Y Combinator Startup Podcast
00:31:40 · Dmitri Dolgov
You know, you can start, the architectures are known, you can start with doing some imitation learning and kind of a black box end to end model will give you very rapid progress and you'll ride that very, you know, initial steep part of the curve, and for some products that's enough, but if you need to reach superhuman levels of performance in a fully autonomous agent, in a safety critical environment, just doing kind of that basic vanilla end to end is not enough, And this is where structure comes in. And the key question here is, does the structure boost scale or does it fight it? Does it limit and constrain your solution space, or does it help you scale without loss of generality? So let me give you an example, let me illustrate this point with kind of a simple thought exercise and a toy problem. Imagine you're building a robot that will play the game of Go. And it wants, you know, you want it to play the game in the physical world, right? So you have a camera that's observing the board and you have an actuator that will actually move the pieces around. Now one way you can build such a robot is to, you know, have an end to end system that goes directly from pixels to actuation, and maybe you train it by giving it some videos of, you know, how humans play the game. And that could be a very interesting research exercise, however, if your goal was to build the world's best playing Go robot, that's probably not the most efficient way to go. And the reason for that is that there is a very simple intermediate representation that captures completely the state of the game, the state of the task that you're trying to solve, as you know, 19 by 19 board, and that gives you a fully observable and complete state of the world that you care about, at least for the game plan and link part. So leveraging that structure, it doesn't limit your model, it doesn't constrain your solution space, but it gives you a very helpful way to scale. Now that of course was a toy example. Anything that's not trivial that you're trying to deploy in the physical world will not have that property, and the fact that such a simple, clean, engineered representation doesn't exist in the physical world is the whole reason why we need end to end systems and learned representations and learned embeddings, but in the physical world, there does exist structure.
00:34:23 · Dmitri Dolgov
You have laws of physics, you have rules of the road, you have objects that behave in reasonably predictable ways, and you can use that structure in addition to the learned representations to boost your performance, simplify validation, and at the end of the day, just get better scaling laws. And this is the approach that we are pursuing at Waymo, which we call structure augmented end to end. So we go beyond the basic vanilla end to end by augmenting the learned embeddings with materialized structured representations. And that gives us a few very important advantages. So first is validation at inference time. Now because the model isn't just a black box where sensors go in and actuation commands go out, we can create a very powerful correctness and safety validation layer that you can run-in real time when the agent is deployed on our vehicles. And this is really important for any agent that's operating in the physical world. Secondly, we get great wins in efficiency when it comes to large scale training and evaluation of the generative part of the model, the decoder. Now if all you have is a black box end to end system, you are forced to do all of your evaluation and all of your training in the end to end setup all the way from sensors to decisions to actuation. Having that intermediate structured representation allows you to kinda mix and match. You can do some training at larger scale and some evaluation in the space of those compact structured representations, and some in the full space of end to end from sensors to decisions. And finally, we get strong, verifiable feedback signals for both evaluation and for training, training recipes to support things like reinforcement learning. That additional materialized structure just gives you much more powerful tools for evaluation, for metrics, as well as crafting your loss function or, you know, reinforcement learning recipes. So the lesson here is to bet on a system that's maximally learned and minimally constrained, and leverage structure intentionally to boost performance and scaling laws, both in training and in evaluation. Now that raises the question of how do you actually train and evaluate your physical AI agent? And that brings us to the next lesson. To build and safely deploy an agent in the physical world, it is absolutely critical to have a good, large scale, realistic, high fidelity simulator.
00:37:08 · Dmitri Dolgov
Now there's two ways you can do training and evaluation. You can do open loop and you can do closed loop. In open loop, you're kind of passively observing input to output pairs, and you can use that for evaluation, for training, imitation learning works like that. Evaluation takes usually the shape of, you know, if you find yourself in this situation, what would you do? And then you score that. And that's in contrast with closed loop, where in closed loop you take an action, you see the effect that that action has on the world, then you update your, through your sensors, the view of the world, you take another action and so forth and so on, and you evaluate and you train on those sequence of actions and sequence of world evolutions. Now the ability to take an action and evaluate that counterfactual is absolutely vital for building and deploying safety critical agents in the physical world. So a real simulator is how you do that, and a real simulator isn't just some lightweight tooling that sits next to your AI, it is a big AI model in of itself. And the problem of building a good realistic simulator is just as hard as building the agent itself. So the AI behind the simulator really needs to understand how the world works, the physics, the semantics, the traffic, the weather, and so on and so forth. And the quality of that simulator has to be high enough so that it doesn't only look good, but it's sufficient to train and evaluate with high confidence an agent that you're gonna be putting in the world in a safety critical environment. So in other words, you have to build a highly accurate generative world model. And at Waymo for years, we've been building what we called behavioral world models, and we're doing that way before the term world models even became popular. And now in the era of end to end models, you also need, on top of behavioral realism, you need sensing realism as well. And in fact, building an end to end model has been fairly easy for quite a while now, but evaluating it in closed loop, that was the hard part of the problem. So we've moved on to building sensing world models, and because we're using the structure augmented representation in our models, we can also leverage that structure in our simulation.
00:39:45 · Dmitri Dolgov
Our behavior world model operates in the space of structured intermediate representations, and the tightly coupled sensor world model then produces realistic sensor simulations. Our world model leverages the great work of Google DeepMind on Genie three, and that gives us the ability to produce controllable and highly realistic scenarios, both in the behavioral as well as sensing aspects. And that in turn allows us to not just evaluate our agent and train new versions of our agent in situations that we've previously encountered, but it allows us to train and evaluate in purely synthetic rare scenarios that we've never seen in the real world. So what you're seeing here is not just a generated video, it's a a full generative simulation of the Waymo Driver in operating closed loop. So here we're simulating what would happen if it came across a car that was stopped in a lane on the freeway. And you can go further than that. Here's a plane that's landing on the freeway in front of us, Or you can simulate an elephant on the loose walking through the intersection, snow on the Golden Gate Bridge or a dinosaur walking around. So the lesson here is that closed loop simulation is absolutely required for evaluation and is extremely valuable for training of your physical AI agents. So you need highly realistic, large scale simulation to train and evaluate. And this brings us to lesson number six. When you're dealing with a problem of that complexity, you can't just build a model and call it a day. You have to build an entire ecosystem, and then you also need a flywheel that powers it, because to make this work at scale, you can't just build the agent and one AI, you need to build three. You're building the agent, for us as the driver that drives the car, you also have the simulator, which is that virtual playground for the agent to learn in, and then you have the critic, and the critic is what rigorously evaluates and judges the performance of the agent and tells it how to improve. And the good news is that the fundamental reasoning and the generative capabilities of all three of those are shared, and that's why in our case, they're based on the same foundation role model. Now once you have these three pillars, you can create an incredibly powerful flywheel to accelerate your progress. So a deployment of your agent in the real world generates data.
00:42:29 · Dmitri Dolgov
That data then grounds the simulator and makes it more realistic. The simulator generates harder age cases for the critic to score and for the agent to learn from, so the agent gets smarter, gets deployed in the physical world, generate more data, and that powers the flywheel and accelerates progress. But a flywheel of course will, you know, spin in any direction or in place, so in order to make it go in the direction you want, you need to guide it by metrics. And that brings us to the final lesson, that your model is really table stakes, but eval and metrics, that's your most important, that's your strategic mode. So build your eval before you build your technology, build your eval and your metrics before you build your product. If you can't quantitatively define what good enough means, you're not really building a product, you're just iterating on your demo. So nowadays, the best model architectures are fairly well known and new ideas tend to proliferate fairly quickly. Data is incredibly important, but without good metrics, you're just flying blind. You aren't leveraging the best data and you can't really evaluate the ROI on making changes to it. So really eval and metrics, that's your foundation, and that what steers your whole tech stack. But for physical AI agents, model level evaluation is not enough. When you're putting an AI agent into the physical world, your eval and your validation needs to go much deeper and much broader. You need to evaluate and validate every component of your system from the physical layer to the behavioral layer that's running onboard in the physical world, as well as the offboard components, and all of the operational processes around it. So for us, we call that the safety and readiness framework, and we spent years building and refining it, and that's what guides our development and our deployment and our scaling, and I consider that to be one of our most important assets. And this, again, the reason it's important is because in the physical world trust is everything, and eval and metrics is how you go about earning that trust. You don't just win trust by talking about you know, the clever technical solution or the clever state of the art architecture of your models, or doing you know, flash a demo.
00:45:06 · Dmitri Dolgov
You earn it gradually, day by day, by in the field, by relentlessly proving that your system is safe and that your system works. And of course you can just prove that to yourself behind closed doors, and this is exactly why we openly publish our safety data and our safety, ongoing safety research. So then that earned trust becomes your ultimate business advantage, right? Your models can be leaked, algorithms can be replicated, but hundreds of millions of miles of fully autonomous operations in the real world backed by evidence grade evaluation and publicly audited proof, that is much, much more difficult to replicate. So when you zoom out and look at this playbook as a whole, you realize that none of these lessons works alone. So the nine set your bar and ensure that you pick the right technology and the right technical approach so that you don't get stuck in the local minimum. Then intentional use of structure to boost scaling and the ability to arrive technical waves of innovation, I guess helps you get to the right level of nines. And your AI ecosystem with the agent, the simulator, and the critic guided by your eval and metrics, that's what allows you to build that powerful flywheel, and that's how all of these effects compound. And it's this playbook that we've been refining over the years is what allows us to achieve the strongly superhuman safety performance of the Waymo Driver. This is a snapshot of the latest safety data we've released. It's based on over 220,000,000 fully autonomous miles, and we're seeing there that in the areas where we operate, the Waymo Driver is about 17 times better than human drivers when it comes to crashes that cause serious injury. And that really matters, because today, somewhere in the world, every twenty six seconds, someone loses their lives on a road to a crash event. And on the current scale, what that means is that Waymo is preventing a serious injury every eight days. And this isn't just a metric on a dashboard, that means that someone's loved one got to walk through the front of the door at the end of the day safe and unharmed. So these are just the early safety benefits of AI in the physical world, and they will only grow from there. If you look at the broader landscape, the opportunity here is absolutely massive.