Summary of Dmitri Dolgov: Waymo and the Future of Self-Driving Cars | Lex Fridman Podcast #147

This is an AI generated summary. There may be inaccuracies.
Summarize another video · Purchase summarize.tech Premium

00:00:00 - 01:00:00

In this podcast, Dmitri Dolgov of Waymo discusses the company's journey from developing self-driving cars to having a public commercial service available to users. He talks about the fifth generation of hardware that is on the new vehicle and how it is qualitatively different from the previous generations. He also discusses how Waymo's self-driving cars are built and operated, and how Alphabet's ML infrastructure helps the process.

  • 00:00:00 Dmitri Dolgov discusses his love for robotics and computer science and how it led to his work at Waymo, which is currently leading in the fully autonomous vehicle space. Dolgov also discusses his experience as a video game developer and how he put his game online for free for people to enjoy.
  • 00:05:00 Dmitri Dolgov discusses his childhood dream of becoming a traffic control cop and how it led to a career in transportation.
  • 00:10:00 Dmitri Dolgov discusses his experience working on autonomous vehicles, including his time working on the Darpa Urban Challenge team which focused on driving in a more dynamic environment.
  • 00:15:00 Dmitri Dolgov discusses his time working on the self-driving car project at Waymo, the challenges and lessons learned during that time.
  • 00:20:00 Dmitri Dolgov discusses Waymo's origins and its mission to build self-driving cars. He shares the story of the Stanford victory lap, a metaphor for the project's successes.
  • 00:25:00 Dmitri Dolgov discusses the progress Waymo has made in developing self-driving cars, including their success in driving 10 routes without any human intervention. He also talks about the importance of believing that the technology is feasible and discusses some of the lessons he learned while working on the project.
  • 00:30:00 In October 2017, Waymo began regular driverless operations on public roads in Phoenix, Arizona. This was a big milestone as it signaled the beginning of a journey to fully autonomous cars.
  • 00:35:00 Dmitri Dolgov of Waymo discusses the company's journey from developing self-driving cars to having a public commercial service available to users. He talks about the fifth generation of hardware that is on the new vehicle and how it is qualitatively different from the previous generations.
  • 00:40:00 Dmitri Dolgov, a Waymo engineer, discusses the fifth-generation of the company's drivers' hardware, which features lidar, cameras, and radars. He explains that the hardware is designed, engineered, and built in-house, and that it is a qualitative jump forward in terms of the capabilities and specs of the sensors. He also mentions that the company has shared videos of its lidar and camera sensors in action.
  • 00:45:00 Dmitri Dolgov discusses how Waymo's self-driving cars are built and operated, and how Alphabet's ML infrastructure helps the process. He notes that while self-driving cars are still in the early stages of development, they are making great strides in terms of data collection and analysis.
  • 00:50:00 Dmitri Dolgov discusses Waymo's plans for self-driving cars and the future of the technology. He discusses the importance of customer service and the role that humans will play in the cars' operation. Waymo plans to open its self-driving service to the public in Phoenix on October 8th.
  • 00:55:00 Dmitri Dolgov discusses how Waymo gets feedback from its riders, how it measures the success of its self-driving technology, and how it tries to ensure that its riders have the best possible experience.

01:00:00 - 02:00:00

Dmitri Dolgov, head of Waymo's Self-Driving Car Project, discusses the company's approach to autonomous driving. He explains that Waymo is focused on safety, reliability, and comfort for passengers, and argues that the company's driverless technology has progressed from a proof-of-concept to widespread deployment. Dolgov also discusses the challenges of scaling up this technology, including improving the hardware and software architecture, developing a reliable deployment process, and developing a good product quality assurance process.

  • 01:00:00 Dmitri Dolgov discusses how Waymo has learned from previous self-driving car projects and incorporated user feedback into its system. He also discusses why it is important for a self-driving car to arrive reliably at the destination.
  • 01:05:00 Dmitri Dolgov discusses the potential for self-driving cars to lead to more user-friendly car interfaces and delightful experiences for riders. He also discusses the potential for ride-sharing companies to improve the customer experience by removing any friction or complexity.
  • 01:10:00 Dmitri Dolgov discusses Waymo's approach to self-driving cars, focusing on the company's focus on safety, privacy, and comfort for passengers. He argues that humans must "flirt" with the rules of the road in order to be efficient and safe, and that Waymo wants to create a driver archetype that is both safe and efficient.
  • 01:15:00 Dmitri Dolgov discusses his experience riding in Waymo's self-driving cars, which he found to be more assertive and efficient than autonomous vehicles typically are. He also discusses how Waymo's trucks differ from autonomous trucking and how the company's fundamental challenges in this area carry over to other aspects of driver training and trucking.
  • 01:20:00 Dmitri Dolgov discusses how Waymo and other self-driving car companies approach decision making and planning. He notes that the technology and principles that apply to self-driving cars across all domains continue to carry over between domains.
  • 01:25:00 Dmitri Dolgov discusses the current state of Waymo's driverless technology and how it has progressed from a proof-of-concept to widespread deployment. He also discusses the challenges of scaling up this technology, including improving the hardware and software architecture, developing a reliable deployment process, and developing a good product quality assurance process.
  • 01:30:00 Dmitri Dolgov discusses how Waymo is incorporating lessons learned in Phoenix into its next system to increase efficiency and scale. He also mentions plans to move to Austin, Texas.
  • 01:35:00 Dmitri Dolgov, head of Waymo's Self-Driving Car Project, discusses the importance of lidars and radars in self-driving cars, and how they complement each other. He says that while there are some trade-offs to make, they are ultimately worth it because they make the system more capable and safer. He also discusses Elon Musk's comments that lidars are a crutch.
  • 01:40:00 Dmitri Dolgov discusses the potential benefits of lidar technology, noting that it can be used in a variety of ways to improve self-driving car performance. He also discusses the philosophical question of how much of driving can be learned by computers, and stresses the importance of combining lidar data with other sensing technologies to achieve the best results.
  • 01:45:00 Dmitri Dolgov discusses how self-driving cars are becoming more and more reliant on data-driven development, and how various models of machine learning are used to fuse sensory data together to make accurate predictions. He also discusses how to inject inductive bias into a self-driving system in order to encourage predictability and human-like decision-making.
  • 01:50:00 Dmitri Dolgov discusses how machine learning is playing a role in autonomous cars, including how it is used to predict what other drivers will do, make decisions in a simulator, and handle pedestrians and cyclists.
  • 01:55:00 Dmitri Dolgov discusses the challenges of autonomous driving and how Waymo is working to overcome them. He explains that Waymo is particularly focused on safety around vulnerable road users, such as pedestrians and cyclists, and emphasizes the importance of performance across all senses in order to be successful. Dolgov also discusses a recent incident in which a woman tripped and fell into the path of a Waymo car, illustrating the importance of strong reactions in autonomous driving.

02:00:00 - 02:20:00

In this video, Dmitri Dolgov discusses how Waymo's self-driving cars are performing and how they are able to avoid collisions. He also discusses the trolley problem, which is a difficult philosophical construct that highlights the difficult ethical decisions that humans must make. He says that although humans are not particularly good at making these decisions, the algorithm must still make them.

  • 02:00:00 In this video, Dmitri Dolgov discusses how Waymo's self-driving cars are performing and how they are able to avoid collisions. He also discusses the trolley problem, which is a difficult philosophical construct that highlights the difficult ethical decisions that humans must make. He says that although humans are not particularly good at making these decisions, the algorithm must still make them.
  • 02:05:00 Russian author Mikhail Bulgakov's novel "The Master and Margarita" is a philosophical and humorous story that is enjoyable to read multiple times. Dolgov discusses how books have had a significant impact on his life, and how two of his favorites, "Death of a Salesman" and "The Catcher in the Rye," have affected him in different ways. He also mentions a Russian author who is not well-known in English, but whose work he has read in Russian.
  • 02:10:00 Dmitri Dolgov discusses how his Russian skills helps him appreciate some of the depth and beauty in works of science fiction and fantasy, such as the Stragoski brothers' Roadside Picnic. Dolgov then discusses his favorite science fiction book, 1984, and how it has lessons for today's world.
  • 02:15:00 Dmitri Dolgov discusses how our society may go in directions that we wouldn't like it to go, but if anything, that's been heartbreaking to an optimist about 2020 is that society is kind of fragile. He believes that love will be the solution, but unfortunately, he thinks it's less of a flamethrower type of next and more of a slow boil.
  • 02:20:00 Dmitri Dolgov discusses his admiration for Google's self-driving car project and how it has changed his perspective on life. Dolgov says that he's a huge fan of both Waymo and technology in general, and he looks forward to seeing what the future holds for self-driving cars and technology in general.

Copyright © 2023 Summarize, LLC. All rights reserved. · Terms of Service · Privacy Policy · As an Amazon Associate, summarize.tech earns from qualifying purchases.