Dmitri Dolgov

Their words

that your model is really table stakes, but eval and metrics, that's your most important. That's your strategic moat. So, build your eval before you build your technology.

Their words

So, a deployment of your agent uh in the real world generates data. That data then grounds the simulator and makes it more realistic. The simulator generates harder edge cases for the critic to score and for the agent to learn from.

Their words

And the problem of building a good realistic simulator is just as hard as building the agent itself.

Their words

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.

Their words

But if you need to reach superhuman levels of performance in a fully autonomous agent in a safety-critical environment, uh, just doing kind of that basic vanilla end-to-end is not enough.

Their words

Essentially, structure that fights scale will always lose. And structure that channels scale always wins.

Their words

Turns out, the task of driving is not that dissimilar from the task of modeling language uh because of the social aspects of driving. You're kind of having a conversation with other dynamic actors in the world, but you're doing that in the space kind of body language of your agent, your car, as opposed to just the language of words.

Their words

So, betting your company, betting your approach on today's hardware prices is just betting your company on a number that has a fairly short shelf life and is going to expire.

Their words

However, if you are targeting full autonomy, and you're targeting superhuman, strongly superhuman performance, you find that weak sensing just leads to a safety curve that flattens out way too early.

Their words

common failure mode is picking the tech that gives you the fastest early ramp, riding that steep curve, feeling like you're winning, projecting that, you know, steep slope into the future and feeling like the sky is the limit, and then hitting the plateau, and discovering that the technology path that you picked actually flattens out way before the performance that is required by your product.

Their words

And that's why every hype cycle produces a wave of absolutely spectacular demos and very few real products. And the recurring mistake of every cycle is spending on the demo when you should be saving for the nines.

Their words

And at scale, the long tail is the problem space, is your entire problem statement. When you drive millions of miles per week, a rare event that might happen once in a million miles, that just becomes your daily reality.

Their words

reliability and performance lives on this exponential ladder of nines. So, getting to that first 90% or 99%, that's the easy part. But then every next nine that you want to add, that takes about 10 times more effort.

Their words

It took us about 10 more years to begin providing a service, and then 5 more years to scale to half a million trips per week. So, the demo took 18 months, the product took about 15 years.

Their words

And a working demo is 1% at best of the work that you have to do. The many nines of performance, the many nines of reliability that follow, that's where the real work happens.

Their words

Given the high cost of errors, you need to have a very high level of safety and a very high level of confidence on day one before you deploy your first robot, before you drive your your first autonomous mile.

Their words

In the physical world, the cost of a mistake can be measured in human lives, not tokens. There's simply not an undo and a retry button.

Their words

in Silicon Valley, there's a common mantra to move fast and break things. However, when you're dealing with atoms instead of bits, breaking things is not really okay. So, the thing you have to do is to move fast and ship safely.

Their words

That the best AI moments will look like nothing happened. It's just the task got done safely and smoothly.