, so that's your horizontal component within the human. Exactly. For example, would you use the same brain or sub components to make a human as you would to make a drone or to make some other form factor? [00:09:03] S0: No. But you can make that same analogy in compute. Right? We have CPUs. We have GPUs. We have basics. Right? So I think you track the maturity of the market, and you start in this world of specialization, custom machines, because that's the only thing the market can support. You move towards general purpose, and in robotics, this is gonna look like humanoids and similar form factors. And then eventually, you kind of break past to that re specialization as the market is big enough. [00:09:25] S1: Why is this thing called Tutor Yeah. Intelligence?
00:09:29 · Josh Gruenstein
[00:09:29] S0: Tutor was founded kind of with this thesis that there is a whole Internet of human intelligence that is networked, That is posting on social media, that's writing emails, and that network of human intelligence for LLMs and other digital world models, by natural emission, has created this beautiful dataset that we can use to train general purpose models. The issue is, our work in the physical world is not similarly networked. Right? Like, I pick up my coffee cup and nobody knows. Right? Like, nobody recorded all of the pressures of my fingertips and where they were in the poses, and so we don't have this asset. And the core idea behind Tutor is what if instead of trying to find kind of this algorithm where from no data, we could build urban intelligence? And for us, we're like, yeah, that seems like a problem that's rigged against us. Right? That's not how this worked in computer vision. It's not how this worked in NLP. What if we could build this network of robots out in the real world? We could connect them with human tele operators, which we call tutors, that teach the robots how to do new tasks. And in the same way the Internet, kind of by a mission, created this dataset to train these frontier models, we can do the same with robots. And we can build this internet of robot training data. And there are all these beautiful emergent properties where suddenly it's not this massive capital project. It is to go build this network, but also you're creating incredible economic value as you're doing it. Right? You're making everybody's goods cheaper as you deploy these robots and collect this data. And so it's a very positive sum, I think, this very unique way. And so when we thought about, okay, what are we gonna name this business? You know, it's not strictly artificial intelligence, there's some artificial intelligence, but there's ultimately like this Tutor Intelligence that we're building that was connecting and controlling these robots.
00:11:14 · Raghu Raghuram, Josh Gruenstein, Martin Casado
[00:11:14] S1: Yeah, that's an amazing insight. And the funny thing is when we talk to your customers before we invested, most of them didn't seem to know or didn't even seem to care that there was a Tutor or when the Tutor was operating and when your learning model was operating. And that, credit to you guys, you made those very, very seamless too. [00:11:37] S0: Yeah, this is, to me, a signal that you know the robots are good. Right? You want the feedback to be like, my shampoo is that I'm manufacturing at volume, it's now 5¢ cheaper than it was, if the feedback is like, the robot is pretty, it's like, it's probably a really bad robot. [00:11:51] S3: Is the goal to not have tutors in the future, or does it matter? [00:11:57] S0: We think of it at a very pragmatic level, of what is the operational complexity of tutors, what is the economic impact of tutors, right? Think about what is the cost to serve models with an H200, right? And you can map that out, and what's the duty cycle and utilization of that H200, and you can think very similarly about human intelligence. What is sort of the heterogeneous profile of intelligence that you really want from humans that you can't get yet from your models? And then it becomes a unit economics question. Right? Like, how much engineering do you have to do to get that last nine out of your models versus using tutors? And really, you know, going back to what we were talking about, you wanna be able to make as many trade offs here as you possibly can to make a business like this work.
00:12:38 · Martin Casado
[00:12:38] S3: I mean, know, it's tempting to say that the economic marvel that happened with generative AI could also happen here. And the marvel that happened with generative AI is you train these models with some distribution of data, and it turns out basically the same distribution is economically viable. It's like you train it on images and people want images, you train it on text and people want, you know, it's that kind of distribution in, distribution out. It's not clear that this is actually the case with these robots, right? You could train it on a distribution of data, but then actually what you need is just slightly out of that distribution. So like, how, I don't know, you probably have the best visibility into how this scales and what the convergence is, and whether you can actually remove Tutors, like what is your intuition on that?
00:13:24 · Josh Gruenstein, Martin Casado
[00:13:24] S0: I think this is one of the big questions, and I think really it ultimately comes down to what is the data and where are getting it from? And the magic of the internet, right, is if you are a company and you provide at the application layer tools for software engineers to accelerate their software development, you are naturally in flow of all of this really, really, really economically valuable work. Right? If you can make money as a business, then you're probably also making data as a business. Right? Those economic incentives are aligned. And similarly for robots, if you're like, okay, we're just gonna strap cameras on everybody's heads and send them throughout Indonesia to collect data doing x, y, and z, you you have no guarantee. Right? You you can guess, but there's no, you know, invisible hand on the market that's guiding you to ensure that that data is actually economically valuable. [00:14:12] S3: Right. [00:14:13] S0: But if it's actually this, like, kind of emission byproduct, if you have robots in real customer sites that customers are paying for, for those robots to do that task, you know for sure that that data is economically valuable. [00:14:27] S3: Doesn't it kind of depend on also the dimensionality of the solution set though? It could be the case that the economically valuable stuff is so high dimensional that you could solve 90% of it, but the last 10% you'll need a human being to do for a while just because getting that data set is just gonna require too many tutors over a period of time.