Sebastian Thrun: Flying Cars, Autonomous Vehicles, and Education | Lex Fridman Podcast #59

TL;DR

Sebastian Thrun, a renowned roboticist and computer scientist, discusses his work in developing autonomous vehicles, pioneering online education, and building flying cars.

Transcript

following is a conversation with Sebastian Thrun he's one of the greatest roboticists computer scientists and educators of our time he led the development of the autonomous vehicles at Stanford that one 2005 DARPA Grand Challenge and placed second in the 2007 DARPA urban challenge he then led the Google self-driving car program which launched the s... Read More

Key Insights

• 💦 Sebastian Thrun's work in robotics and autonomous vehicles has revolutionized the transportation industry.
• 👻 Machine learning is a powerful tool in building intelligent systems and robots, allowing them to learn from data and experience.
• 😀 The development of autonomous vehicles faces challenges in ensuring safety, affordability, and acceptance by the general public.

Questions & Answers

Q: How did the DARPA Grand Challenge inspire the development of autonomous vehicles?

The DARPA Grand Challenge in 2005 and 2007 sparked innovation in the field of autonomous vehicles, challenging teams to build self-driving cars that could navigate through rough terrains and urban environments. This competition led to significant advancements in the industry and inspired a new generation of roboticists.

Q: What is the role of machine learning in building intelligent systems and robots?

Machine learning plays a crucial role in building intelligent systems and robots as it enables computers to learn from data and experience. By observing experts and analyzing patterns, machines can acquire skills and knowledge without explicit programming. This approach allows for more scalable and adaptable systems that can surpass the limitations of rule-based programming.

Q: How can autonomous vehicles revolutionize transportation and improve safety?

Autonomous vehicles have the potential to transform transportation by reducing traffic accidents and congestion. With the ability to learn from vast amounts of data, these vehicles can make better decisions and adapt to unpredictable situations on the road. The ultimate goal is to improve safety and efficiency while minimizing the environmental impact of transportation.

Q: What are the challenges in achieving widespread adoption of autonomous vehicles?

While significant progress has been made in developing autonomous vehicles, there are still challenges to overcome. Ensuring public safety remains a priority, and current systems need to demonstrate robustness in handling rare or unexpected situations. Additionally, the affordability and social acceptance of autonomous vehicles are crucial factors in driving widespread adoption.

Key Insights:

• Sebastian Thrun's work in robotics and autonomous vehicles has revolutionized the transportation industry.
• Machine learning is a powerful tool in building intelligent systems and robots, allowing them to learn from data and experience.
• The development of autonomous vehicles faces challenges in ensuring safety, affordability, and acceptance by the general public.
• Thrun's vision extends beyond autonomous vehicles to building flying cars and revolutionizing education through online platforms like Udacity.

Summary

Sebastian Thrun is a roboticist, computer scientist, and educator who has led several groundbreaking projects, including autonomous vehicles and online education. In this interview, he discusses various topics such as the possibility of living in a simulation, the evolution of self-driving cars, and the challenges and future of autonomous vehicles. He also shares insights on leadership, the importance of empathy, and the role of academia in solving real-world problems.

Questions & Answers

Q: Do you think we are living in a simulation?

Sebastian thinks it's irrelevant to how we should act as human beings. While it may be an interesting thought experiment for some, it doesn't impact our daily lives. He emphasizes the importance of focusing on the here and now and caring about people.

Q: What does it take to build an intelligent system and an intelligent robot?

Sebastian believes that the biggest innovation in recent years has been machine learning. Computers can now teach themselves through experience and data, similar to how humans learn. This revolution has enabled computers to learn from experts and become as good as the best in different fields.

Q: How do you choose what problems to solve?

Sebastian's main desires in life are to make the lives of others better and to learn. He seeks out problems that have a maximum impact on society, such as transportation, which has transformed the world. He aims to solve problems that can make the world a better place and challenge him to learn and grow.

Q: What are some inspiring or enlightening experiences from the autonomous vehicle projects you led?

Sebastian shares a painful experience of facing complicated bugs in their autonomous car during the DARPA Grand Challenge. It took a long time to find the cause of the bug, which was a clock synchronization issue. On the enlightening side, their team's focus on machine learning and software development set them apart from others who focused more on hardware. This approach allowed them to build a learning machine that could watch experts and improve over time.

Q: What does it take to finish a product/system on time and ahead of schedule?

Sebastian attributes their success in finishing their autonomous vehicle early to good time management and being done with everything a month before the race. This early completion allowed them to thoroughly test the system and find and fix minor bugs before the race. He advises focusing on the weakest part of the system and continuously improving it.

Q: What does it take to be a good leader?

Sebastian believes that the biggest skill a leader should acquire is the ability to put themselves in the position of others and listen. He emphasizes the importance of empathy and understanding others' perspectives. A good leader should empower their team, make them feel good about themselves, and help them achieve their goals.

Q: How do you balance public safety with the drive to innovate?

Sebastian mentions that balancing public safety with the drive to innovate is a fundamental dilemma faced by all innovators. He believes that the United States has been successful in creating a balance between the two, especially in industries like aerospace and nuclear energy. While achieving 100% absolute safety is not possible, there are methods and procedures in place to ensure public safety and progress in innovation.

Q: What are the challenges in achieving full autonomy for autonomous vehicles?

Sebastian acknowledges that building a self-driving car that can handle 90% or even 99% of everyday driving scenarios is relatively easy. The real challenge lies in the remaining 1% where unexpected and dangerous situations may occur. Achieving full autonomy necessitates addressing this last 1% to ensure safety. He expresses optimism about the progress made so far and mentions that some companies are already testing driverless cars with safety drivers present.

Q: What are your thoughts on the contrasting approaches of companies like Tesla and Waymo in developing autonomous vehicles?

Sebastian expresses his admiration for both Tesla and Waymo. He personally uses Tesla's Autopilot and believes it is a great technology for highway driving. He also values the cautious and safety-focused approach taken by Waymo. He acknowledges that there are risks and challenges involved in developing autonomous vehicles but has faith in the methods and procedures in place to ensure public safety.

Q: What role does academia play in solving real-world problems?

Sebastian believes that academia's primary role is to educate young people and help them develop critical thinking skills. He emphasizes the importance of focusing on real-world problems and building prototypes that offer solutions. He suggests that academia should focus on interdisciplinary problems and engage in more debates about the big problems that society faces. He also highlights the symbiotic relationship between academia and industry in Silicon Valley, where research is applied to real products and has a significant impact on society.

Q: What is your vision for the future of autonomous vehicles?

Sebastian is optimistic about the future of autonomous vehicles and believes that they can have a profound impact on society. He mentions the need to down-cost and harden the technology further, improve mapping and perception systems, and work on business models to make autonomous vehicles affordable and socially accepted. He looks forward to the day when fully driverless cars operate without a safety driver present.

Takeaways

Sebastian Thrun's interview provides valuable insights into various topics, including the possibility of living in a simulation, the challenges and progress of autonomous vehicles, and the role of leadership and academia in addressing real-world problems. His emphasis on empathy, listening to others, and empowering teams highlights the importance of human connection and understanding in innovation and leadership. He expresses optimism for the future of autonomous vehicles and believes that they can make a significant impact on society, provided a balance is maintained between public safety and innovation.

Summary & Key Takeaways

• Sebastian Thrun's accomplishments include leading the development of the autonomous vehicles at Stanford, heading the Google self-driving car program, and co-founding Udacity, an online education platform.

• Thrun emphasizes the importance of focusing on solving real-world problems and building end-to-end systems, rather than just researching separate components.

• He highlights the potential of machine learning in revolutionizing different fields, such as healthcare, by assisting doctors in making more accurate diagnoses.

• Thrun discusses the challenges and advances in the development of autonomous vehicles, including the use of deep learning and the need to balance public safety with innovation.