Real Alpha
Watch modeSubscribed

Minus One · 3 min · Oct 8, 2026

Start physical AI in settings where the problem can be simplified

Dmitri Dolgov

Co-CEO, Waymo

  • Aditya AgarwalHost · Partner, South Park Commons

    One of Facebook's first engineers, he helped build its first search engine; later CTO of Dropbox.

  • RS

    Ruchi SanghviHost · Partner, South Park Commons

About this segment

Dolgov advises small physical-AI teams to start with settings they can constrain or systems that assist people. Fully autonomous systems require early investment in simulation, world models and evaluation.

From “How Waymo Solved Self-Driving | Dmitri Dolgov, Co-CEO”, Oct 8, 2026. Watch the full episode

“Maybe you can control the environment or maybe you can build a system that helps humans as opposed to building fully autonomous.”
Dmitri Dolgovat 35:54 · Minus One35:54

Takeaways

  1. Controlled settings can make deployment more achievable.
  2. Human-assist products can avoid the full-autonomy bar.
  3. Fully autonomous bets need fundamentals before fast iteration.
Watch 2:32 from 35:06

In this episode

Full episode
  1. 15:00Closed-loop validation requires a driver, simulator and critic4:15
  2. 22:07Waymo’s moat is its system for improving the product5:01
  3. 26:59Each additional safety nine demands more rigorous evaluation8:07
  4. 37:27A stable scorecard can show a system is nearing deployment3:37

  5. 35:06Start physical AI in settings where the problem can be simplified2:32

  6. 41:04AI will speed simple builds, while safety-critical work stays hard3:30Up next
  7. 12:37Waymo began by testing feasibility, not building a product2:22
  8. 44:24Perception transfers more easily than a physical agent's actions3:00

Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

  1. 35:12

    where those fundamentals matter tremendously. If you had three exceptional engineers who were starting today, they had way more sized ambitions but a startup budget, where would you tell them to start? What space? Physical AI, anything? Physical AI. Let's keep to physical AI. Oh, man. There's so much. I'm super excited about all the things that we could be building in physical AI. If you have a way more than an ambition, but a qualitative, a different budget, if I think that puts you on that side of the spectrum where you need to today look at applications, where you can simplify the problem.

  2. 36:02

    Maybe you can control the environment or maybe you can build a system that helps humans as opposed to building fully autonomous. Again, this, I think in my mind, this is the big divide in our domain, right? Are you building a driver-assist system or are you building a fully autonomous system with superhuman safety that you're going to deploy and scale? Qualitatively different problems. So I think there's a similar analogous divide in other applications of AI in the physical world.

  3. 36:27

    So if you are looking for a relatively quick product you can deploy, a problem you can solve, an application you can deliver, then I would look for something where it can be simplified. Again, maybe something in, let's say, in a warehouse, you can control the environment, you can maybe put some guardrails in the system. There I think you can make a lot of progress by taking the ingredients that already exist kind of off the shelf, again, doing some open loop imitation things and deploy it and then iterate from a deployed product.

  4. 37:05

    If you're on the other side of it, then I think you have to be realistic that you need to invest in the fundamentals, which is less about the product and deploying and iterating, but more about focus on the fundamentals like world models, realistic simulators for whatever agent you're building and the whole infrastructure and evaluation machinery. so in your case like you know let's say someone was exploring autonomous systems you mentioned like you know you did a lot of research