Artificial Intelligence

Elon Musk: Tesla Autopilot | Artificial Intelligence (AI) Podcast

Elon Musk is the CEO of Tesla, SpaceX, Neuralink, and a co-founder of several other companies. This conversation is part of the Artificial Intelligence podcast. The series includes leading researchers in academia and industry, including CEO’s and CTO’s of automotive, robotics, AI, and technology companies. See the comments I make in the introduction to this video below in text form.

INFO:
Podcast website: https://lexfridman.com/ai
Course website: https://deeplearning.mit.edu
YouTube Playlist: http://bit.ly/2EcbaKf
Paper: https://hcai.mit.edu/human-side-of-tesla-autopilot/
Podcast transcript: http://bit.ly/2KXRxOP

OUTLINE:
0:00 Introduction
2:35 Start of conversation: Autopilot motivation
4:01 Display the vehicle’s perception of the driving scene
7:11 Algorithms, data, and hardware development
10:23 Edge cases and common cases in driving
12:18 Navigate on Autopilot
13:57 Hardware and software path toward fully autonomy
17:08 Driver supervision of Autopilot
20:13 Human side of Tesla Autopilot (driver functional vigilance)
23:13 Driver monitoring
24:30 Operational design domain
26:57 Securing Autopilot against adversarial machine learning
28:29 Narrow AI and artificial general intelligence
30:10 Physics view of love
31:53 First question for an artificial general intelligence system

CONNECT:
– Subscribe to this YouTube channel
– Twitter: https://twitter.com/lexfridman
– LinkedIn: https://www.linkedin.com/in/lexfridman
– Facebook: https://www.facebook.com/lexfridman
– Instagram: https://www.instagram.com/lexfridman
– Medium: https://medium.com/@lexfridman

Comments Made in Video Introduction:

This conversation is part of the artificial intelligence podcast. The series includes leading researchers in academia and industry, including CEO’s and CTO’s of automotive, robotics, AI, and technology companies.

This conversation happened after the release of the paper from our group at MIT on driver functional vigilance during use of Tesla’s Autopilot. The Tesla team reached out to me offering a podcast conversation with Mr Musk. I accepted, with full control of questions I could ask and choice of what is released publicly. I ended up editing out nothing of substance.

I’ve never spoken with Elon before this conversation, publicly or privately. Neither he nor his companies have any influence on my opinions nor on the rigor and integrity of the scientific method that I practice in my position at MIT. Tesla has never financially supported my research. I’ve never owned a Tesla vehicle. I’ve never owned Tesla stock.

This podcast is not a scientific paper, it is a conversation. I respect Elon as I do all other leaders and engineers I’ve spoken with. We agree on some things and disagree on others. My goal as always with these conversations is to understand the way the guest sees the world.

One particular point of disagreement in this conversation was the extent to which camera-based driver monitoring will improve outcomes and for how long it will remain relevant for AI-assisted driving. As someone who works on and is fascinated by human-centered AI, I believe that if implemented and integrated effectively, camera-based driver monitoring is likely to be of benefit in both the short-term and the long-term. In contrast, Elon and Tesla’s focus is on the improvement of Autopilot such that its statistical safety benefits override any concern of human behavior and psychology.

Elon and I may not agree on everything, but, I deeply respect the engineering and innovation behind the efforts that he leads. My goal here is to catalyze a rigorous, nuanced, and objective discussion in industry and academia on AI-assisted driving, one that ultimately makes for a safer and better world.

Please follow and like us:
error

Comments (14)

  1. This was an insightful discussion on various aspects of Tesla Autopilot that I hope catalyzes further nuanced conversation on the future of AI-assisted driving. Starts at 2:35. The full outline of the video is as follows:

    0:00 Introduction

    2:35 Start of conversation: Autopilot motivation

    4:01 Display the vehicle's perception of the driving scene

    7:11 Algorithms, data, and hardware development

    10:23 Edge cases and common cases in driving

    12:18 Navigate on Autopilot

    13:57 Hardware and software path toward fully autonomy

    17:08 Driver supervision of Autopilot

    20:13 Human side of Tesla Autopilot (driver functional vigilance)

    23:13 Driver monitoring

    24:30 Operational design domain

    26:57 Securing Autopilot against adversarial machine learning

    28:29 Narrow AI and artificial general intelligence

    30:10 Physics view of love

    31:53 First question for an artificial general intelligence system

  2. “ What’s out side the simulation?”
    AI: sorry Elon the simulation is running inside a VMWARE “os” 😛

  3. Elon's answer about adversarial machine learning is disappointing. He basically says adversarial training is the fix, but many papers have shown that this is not the case, and doesn't really take into account that new attacks can appear which the net has never seen and thus will still break.

  4. Great last question!

  5. From perhaps a layman point of view I’m curious. Assuming the fleet is feeding all this data back to a central system and improving navigation constantly, for countries that have sub par mobile & broadband networks, with shoddy infrastructure, say some cities in Africa, Indonesia, other parts of south east asia, is the infrastructure (pot holes, narrow streets, dirt roads etc ) a problem that would slow down the adoption of autonomous vehicles and rely on human intervention more. Or is this designed to help navigate challenging terrain even in these circumstances. Would love to know more

  6. Really wish the conversation started with the question at 28:29

  7. elon musk is GOD OF HUMANITY !!!!!

  8. I have never riden a horse so I have a question, Aren't horses version 0.1 of Self Driving cars? They don't collide with other and stop when danger is ahead.

  9. I don't think Elon passed the turing test on this one. Try again later.

  10. I cannot handle the way you talk you speak so monotone

  11. Cool, how you get to make the movers and shakers to your studio.

    God bless, Proverbs 31

  12. This video looks like a deposition. You need a better cam! Amazing interviews though on this channel.

  13. sounds so much like knight rider for everybody but of course its really nice to know that car can park itself expertly without one being getting hassled [pun on hasselhoff] for real so many times parking in short or small tight space is headache but if car knows what to do and becomes knight rider on its own why not

  14. wasn't that turned out to be risky, there was one car which blew up or something.. people shuld be in control of their car, but with more safety elements on busy highways or even in crash course but of course the other electric car without gassy fuels are very good idea

Comment here