Skip to content
The Joe Rogan ExperienceThe Joe Rogan Experience

Joe Rogan Experience #1422 - Lex Fridman

Lex Fridman is a research scientist at MIT working on human-centered artificial intelligence and autonomous vehicles. Check out is podcast "Artificial Intelligence Podcast" available on Apple Podcast & YouTube - https://www.youtube.com/channel/UCSHZKyawb77ixDdsGog4iWA

Joe RoganhostLex FridmanguestGuestguest
Feb 4, 20202h 50mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:0015:00

    ... here we go.…

    1. JR

      ... here we go. Three, two, one. (slaps table) Lex, handsome as ever.

    2. LF

      Thank you.

    3. JR

      Well dressed. I always feel like a slob when I'm around you.

    4. LF

      (laughs)

    5. JR

      Do you dress like that in real life? Or only when you do podcasts?

    6. LF

      Yeah. So I have two outfits, this and black shirt and jeans.

    7. JR

      Slick outfit.

    8. LF

      Yeah.

    9. JR

      There's noth- nothing more classic than a, uh, dark suit with a white shirt and a black tie.

    10. LF

      Yeah.

    11. JR

      Is that a black tie or is that a dark blue?

    12. LF

      Black tie.

    13. JR

      Black. Black tie. Black suit, black tie.

    14. LF

      It's armor.

    15. JR

      Yes.

    16. LF

      It makes, makes me feel, uh, like it focuses the mind.

    17. JR

      Mm.

    18. LF

      Somehow.

    19. JR

      Like a professional.

    20. LF

      Yeah.

    21. JR

      Yeah.

    22. LF

      Like I'm taking this seriously.

    23. JR

      Yes, yes, yes.

    24. LF

      (laughs)

    25. JR

      Like you're, you're, you're fucking for real, man.

    26. LF

      Got a-

    27. JR

      You got notes and shit?

    28. LF

      Yeah, I got notes and shit. But I... (laughs)

    29. JR

      (laughs)

    30. LF

      But given the suit, like, I like to get, like, dirty. Like I like to work on a car or whatever.

  2. 15:0030:00

    Ma- …

    1. JR

      she's correct. If there's an argument there, if there's an argument, a scientific argument, a- a man is never a woman. But can a man identify v- as a woman and should you respect him, her, then-

    2. LF

      Ma-

    3. JR

      ... and treat them as a woman? Yes.

    4. LF

      So-

    5. JR

      Yes.

    6. LF

      ... the- so, the question is, whether you... I mean, I- I- I'm- I'm not too deep into thinking about these specific issues, but the question is whether you should get banned for being an asshole or you should get banned for being, um, for lying. 'Cause I think lying is okay. Or the tru-... Whatever, that's-

    7. JR

      'Cause a lot of people lie on Twitter.

    8. LF

      They lie-

    9. JR

      Yeah.

    10. LF

      ... and sort of... Or-

    11. JR

      Insult.

    12. LF

      ... or-

    13. JR

      You can insult people on Twitter as long as you're not specific about their gender.

    14. LF

      But see, the- the insult thing, that's where it gets... It's the party thing.

    15. JR

      Mm-hmm.

    16. LF

      Like, if you have the- the- the a- asshole, douchebag, whatever term you wanna use, they show up to the party, and then if a person shows up to the party and a lot of people leave because they're annoying or whatever-

    17. JR

      Yes.

    18. LF

      ... that- that should be... Like, we should do something to discourage that behavior.

    19. JR

      That's a good point.

    20. LF

      Right.

    21. JR

      However, let's- let's paint a different picture of a party. Let's have a party where everyone says, "My pronouns are they/them and ze/zir and zha vu."

    22. LF

      Yeah.

    23. JR

      And th- And- and then you come in and you go, "Come on, bro, you're a guy."

    24. LF

      Yeah.

    25. JR

      And like, "No, no, no, I'm a they, you fucking-"

    26. LF

      Yeah.

    27. JR

      "... cisgendered heteronormative piece of shit."

    28. LF

      Yeah.

    29. JR

      And then they wanna kick you out of the party.

    30. LF

      Yeah.

  3. 30:0045:00

    Oh. …

    1. JR

      experience was, first of all, just i- th- he found a piece of paper that showed, like, a crime docket-

    2. LF

      Oh.

    3. JR

      ... from the 1970s, all the stuff, like, ju- drugs, crime, robbery. It was all the same issues in the same neighborhoods that he was patrolling in today. And he was like, "Holy shit." And he realized, like, "Oh, this is a quagmire." And then he found out about the laws that were in place from way back in the day where you literally, if you were an African American, you couldn't buy a home in certain areas.

    4. LF

      Right.

    5. JR

      They had ... What is that term? Is it redlined? Is that what the term is? Where they, um-... they designate certain areas where they literally won't sell homes to Black people. And he was becoming aware of this shit as he was a cop. And, you know, in the beginning, he was all gung-ho. He's like, "Um, I'm a cop. You know, I'm here to bust bad guys and, and do the right thing." And then along the way, he kind of recognized you're dealing with systemic racism.

    6. LF

      You're right, yeah.

    7. JR

      Redlining.

    8. LF

      Redline.

    9. JR

      Yeah. So-

    10. LF

      Yeah, so that, that's-

    11. JR

      ... that hasn't been addressed.

    12. LF

      It's all about... I mean, there's a million other things at home, education-

    13. JR

      Yes.

    14. LF

      ... everything and-

    15. JR

      All those things.

    16. LF

      All-

    17. JR

      I think Bernie Sanders, when he talks about those things, he, he seems like a guy who really cares about education, healthcare, and people that live in poverty.

    18. LF

      Yeah. And-

    19. JR

      I don't, I don't know if he's gonna be able to do anything. I don't know, uh-

    20. LF

      Well, that's, that's the main thing is, like, people say Democratic Socialists and so on is gonna... He's going to ma- make a slight move into whatever direction he's trying to advocate, which in this case is more investment into the infrastructure and so on, into our-

    21. JR

      Yeah.

    22. LF

      ... at home. But like, you know, he's just one human being. There's, there has to be a Congress that represents the people.

    23. JR

      Mm-hmm.

    24. LF

      And if there's anything, I think Congress is probably the most hated entity in all of the universe. Like, it, you look at all the polls of what people like and hate-

    25. JR

      Mm-hmm.

    26. LF

      ... c- like, like, rats are above-

    27. JR

      (laughs)

    28. LF

      ... in terms of favorability ratings. So Congress is really the broken system. Uh, Bernie won't be able to do much except take a little sort of... The role of the president as I see it is to, one, the terrifying one, is to start wars. And so this ver- it's a very serious responsibility you have to take. And the second is to inspire the population. In terms of executive power of enacting laws, there's not much power. All you can, all you can do is, um, what our current president is doing, sort of, um, inspiring the, uh, in that, in that case, the Republicans in Congress to sort of work together to work on certain legislation. So you can inspire the Congress and you can inspire the people, but you don't have actual direct power. So Bernie's not going to turn America into a socialist, uh, you know, uh, uh, haven. He's going to take a small step into maybe, he'd probably fixing, focusing on one aspect, like healthcare or something like that, like President Obama did and try to m- make a little change. And so, um, so in that sense, people that are genuine and have ideas, like Andrew Yang is another one. Just (laughs) he has, like, a ridiculous number of ideas. I don't know if you've seen, like, his website.

    29. JR

      He thinks all cops should be purple belts in jujitsu.

    30. LF

      Yeah, I like it.

  4. 45:001:00:00

    Mm-hmm. …

    1. LF

      a couple of years, that's able to drive autonomously. I mean, most people would laugh at that idea. Like the most roboticists that know from the DARPA Challenge days, most of them know how hard this problem is. He said, "No, no, no. We're going to ... We're not only gonna throw away LIDAR," which is this laser-based sensor, "We're gonna say cameras only, and we're gonna use deep learning, machine learning, which is learning based system." So it's a system that learns from scratch, and we're gonna teach it to drive from eight cameras and so on. So just talking to somebody like that was the, the, the not ... The fact that he thinks like that, I think it's just fun to talk to people like that.

    2. JR

      Mm-hmm.

    3. LF

      I, I don't meet them often. Let's say, "No, no, no. Stop this bullshit of thinking that this task is impossible. Let's say why is it impossible. Is it really impossible?" Well, you find out when you start to think about most problems from first principles is that it's not actually impossible. And then you have to think, okay, so how do we make it happen? How do we create an infrastructure that allows you to learn from huge amounts of data? So one of the most revolutionary things that Tesla's doing, and hopefully other car companies will be doing, is the over-the-air software updates. Just like the update-

    4. JR

      Hmm.

    5. LF

      ... that you got. The fact that just like on your phone you can get updates over time means you can have a learning system, a machine learning based system, that can learn and then deploy the thing it learned over time, and do that weekly. That sounds like maybe trivial, but it's com- nobody else is doing it and it's completely revolutionary. So cars, once you buy them, they don't learn. Most cars. Tesla learns. That, that's a huge thing. Uh, forget about Tesla autopilot, all the stuff. Just the fact that you can update the software, I think is a revolutionary idea. And then they're also doing everything else from scratch. This is this first principles type of thinking. The hardware. So the, the hardware in your car ... I don't know when you got the Tesla but it should be hardware, um, version two. But that hardware performs what's called inference. So it's already trained, it's already learned its thing, and it's just taking in the raw sensory input and making decisions. Okay. They built that hardware themselves from scratch. Again, ballsy move. Now they're building what they're calling ... Uh, again, he's such a troll. But they're calling Dojo, is the, the name of the, the specialized hardware for training the neural networks or training the models. What training is, is the learning side of it. So they're building their own like supercomputer, like Google has a TPU to improve the training. They-

    6. JR

      TPU? What's that stand for?

    7. LF

      Tensor processing unit. It's the same thing as the more general. NVIDIA has graphics processing unit GPUs that all the nerds, all the people like me have been using on, um, for machine learning to train neural networks. Uh, it's what most also gamers use-

    8. JR

      Mm-hmm.

    9. LF

      ... to play video games, right? But they have, they have this nice quality that you can train huge neural networks on them. Okay. TPU is, uh, is a specialized hardware for training neural networks. GPUs allow you to play video games and train neural networks. TPUs clean some stuff up to make it more efficient, energy efficient, more efficient for the kinds of computation neural networks need. Google has them, a bunch of other companies have them. You know, most com- most car companies would be like, okay, let me partner with somebody else to u- from with Google to use their TPUs or use NVIDIA's GPUs. Tesla's building it from scratch. So that kind of from scratch thinking is i- is incredible. And, uh, the other two things I really l- listen, really l- that I like about Musk is, uh, the hard work. We live in a culture ... Like so many people, like I often don't sleep. I do crazy shit in terms of just focus, stay up nights sometimes. And now o- o- often people recommend to me that, you know, balance is really important.... and taking a break is important, you know, it- it- that you re- re- rejuvenate yourself, you return to it with fresh ideas. All those things are true. Sleep is important. You had, uh, people on the podcast tell you how important sleep is. But what most people don't- don't advise me is hard work is more imp- passion is more important than all of those things. Like, that should come first. And then sleep empowers it, rest empowers it, rejuvenation empowers it. Especially in engineering disciplines. Hard work is everything. And he's sort of unapologetically about that. It's not like a "Come- come to us, come work with us. It'll be a friendly environment with free snacks."

    10. JR

      (laughs)

    11. LF

      It's like, "You're gonna work the hardest you've ever worked on," whether you agree with him or not, "on the most important problems of your life." Okay. I like that kind of thinking, because it emphasizes the hard work. The- the other part, it was ... in terms of meeting him in person, I don't know if you got to interact with that off, because when he was on mic with you, he was very, um, he was very kind of-

    12. JR

      It was hard to bring it out of him.

    13. LF

      Yeah.

    14. JR

      In, in person before that, he was very jovial and friendly-

    15. LF

      Yeah. Yeah.

    16. JR

      ... and huggy. He's great.

    17. LF

      Yeah. He's fun.

    18. JR

      And then once he got on the microphone, I was like, "Oh, this is heavy lifting."

    19. LF

      Yeah.

    20. JR

      "I gotta bring this out of him."

    21. LF

      Yeah.

    22. JR

      So then we started drinking.

    23. LF

      Drinking helps.

    24. JR

      And then ... Oh, yeah.

    25. LF

      Yeah, yeah.

    26. JR

      It helps a lot.

    27. LF

      Yeah.

    28. JR

      And then once the- the drinking, you know, then- then I got to see who he is.

    29. LF

      Yeah. I shouldn't have- I should have done that. But-

    30. JR

      (laughs)

  5. 1:00:001:13:44

    But Wikipedia, bro, what…

    1. LF

      to undertake the effort of proving the Wikipedia article for gravity wrong. So...

    2. JR

      But Wikipedia, bro, what a terrible example.

    3. LF

      Is it?

    4. JR

      Wikipedia's sketchy. It says I'm Bryan Callen's brother. It says I got celiac disease.

    5. LF

      Did you see that?

    6. JR

      Says a bunch of shit that's not real.

    7. LF

      How do you know you're not related? Are you sure?

    8. JR

      I know.

    9. LF

      Okay.

    10. JR

      I know. I'm pretty sure. (laughs)

    11. LF

      (laughs) Does it still say that?

    12. JR

      I mean, he might as well be my brother. I don't know if it says it anymore, but whatever.

    13. LF

      But no, Wiki- Wikipedia-

    14. JR

      Someone put it in there again. Fuck it.

    15. LF

      Uh, Wikipedia is actually another distributor system that's incredibly surprising to me that it works 'cause it's-

    16. JR

      Yeah, it is, right? Because even though there is a lot of misinformation in it and there's a lot of, uh, you know, falsehoods, there's a lot of really good information as well, you know, like particularly about historical figures and interesting stuff, you know, if you wanna find facts on things-

    17. LF

      And on science.

    18. JR

      ... it's great research. Yeah.

    19. LF

      And science and technical topics are not, not like nutrition science or things where there's a lot of debates. On like physics and math and so on, it's really good.

    20. JR

      Mm-hmm.

    21. LF

      Like it's really, really good.

    22. JR

      So it's community-supported.

    23. LF

      Yeah.

    24. JR

      By other physicists.

    25. LF

      ... but moving back from, uh, flat earth, can we go back to why you think we're not going to be, uh, colonizing Mars?

    26. JR

      Oh, n- I'm not saying ever. I'm just saying, the problem to me is the type of people that would want to do it, 'cause they can't return, you know. That's the, the real issue with going to Mars, is that you can't return.

    27. LF

      There's, you don't think there's a huge number of non-crazy explorers in this world?

    28. JR

      That wanna die on Mars?

    29. LF

      Yeah.

    30. JR

      I got a whole bit about it. Um, I really believe that it's th- the, the fringe of the fringe that would be willing to die on Mars.

Episode duration: 2:50:25

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode ZIcQwDkcaWk

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome