Joe Rogan Experience Review podcast - A review of Episode 1188 Lex Fridman

Episode Date: November 3, 2018

  AI scientist working on discovering artificial consciousness. Talks in great length with Joe about how the study of AI is going. Enjoy the review folks! Please email me with any suggestions and questions for future shows : [e-mail]

Hello and welcome to another episode of the JRE Review. The JRE Review podcast reviews the best of the Joe Rogan experience podcast for the week. It's a review show. It's like how events are for the news. There's no direct correlation, but interesting things have happened, and I want to talk about it. Sometimes I have guests on friends of mine or just fans of Rogan that want to talk about particular episodes that they really enjoyed. This week's episode is podcast 1188 Lex Friedman. Lex is a physicist that studies artificial intelligence Or maybe he is more of a computer coder Basically what he does is he spent his life trying to understand and create AI Artificial intelligence and he's a super interesting guy so I can see why Robin wanted him on. He was talking about taking the angle of learning a lot of philosophy to understand consciousness and consciousness really is as far as

these AI guys can tell is the center point of what artificial intelligence is going to be. So how do you make some sort of consciousness? Well first you have to understand what it is. So he's been listening to people like Sam Harris and very smart people that study the mind and that's kind of the angle that he takes towards solving this problem. So I found that really fascinating. They move on to talk about Neil de Grass Tyson for a bit. Neil always comes up because he's very smart dude and has great theories on things and and all things science and they were talking about how he loves to shit on movies and Neil de Grass Tyson was on

Robin not that long ago and one of the things he says he doesn't do anymore is talk about movies because people get mad at him which I think is quite funny. One of the ones they get into is that movie gravity with San Gopola con George Clooney and Neal was saying how to Joe he was saying how silly it is that all the space stations are so close together It wouldn't be like that and it's just one of the things that that he pointed out and then and then Joe wanted to talk about that the newer aliens movie that had the robots in and That was like a version of AI but one of them was very creative

But there was a problem with it because he would, he was so creative that it made, it made the robots act kind of messed up. I don't know if you got, I think that one was called Prometheus or something. Good movie. I really liked the character that played the robot too, and then there were two types of robots in there. One was a newer model, but didn't have the ability to be creative. They took it out. So it was more like a kind of like server robot and you know they like battled it out. So it was like a good representation of like AI and what it could do. Of course with like a Hollywood twist on it. But yeah, I definitely like that. Joe talks a little bit about his very first Jiu-Jitsu class to Lex, and really how his background in martial arts was honestly useless up until then. So Joe Rogen studied Taikwondo and was very good, like national level, good, from like the

ages of I think 15 to 21, and he really shaped a lot of how he sees the world and understands things but when he got to Jigitsu he was up against the guy's similar size and he just got demolished the guy could do anything he wanted to and if you've ever trained Jigitsu you know that is the case it's very humbling it makes you realize that it realize that there's no brute strength stuff. You can just learn the skills of movement and it's just like being a human being, the trains is all you need. It doesn't matter what size you are. I've even seen people that have disabilities like a shorter arm or like a bad leg and they've trained a long time and they're pretty handy. So they can even adjust jujitsu for those different reasons. Joe and Lex kind of had differing viewpoints from what I could tell on this podcast about

When AI was coming about and what it will be able to do and Joe was saying that AI will eventually be at a approve on its own design Limitlessly like limitless understanding just keep improving on its design and make itself therefore infinitely powerful and intelligent Lex wasn't feeling that. He didn't, he didn't almost, I feel like he almost doesn't believe that they're going to be able to make AI in any terms, or at least for a very long time. He was talking about how neural networks,

which are like the ways we see AI now, kind of work in two ways. One is competitive self-play, and the other one is like massive input. So competitive self-play was a cool thing that he talked about, and it was about how the AI in... Oh, and this only really works if you can simulate the world like a game of chess or Sadooku or something like this. Well, I guess not Sadooku because you don't play against somebody, but any game where there's like two sides, right, so maybe like back down or chess or those sorts of things. And what what happened is the computer will then play against itself in many, many different ways and see, many different ways

and see all the different ways that it can win. The problem with that sort of problem solving is in the real world, there's way more variables. So you can't just run those until you can understand the whole picture of that simulated world. So that's very difficult right now for AI to do. Another thing is the mass input. So that means if you want a program to know the difference between a dog and a cat, it's probably going to need like 10,000 pictures of each one, dogs and cats, to be able to be sure that it could tell the difference. So if that's all one good, if you just wanted to do that,

but if you wanted it to be able to solve lots of different things and be able to tell the difference between all kinds of different, I mean, it would be so much data would have to store and process that it's just not really very effective direction of that. Then they talk a little bit about Black Mirror, that really creepy, fantastic kind of sci-fi show that's about, I guess, the sci-fi, it's just kind of about technology in the future. But one of the really good ones is called Meatita. And why it's good is it uses a robot that's very similar to this like dog style robot that Boston Dynamics has.

But what it does is it's hunting humans. If you haven't seen that episode, make sure you do. It's really very good. And Joe was just saying, you know, how long will it be before we can make something like that? Lex didn't really have any sort of an answer, but he did say in all tense purposes, the machines capabilities were like, they have machines that can do all those things now it just can't make the decisions required to kind of run around and figure all this out. Next they talk about the worry and the fear of AI, obviously Elon Musk is scared of AI, Sam Harris, same thing, and in a way that they were saying that it's not always a good thing to pursue technology,

right? For example, like, Nukes are bad, but we have them. And then Lex basically says that it's inevitable because AI started with our desire to create the gods. So in a sense, it was like a, it's kind of a creepy quote, but a really good one all the same, because it's like we want to make the things that are kind of better than us in a way in a way. But again, talking about, you know, machines that can like hunt us down, Lex made it very clear that he's not really concerned about AI.

In fact, Tim Kennedy, UFC fighter and a special forces guy, he was like, Tim Kennedy will be more dangerous for the next 50 years. Obviously, Tim will get old, but he's just saying military guy, being able to use his human brain to chase you down. And chances are they're just going to augment the soldiers rather than replace them with robots, at least for a very long time. Lex is real concerned is drones. Drone's he was saying to Joe or potential concern. And that's kind of it. Then moving on, let's start to just talk about how he positions his own life for success.

And Joe said at the end of the day, you know, being a celebrity is almost like a cheat code, which I found quite fascinating. And, you know, you want to work hard. The end result isn't just you laying on your back with somebody feeding you grapes all day. You're not going to feel good about that. Doing things that are hard, it should never stop. And in a way, scarcity is an excellent driving force. So he takes someone like Joe Rogan.

He has really all the money you could ever need and then all the opportunities to do things you could ever want. But he still brutally works so hard. And it's like, what is that driving factor? What is that piece? And he makes it clear that it's really not anything special. It's just that he does it.

He just does it. And kind of in a way, every day, every week that he does, all the work that he does is working out as you get to, his comedy, his podcast, so you have to see, it just becomes a part of him. And then that's how he gets it done. Finally, the end with something quite, find AI for sex parts, right? So of course, it's going that direction. Eventually, they're going to make sex parts. Talk about it now, women hate it. If you're out there and you have a girlfriend and you talk about, hey honey, will it be bad

if I ever got a sex part? Guarantee they're gonna hate it. They will hate the idea of it. Because they talk about how we'll probably make them insecure and the sex parts can look however you want. But my theory is that the women are gonna get one too. One that sits there and listens and looks hard and does everything they want.

And that's really it. one that sits there and listens and looks hot and does everything they want and that's really it. But anyway it was a great episode, Lex is a fascinating guy. I can't wait to hear more about him and have him back on and see how his research expands and the things that he learns and just him as a character is a young guy. So I hope he gets a voice and more chance to hear about him. He really should have his own podcast. I found him. He was very shy and calm sounding but he was excellent to listen to and I would love to hear more about what he's up to. But anyway, thank you so much again guys for listening and stay

tuned for more episodes. Cheers! Cheers!