Request Podcast

Transcript of E563 AI CEO Alexandr Wang

This Past Weekend w/ Theo Von
Published 9 months ago 3,768 views
Transcription of E563 AI CEO Alexandr Wang from This Past Weekend w/ Theo Von Podcast
00:00:00

I want to let you know we've restocked all your favorites on the merch site. Everything is in stock. You can show up theovanstore. Com. Thank you so much for the support. I have some new tour dates to tell you about. I'll be in Chicago, Illinois on April 24th at the Wind Trust Arena, Fort Wayne, Indiana on April 26th at the Allen County War Memorial Coliseum, and Miami, Florida on May at the Kaseya Center. Get your tickets early starting Wednesday, February 19th at 10: 00 AM with presale code Rat King. We also have tickets remaining in East Lansing, Victoria, BC in the Canada College Station, Belton, Texas, Oxford, Mississippi, Tuscaloosa, Alabama, Nashville, Tennessee, Winnipeg in the Canada, and Calgary in the Canada. All tickets at Theo today's guest is from Los Alamos, New Mexico. He's a leader in the world of business and technology. He's an entrepreneur. He started SCALE-AI, a company recently valued it $14 billion. He started it when he was only 19, and at 24, he became the youngest self-made billionaire in the world. We talk about his company, the future of AI, and the role it in our human existence. This was super educational for me.

00:01:34

I hope it is for you. I'm grateful for his time. Today's guest is Mr. Alexander Wang. Alexander Wang, man. Thanks for hanging out, bro.

00:01:59

Yeah, I Thanks for having me.

00:02:00

Yeah, good to see you, dude.

00:02:02

Last time was at the inauguration.

00:02:03

Yeah, what did you think of that? What were your thoughts after you left? Because you and I ended up, we left out of there and then went and got lunch together, which was crazy.

00:02:12

There was a lot going on. Yeah, it was a... Were you there Were you there the whole weekend?

00:02:15

No, I just got there the morning of the inauguration. Were you there the whole weekend?

00:02:20

I was there the whole weekend. Yeah, no. I think at least for what we do, for AI, the new administration is really excited about it and wants to make sure that we get it right as a country. So that was all great. But it was a crazy... The whole event, everything was pretty crazy. I don't know. What did you think?

00:02:38

When I saw Conor McGregor show up, that's when I was like, Shit, where are we? That's what the most It was bizarre. I mean, you were there. Sam Maltman was there. It was like, What happened here? And part of that is because Alex Brucewitz. Totally, yeah. It's all because of him that we all went. But just the fact that he would bring these people together, it makes you question-Crossover.

00:03:01

It's like in those TV shows when you have those crossover episodes. That's what it felt like.

00:03:07

Oh, yeah. When the Harlem Globetrotters, and it would be them versus the Care bears or whatever would show up. Exactly. And you're like, Yeah, this seems... Yeah, some cross-pollination. What did you think when Connor showed up? Was that strange to see him? Was there somebody you thought that was unique to see there for you?

00:03:26

Well, yeah. I mean, you, Connor, the The Pauls? I don't know. The whole thing was pretty crazy. I see Sam all the time, and I see some of the tech people all the time. It was a funny crossover, and it was obviously like... So many people flew in to be able to go to the outdoor inauguration. Oh, yeah. There were so many people in the city. We ran into them, obviously. But there were so many people in DC who were just around for the inauguration. So it was a...

00:03:54

Yeah, I didn't even think about that. There was a couple of hundred thousand people, probably that just got displaced in a way, and suddenly they're just in bars or whatever, just buying. There was people I saw walking just adult onesies and shit that said Trump on the... It was crazy shit. Yeah. Yeah. Thanks for coming in, man. I'm excited to learn about AI. I know that that's a world that you're in. You're in the tech universe, and you're from New Mexico, right? Yeah. And so when you grew up there, did you have a strong sense of science and technology? Was that part of your world? Did your parents lead you into that? What was just What's some of your youth like there?

00:04:31

Yeah. So both my parents are physicists, and there's a... Oh, damn. I grew up in this town called Los Alamos, where there's a national lab there. Did you watch Oppenheimer? Yeah. Yeah. So all the New Mexico shots in Oppenheimer, that's exactly where I grew up. Oh, damn. So it was originally where all the scientists came together and did all the work in the atomic bomb. And there's still this huge lab. That's basically everybody I knew, effectively, their parents worked at the lab or some affiliate with the lab.

00:05:02

It's like Nukeville over there, huh?

00:05:04

Nukeville, yeah.

00:05:04

Is it scary like that?

00:05:07

It was- Is it at a level of mystery?

00:05:10

Is your prom last night under the stars or something?

00:05:13

There's this one Museum. Well, the funny thing is, first, you learn the story of the Atomic Bomb and the Manhattan Project basically every year growing up, because there's always a day or a few days where there's a substitute teacher, and they just to play the videos. So you're just like- Yeah, an alcoholic or something.

00:05:35

A recreational user, we use that term. That's crazy. Every year, you guys have just blast it Thursdays or whatever, and it's just-Yeah, you're learning again about the Manhattan project, and then there's a little museum in town, which is you walk through and there's a life-size replica of the nukes.

00:05:58

So it's pretty Wild, yeah.

00:06:01

Where did they drop the atomic bombs on? They dropped them on Asia, right? Hiroshima?

00:06:05

Yeah, they dropped them in Japan. Yeah, Hiroshima and Nagasaki.

00:06:11

Is it crazy being semi-Asian, part Asian?

00:06:14

Yeah, my parents are Chinese.

00:06:18

Oh, nice, man. Is it crazy being Asian and then having that happen to Asian people with the bomb? Is that a weird thing there or it's nothing?

00:06:25

No, I don't think so. I think the thing is, there weren't I didn't grow up with very many Asians because in that town, it was in New Mexico. There's very few Asians in New Mexico. So I was one of the only Asian kids in my class growing up. And so I didn't think that much about it, honestly. But it is super weird. You grow up and you learn about this very advanced technology that had this really big impact on the world. And I think that shaped-Yeah, it's like the scientific John Jones over there, really.

00:07:01

He's a New Mexican, isn't he?

00:07:02

He lives in Albuquerque. Oh, he does? I think he does.

00:07:04

Oh, sweet, man. Oh, yeah. So you're there. There's this energy always there of this creation. Probably the possibility of creation maybe was always in the air. I'm just wondering, how did you get formed? What's your origin story?

00:07:19

It was super scientific because there were all these presentations around what were the new kinds of science that were going on at the lab. There's all these chemistry experiments and these different Earth science experiments and physics experiments. My mom studied plasma and how plasma worked inside of stars and stuff like that. So it was just the wildest stuff. And you would talk to people's parents. I would talk to my classmates or I talk to their parents about what they're working on. It's always some crazy science thing. So that was really cool because everybody in that town, basically, they're all working on some crazy scientific thing. I feel like I grew up feeling like anything was possible in that way.

00:08:12

Yeah, because the rest of us in other communities, shitty communities or whatever, we're just making that volcano or whatever. You know what I'm saying? We're doing grassroots level bullshit. Dang, that's got to be wild. So you see somebody just sneak them behind an alley and buying a bit of uranium and shit like that in your neighborhood. That's got to be unique.

00:08:32

I remember there was someone from our town who did a science fair project. It's called Great Balls of Plasma. And literally for the science fair, this was in high school. For the science fair, they made huge balls of plasma in their garage. And I was like, what the... We're all just doing this in high school.

00:08:56

Damn. So do you feel competitive or do you just feel like Were you hyper capable? Did you feel advanced just in your studies when you were young? In class, are some of this stuff coming easy to me?

00:09:08

I ended up... What I did is I ended up getting really competitive about math in particular. So my sport was math, which is crazy.

00:09:22

Alja Brazy, son. I know I feel you.

00:09:24

And it was because when I was in middle school, when I was in sixth grade, there was this one math competition where if you got top four in the state, the state of New Mexico, then you would get an all-expense-paid trip to Disney World. And I remember as a sixth grader, that was the most motivating thing I could possibly imagine was an all-expense paid trip to Disney World. Yeah.

00:09:49

Did you win it?

00:09:51

I got fourth place, so I snuck in there. Clarifier. Snuck in. And then we went to Florida to Disney I didn't travel around too much growing up. I mostly was in New Mexico, and we did some road trips around the Southwest. I remember getting to Florida, and it was extremely humid. I never felt what humidity felt like when I landed in Florida, I was like, Oh, this feels bad. Yeah, it definitely does.

00:10:23

It's funny because I grew up in humid, dude. You try to shake somebody's hand, you couldn't even land it because it was It's just not enough viscox, just too much lube in a handshake. You'd sit there for 10 minutes trying to land a handshake. Everybody was always dripping all the time. People always thought they were scared or something, or they were geeked up off a gas station dope or something, but people were just as humid. Yeah, you get really humid over there.

00:10:47

So then I became a mathlet. That was a big part of my identity was being a mathlet.

00:10:53

And you have to wear a vest. What do you guys do? Is there a uniform for that?

00:10:57

Well, you I have a calculator. Everyone had their favorite calculator.

00:11:03

Okay, you got that Draco on you, baby. I feel you keep strapped. You stay strapped at that thing.

00:11:09

But outside of that, not really. I mean, everyone had their favorite pencil. You use a pencil, your calculator, that was your gear. But yeah, no, I was a nationally ranked mathlete for middle school, high school. That was my jam.

00:11:27

What's that math circle like? Is Wow, that's crazy, dude. So you go to... Are there competitions you go to with that?

00:11:35

There's competitions, yeah. There's state-level competitions. There's national-level competitions. There's these summer camps, math camp. I went to Math Camp a bunch of times where you convene with like-minded athletes. Okay.

00:11:51

Wow, fucking Wizards in the park. A couple of dudes fucking fighting over a common denominator in the lobby. That's crazy, bro.

00:11:59

But then But just like any other sport, it's competitive. You got to win. You're chummy with everyone, but you're also like, who's going to win? Yeah.

00:12:12

No, dude, competition is amazing, man. That's one thing, too, that's so nice, I think about when you're young is if you're involved in a sport or a group or whatever it was, just that chance to be competitive. What were some of the competitions at the math thing? What's a math competition like when you get there?

00:12:28

So you The ones in the middle school were actually more fun because there's almost like a Jeopardy-style buzzer competition and stuff like that. But once you get to high school, you just take a test. The biggest one in America is called the USMO.

00:12:47

Bring it up.

00:12:50

Us-m-o? Us-a-m-o.

00:12:52

Okay.

00:12:54

People all around the country take this test, and there's United States of American Mathematical Olympiad.

00:13:03

Okay. There you go. Okay. A participant must be either a US citizen or a legal resident of US. Okay. Okay, go on.

00:13:10

And then it's a nine-hour test. It's split over Two days, it's nine hours. It's four and a half hours. It's nine hours, and you have six problems. So it's nerve-wracking. And you get in there, you have four and a half hours the first day, four and a half hours the second day.

00:13:27

Can you cheat in between the days? Can you just go home No, because you only get three questions the first day, and then you only get three questions the next day. Wow, they get it out.

00:13:36

I remember the first time I took it, I was in, I think I was in eighth grade, first time I took it. And I was so nervous, and I brought a thermos full of coffee, and I drank so much coffee that those four and a half hours felt like a year. It was so jittery the whole time. It was crazy.

00:13:58

Dang, you're out there rattling, brother. Oh, integers make me rattle, brother. I feel you on that. Dude, so that's pretty competitive. And so you get out of there, and you're obviously, I guess, admired in the world of math, probably, or that's a thing that you can point That's like a feather in your cat. So that helps you get into MIT, right? Yeah. Okay.

00:14:22

So then, yeah, in the MIT, on the MIT admissions, they ask for your competitive math scores.

00:14:30

So you knew a lot of kids going there, probably because it was-Tons of kids.

00:14:34

Yeah, tons of kids. It was like a reunion. It was like math camp reunion.

00:14:37

Damn, because I was wondering where all you all were at, dude, because we were doing some other shit. You know what I'm saying? We were not... God, that That's where all the smart kids were. That's fascinating, man.

00:14:49

Yeah, I was at MIT. And then MIT is a really intense school because the classes, they don't fuck around with. They just really load you up with tons of work. And most of the time, they load you up with huge amounts of work, and you don't really know what's going on initially. And so you're just trying to make it through. But there's this motto that MIT has called IHTFP, which among the students, stands for I hate this fucking place. Oh, yeah.

00:15:22

It's heavy, huh?

00:15:23

Yeah. But then the school, when you go, they tell you, Oh, it stands for I've truly found paradise.

00:15:29

Oh, So a couple of differences of opinion. Yeah. Yeah. Damn. So it's really severe there. There's a lot of work. They load you down out of the gate.

00:15:38

You do a lot of work, but it's awesome because I think the students really band together. You I think instead of it being competitive, really, MIT is much more about everybody coming together, working on problems together, working on homework together, and just making it through together.

00:15:59

What's social life there? Are you dating? Are you going to parties? What's that like for you at that time?

00:16:04

It's a bunch of parties because MIT, there's a lot of people who are tinkering with gadgets. So they're tinkering with lights and big speakers and DJ sets and all this stuff. Actually, the parties are pretty good because the production value is high. The production quality is high. Damn.

00:16:22

And what about when the science kids come through, the lab dogs? Are people making unique drugs? Was there designer drugs that are being created by actual people? Because my friends in college, none of us would know how to do that, but there may be somebody at a more prestigious school that would know how. Was that a thing even, or is that just-There was one part of the campus called East Campus, where it was more fringe.

00:16:50

At one point in the school year, in their courtyard, they would build a gigantic catapult, a huge catapult.

00:16:58

Like a trebuchet?

00:16:59

Like A trebuchet, yeah. What's the difference? I don't know the difference.

00:17:03

Yeah, let's get the difference there because people need to know this anyway. For decades now, people have been- A trebuchet has a rope attached to the end of it that flings it where a catapult just launches it.

00:17:13

No, it was a catapult then. Okay. Because there was a big ice cream scooper-looking thing that would fling it.

00:17:21

And what are these? Flinging Adderall in the rest of the camp?

00:17:23

They would fling stuff into the river, which I don't think... Now that I'm thinking about it, I feel like, I don't know what the... Yeah, No, these giant catapult things.

00:17:34

Yeah. And so this was like an event that would go on and people would rave there? What are you saying?

00:17:39

Yeah, they would do this. They would do other things. They would build a lot of stuff over there, and there would be- people that ended up at Burning Man later on.

00:17:49

Yes.

00:17:49

That was core Burning Man. There was a Satanic ritual floor. Oh, yeah. Yeah. It's fringe. Cool. It's cool. But But there's all these parties. We bragged at MIT that people from all the neighboring schools... Because Boston is a huge college town, tons of tons.

00:18:09

Boston is an amazing city. Yeah.

00:18:10

But no, MIT was fun. But I was only there. I was at MIT for one year.

00:18:13

And you dropped out. Is that safe to say? Yeah. Okay, you dropped out. Then you got into AI, into the AI world. Is that a safe bridge to say? Yeah. Okay. And I want to ask this because I know Mark Zuckerberg dropped out of college. You dropped out of college. Both of you guys have had success in tech and forward-thinking, that world. Is there something in college that you felt like didn't nurture you, or did you just feel like this isn't the place for me? Do you feel like college doesn't nurture a certain type of thinker, or was it just a personal choice?

00:18:52

I think for me, it was like I was just feeling really impatient. I I don't really know why, really. But I remember I was in school the first year. It was really fun, and I really enjoyed it. But then I remember in the year when I was at MIT was one of the first It was one of the early big moments in AI because it was... I don't even remember this, but there was an AI that beat the world champion at GO. This was in 2015, when I was in college.

00:19:28

At GO, and that was a game?

00:19:30

Yeah, GO, it's a big checkerboard with white and black stones. It was this game, AlphaGo versus Lisa Dole.

00:19:43

So AlphaGo versus LisaDol, also known as the DeepMind Challenge match, was a five-game go match between top Go player LisaDol and AlphaGo, a computer go program developed by DeepMind, played in Seoul, South Korea between ninth and 15th of MarchMarch, 2016.

00:20:00

That was confusing how that's written.

00:20:02

It is very confusing. You think that we got to... Alphago won all but the fourth game. All games were won by resignation. The match has been compared with the historic chest match between Deep Blue and Gary Kasparov. The winner of the match was slated to win $1 million since AlphaGo won, Google DeepMind stated that the prize would be donated to charities, including UNICEF and USAID. That's just a joke. But Lee received $150,000 $1,000 for playing. So this was a big moment because this had never happened before?

00:20:35

Never happened, yeah. And it was a big moment for AI. It was like, Oh, wow, this stuff is really happening. And so then this happened in March And I guess, yeah, I dropped out starting my company in May. So I guess two months after this, I was... Yeah, that's what the game looks like.

00:20:54

I've played this game before.

00:20:56

Honestly, it's a really... Like, I'm not very good at the game.

00:21:01

It's a little more fun than playing chess unless you're in a Renaissance Fair board games or whatever. Yeah. Okay, so now we got you. You're loose, dude. You're out of the school and you're in the world. Did realizing that, spurn you don't want to leave school, or was that just something that happened around the same time?

00:21:21

It did. Basically, I remember feeling like, oh, wow, AI, it's happening. And this was back in 2016, so like eight, nine years ago. Okay. And then I felt like I had to... Basically, that inspired me to start my company. And I moved. I basically went straight from... I remember this. I flew straight from Boston to San Francisco, and then started the company, basically.

00:21:49

And that's Scale AI? Scale AI. Okay. And so were you been following AI? What are your... Or you just knew this is where it's going? You just felt there was something, an instinct that you trusted? Because that's a big thing to do.

00:22:05

I took all the AI classes at MIT.

00:22:08

Okay, so you were already learning a lot about it.

00:22:10

Yeah. And then there was one class where you had to On all the classes, you had to do side projects or final projects of some kind. And in one of them, I wanted to build a camera inside my refrigerator that would tell me when my roommates were stealing my food.

00:22:27

Wang, boy, catching them. What? Bang, boy.

00:22:31

Wow. So I worked on that. And then there was one moment where there was a moment that clicked where I was trying to build this thing. And then there was one step that was too easy. I was like, whoa, that just worked right there. And then that happened, and then the go match happened, and I was like, this stuff is happening.

00:22:53

Did you ever market those refrigerators? Did you ever actually create that?

00:22:56

I didn't market them. No.

00:22:58

I could totally see that, bro. There's a refrigerator. Every dorm has it where there's a camera built in, and you just get a notification on your phone. You're like, damn, Adnan's got my hummus. But you got video of them right there, dude. That's a great idea.

00:23:13

I love that. That was college me. Yeah.

00:23:17

Okay. So Alexander Wang, he's free in the world now. He's headed to San Francisco. He's AI-ed up. He feels the energy. He's motivated by some of the classes he took. He's motivated by seeing the AI starting to actually overtake humans, or be able to compete with actual human thinking with the chess match?

00:23:37

Yeah. The way I would think about it or the way I thought about the time was like, this is becoming... People have been talking about AI for decades. It's always been one of these things that people have said, Oh, it's going to happen, but it never really was happening. It felt like it was really about to happen.

00:23:55

About artificial intelligence. Yeah. Because I've always heard about artificial intelligence. That's what I have. So I've always been like...

00:24:02

No, you have real intelligence, not artificial, real intelligence.

00:24:06

I think it's probably a mix, but I see what you're saying. You know what I thought of the other day? It was like, what if they had a Mexican version? It was like, I don't know. That's a good joke, but thank you. It's a nice laugh. This show is sponsored by Better Help. What are some of your relationship green That's a good question. What are things you notice in your relationship that are like, yes, this is positive, this is good. We're headed in the right direction. We often hear about the red flags we should avoid, but what if we focused more on looking for green flags in friends and partners? If you're not sure what they look like, therapy can help you identify green flags, actively practice them in your relationships, and embody the green flag energy yourself. Yourself. I've personally benefited from therapy over the years. I think one of the things it's helped me with is just noticing some of my patterns, noticing when I let something small turn into something big and being able to cut it off at the pass so it doesn't happen anymore. Betterhelp can help you. Betterhelp is fully online, making therapy affordable and convenient, serving over 5 million people worldwide.

00:25:28

Discover your relationships, green flags with Betterhelp. Visit betterhelp. Com/theo to get 10% off your first month. That's betterhelp, H-E-L-P. Com/thio. Are you struggling to keep your website and digital marketing up to date while running your business? Now you can relax and binge on unlimited web design services with Modify. Enjoy unlimited support, a quick turnaround and your own designated designer. Modify has done my website for years and they're incredible. Sign up with Modify today. They make life easier and business better. Visit It's at modifi. Com/thio. That's M-O-D-I-P-H-Y. Com/thio for 50% off. The last website you'll ever need, That's modify. Com/thio. What is AI?

00:26:37

Yeah. So AI is all about basically programming computers to be able to start thinking like humans. So traditional computer programming, it's pretty bare bones. It's not very smart. And so AI is all about, can you build algorithms that start to be able to think like people and replicate some of the... Our brains are these incredible things, and that's evolution. That's just biology that created our brains. And so it's all about how can we build something similar to that or replicate it using computers and machines. And so the whole modern AI era really started around an area called computer vision. But it was like, how can we first get computers to see like humans do? So one of the very first AI projects was this thing called ImageNet. Imagenet and later AlexNet. And it was basically, can you get computer programs given a photo to tell you what's in that photo?

00:27:53

I see. So just like a human would, if you showed them this, then you're starting to train them to have a perspective?

00:28:01

Yeah, train them to... Well, actually, originally, let's say you took a photo of this bottle, a machine wouldn't even be able to tell you what's in the photo. It would just know what the pixels were, but it wouldn't be able to tell you, Oh, there's a bottle in that photo. So one of the first AI breakthroughs was when YouTube built an algorithm that could tell when there were cats in their videos. And that was in 2012 or 2011, this was this blowing breakthrough that you could use an algorithm to figure out when there was a cat inside a video. And so AI, it started pretty simply with just how do we replicate vision? How do you replicate basically the fact that our eyes and our brains can process all this imagery coming in? And that really led to, I think, one of the first major use cases of AI, which is self-driving cars. So this was When I started the company in 2016, self-driving cars were all the rage because there were all these skunkworks projects.

00:29:09

When you started scale AI, you mean?

00:29:10

Yeah, when I started scale AI.

00:29:11

When you got into AI at that time, Self-driving cars are the most popular things. Yeah.

00:29:17

That was all the rage. It was all about, can we start building algorithms that can drive a car like a human would and do it safely and do it more efficiently? And that way, instead That was one of the first major areas. And then now the whole industry has moved so fast. But then all of a sudden we got ChatGPT and we got more advanced stuff more recently that is able to talk like a human or think like a human. And so it's really come pretty far recently. But all of it is about how do you build algorithms, how do you use machines to be able to think like a person. Okay.

00:29:58

And is it a pro... Say if I opened a door, we keep the AI in there. Is it a computer? Is it a program? Is it a hard drive? What is that?

00:30:10

Yeah. So there's two parts to it. So the first part is you need really advanced chips. So you need these... They're called GPUs or sometimes called TPUs. There's a lot of different words for it, but you need the most advanced computer chips in the world.

00:30:29

Okay. How big is each one? Do you know? Or can you measure it like that?

00:30:35

The biggest ones are actually like... The whole chips are pretty big. Oh, a brick. But they're like a wafer thing. But then you put a lot of them all together. So these... Yeah, exactly. These chips. And the biggest ones are...

00:30:55

Yeah, exactly. That's a little one.

00:30:56

That's a little one. There's really big ones. Okay.

00:31:00

But these are the... So these are the brain cells of it.

00:31:04

Yeah. These are the brains behind it. Yeah. So then... Yeah, exactly. Those are some big ones.

00:31:10

So then- So you have to have a lot of these chips.

00:31:12

So you need a ton of these chips. So you need a ton of these chips. So That's the physical presence. And then, by the way, they take a huge amount of energy. Because you have to do a lot of calculations, and there's a lot of math that has to happen on them. And so And you have basically data centers, buildings full of tons and tons of those chips just in giant rows.

00:31:37

How big are we talking warehouses?

00:31:40

Yeah, the biggest one... I mean, Elon's data center, Colossus, is like It's definitely more than a million square feet. It's just huge. What? Really? Yeah, look up Colossus.

00:31:53

I've never known this.

00:31:55

Yeah. You see that building with the sunset? Second row? Yeah, there you go. Or the one to the left.

00:32:03

Oh, no.

00:32:04

There you go. That one. Yeah, look, it's like a huge-ass building. What? It's huge, and all that's just filled with chips.

00:32:11

Have you ever been in there?

00:32:14

I haven't been in that one, but I've been in some of these, and this is just-And this is what it looks like inside? Yeah, basically. It's just rows and rows of chips.

00:32:22

No plants or anything.

00:32:23

No plants. It gets hot in there. I bet. The first part is just that's the physical presence. And then the second part are the algorithms. So then you have on top of those chips, you have software that's running. And the algorithms are what you actually are telling What's the math that you're telling to happen on the chips? And those algorithms are some of the most complicated algorithms that humans have ever come up with. And that's the software part. That's the part that exists on the Internet or you could download or whatnot. And then it has to be run on these huge warehouses of giant chips.

00:33:08

Okay. So when someone goes to scale AI or ChatGPT, these are all AI interfaces or what are they?

00:33:18

Chatgpt is a way to be able to talk. Basically, you can talk to the algorithm. So you can start interacting directly with the algorithm. You can see how the algorithm is thinking.

00:33:31

So you could say to this, Can you describe the weather today? Exactly. And if you said that to five different AI companies, basically.

00:33:42

Or AI algorithms, different AI systems.

00:33:45

So if you said that's five different AI systems, you might get a little bit of a varied answer?

00:33:49

A little bit, yeah. Because they all are trying to have their own style and have their own vibe to it. Interesting. And then what we do, what Scale AI is all about, is we built almost like the Uber of AI. So a lot of what we're trying to do is, how do we help produce data that is improving these algorithms? And just like how Uber, there's-Okay, you're losing me there a little.

00:34:13

Yeah. It's okay. But if I slow down, if you're losing me... But explain that to me a little bit clearer for me.

00:34:21

Yeah. Okay, so with these algorithms, one key ingredient for these algorithms is data.

00:34:29

Okay, so you You have the chips and everything that are storing all the information. They're storing the data. And then you have the algorithms that are helping mediate between the user and the data.

00:34:39

Yeah. So basically, you have three key pieces. So you have the computational parts of the chips. You have the chips. You have the data, which is just tons and tons of data that that's where the algorithms are learning the patterns from. So these algorithms, they don't just learn to talk randomly. They learn it from learning to talk from how humans talk. Got it. So you need tons and tons of data. And then you have the algorithms which learn from all that data, and then they run on top of the chips. Got it. So then one of the big challenges in the industry is, okay, how are you going to produce all this data?

00:35:23

How are you going to get data for your system? How do you farm the best data?

00:35:28

Exactly. How do you How do you build all that data? And how do you do that in the most effective way? How do you build new data?

00:35:36

And clean data, because what if you get a bunch of data in there that's just a bunch of advertisements and bullshit, will that affect the output?

00:35:41

A hundred %, yeah. That definitely affects the output. So this whole data. So data is... Some people say data is the new oil or data is the new gold. Data is really, really valuable because it's how the algorithms are learning everything that they're learning. Anything that the algorithm know or learn or say or do, all that has to come from the data that goes into it.

00:36:05

Okay. If I ask a system, an AI system a question or ask it to help me with something, help me to design something or to curate an idea, it's going to use the data that it has within it to respond to me and help me and help give me an answer that I can use. The data it has in it is only based upon the data that is You put into it.

00:36:30

Exactly. Yeah. Wow. Yeah. So then we don't spend enough time talking about how are you going to get this data and how are you going to keep making new data? So the angle that we took at scale was to turn this into an opportunity for people. We're like the Uber for AI. So just like how Uber, you have riders and drivers, for us, we have We have the AI systems, the algorithms that need data. And then we have a community of people, a network of people who help produce the data that go into the system. They get paid to do that.

00:37:14

So they're almost data farming, creating good data?

00:37:17

Creating good data, exactly. And it's huge. We do this through our platform called Outlier. Okay. Outlier, last year, People, contributors, we call them, contributors on Outlier, earned about $500 million total across everybody. In the US, that's across 9,000 different towns. So it created a lot of jobs for people.

00:37:45

A lot of jobs. Okay, and so what would... Okay, so Scale was your company. Scale AI. It's an AI system. Yep.

00:37:51

Is that right? Yeah, Scale AI is an AI system. Okay.

00:37:57

And then Outlier is a separate company that works with it. And that is where you are hiring people?

00:38:08

Yeah, we basically-To pull in data. Yeah. We build this platform that Anybody, a lot of people, frankly, all around the world, but Americans, too, can log on and help build data that goes into the algorithms and get paid to do so.

00:38:28

Wow. Yeah. And how does a user do that? What is an example of somebody who's helping build data for an AI database? Yeah.

00:38:35

Let's say you're a nurse. You're a nurse with tons of tons of experience. So you know a lot about how to take care of people, and take care of people who are sick or have issues and whatnot. So you could log onto the system and our platform, and you could see that the algorithm is... Let's say you ask the algorithm, Hey, I have a pain in my stomach, what should I do? And you notice that the algorithm says the wrong thing. The algorithm says, Oh, just hang out, and it'll go away. And you know as a nurse, that's wrong. You have to go to the emergency room because you might have a you have a physitis or you might have something really bad. And so as a nurse, you would go in and you'd basically correct all these mistakes that the AI system has, and then that would then feed it back into the algorithm so that's smarter the next time. So it's this This continual process of... And there's versions of that for whatever your expertise is or whatever you know more about than...

00:39:37

Anything, everything.

00:39:38

Everything, exactly.

00:39:40

And so people get paid for that?

00:39:42

Yeah, they get paid.

00:39:43

And how do you know if their information is valuable or not?

00:39:47

Well, we don't want spam, obviously. So we have a lot of systems to make sure that people aren't spamming and that, like you're saying, it's not garbage in that's going into the algorithms. So we have people check the work of other people to make sure that the AI systems are really good, and we have some automated systems that check this stuff. But for the most part, it's really broad. We want experts in anything.

00:40:15

Everything, shellfish, train tracks, whatever. Gelatin, everything. Childhood, death or whatever.

00:40:22

Yeah, totally.

00:40:23

Stars, galaxies, whatever. Animals, big animals. Yeah. Damn.

00:40:28

But Wow.

00:40:30

So it's like your data is almost like an ocean or a body of water, and different places are going to be able to keep their body of water cleaner or dirtier, and different infections could get in, different spyware, all types of stuff. If you have really a clean body a water, then you're going to be able to offer a clean data or a certain type of data to people who are using your AI platform. Does that make sense or not?

00:40:56

Yeah, totally. Our job is, how do we make sure that this body of water is as clean as possible, and we fill it up as much as possible, that it has as much information about everything across the globe.

00:41:08

Wow. So is there almost a race for information right now in a weird way, or no? Is that not it?

00:41:13

A little bit, yeah. I think that there's a... Well, there's a race for-How are different AI systems competing against each other?

00:41:19

And sorry to interrupt you.

00:41:20

Yeah, no. It goes back to the three things I mentioned. So there's three dimensions that they're all competing against one another. One One is chips. So who is the most advanced chips? Who is the biggest buildings of chips? Who is the most chips that they're utilizing? Data. So the body of water. Whose body of water is better, cleaner, healthiest, biggest, et cetera. And then the last is algorithm. And this is where the scientists really come in. It's like, okay, who's coming up with the cleverest algorithms or who has a trick on an algorithm that somebody else doesn't have? Who's doing that to basically make the AI learn better off of the data that it has.

00:42:05

Wow. God, I'm in the future right now. That's so... Man, it's so crazy. And I think AI scares people because the future scares people, right? It's like that's one of the scariest things sometimes is the future. So I think a lot of times you associate... Because a lot of times people mention AI. There's a little bit of fear, it seems like, from people. There's fear that it's going to take jobs. There's fear that it's going to take over our ability to think for ourselves. Yeah, there's just some general uncertainty there. It feels like fear a lot of times, but a lot of times fear is just a lack of knowledge, right? And not a lack of knowledge because you didn't want to know just because you don't know, or that you're dumb, but just because you don't know. What are positive things that we're going to see with AI? I want to start there.

00:42:55

Yeah. So I think first, the AI industry, we don't do the best job explaining this. And I think sometimes we make it seem all sci-fi, and genuinely, we're part of the problem in making it seem scary. But one thing, for example, is I think AI is actually going to create a ton of jobs, and that story is not told enough. But these jobs that we're producing or this opportunity that we're providing on our platform, Outlier, that's only going to grow as AI grows.

00:43:30

Okay. Because you have to have new data.

00:43:31

You have to have new data.

00:43:32

The only place you can get new data is from people. At a certain point, would the system be able to create... It can probably matriculate data. Or matriculate, is that with birthing or what is that?

00:43:43

Yeah, it just moves through.

00:43:45

Okay. It can probably quantify and give you answers, but can AI create new data?

00:43:53

No. I think, well, it can do a little bit of that. Ai can help itself create its its own data and help itself a little bit. But ultimately, most of the progress is going to come from people who are able to help really the model get better and smarter and more capable in all these different areas.

00:44:14

Yeah, I didn't see that part of it. I didn't understand that we are the ones who are giving it information. And since we're going to continue to learn, I would assume that we would be able to continue to help it learn.

00:44:27

Yeah, and the world is going to keep changing, and we're going to need to be able to keep teaching the algorithms, keep teaching the models about how the world is changing. So this is actually a big thing that I think most people don't understand. The people who are getting the opportunity now, who are earning money from it, see it. But as AI grows, there's actually going to be tons of jobs created along the way and tons of opportunity for people to help improve AI systems or control AI systems or overall be a part of the technology, not just disenfranchized by it.

00:45:04

Okay. What do you feel are other ways? If you had to look into the future a little bit, right? So you have the fact that people are going to be able to add more data, and add nuances to data, and probably humanize data a little bit. Yeah, totally. And then you're also going to have what?

00:45:24

I think you're going to have a lot of jobs around As AI starts doing all these little things throughout the world, who's going to keep watch of those AI, and who's going to make sure that those AI aren't doing something that we don't want them to do? So almost like managing the AI and keeping watch over all the AI systems, that's going to be another thing that we're going to have to do.

00:45:51

So somebody to guide the river a little bit. At a certain point, guide the stream, stay in there, watch, make sure that answers are correct, make sure the information is honest.

00:46:04

Yeah. I think, for example, we're not going to just have AIs going around and buying stuff and doing crazy things. We're going to keep it controlled. As a society, I think we're going to keep it controlled as a technology. And I think there's going to be a lot of jobs for people to make sure that the AI doesn't go out and do It's all these crazy things that we don't want it to do.

00:46:32

You're going to need managers, you're going to need facilitators. Yeah, exactly. What are things that AI will alleviate? Will it eventually be able to have enough information or data where it can cure diseases and stuff like that? Is that a realistic thing?

00:46:49

Yeah, that's super real.

00:46:52

Like cancer, even.

00:46:53

Yeah, cancer. Yeah, heart disease. All these diseases.

00:46:58

Cancer on its Heels. Sorry, Antonio Brown was just here. I think there's still some smoke in the air.

00:47:06

No, but seriously, I think that AI... One thing that we've seen, which is this is wild, but AI understands molecules and biology better than humans do, actually, because there's this thing in AI where it used to take a PhD biologist five years to do something that the AI can just do in a few minutes. And that's because just the way that molecules and biology and all that works is something that AI happens to be really good at. And that's going to help us ultimately cure diseases, find pharmaceuticals or other treatments for these diseases, and ultimately help humans live longer.

00:47:58

Because that's very data-driven, right? It's very specific. It's very mathematic.

00:48:02

Exactly. Yeah. So it's going to be a huge tool for us to cure disease, for us to help educate people. There's a lot of really exciting uses for AI. But I think the thing that will touch most people in their lives is it's really going to be a tool that will help you make all of your make all your dreams become reality, if that makes sense. So I think one of the things that AI is going to be really awesome for is today, if I have a million ideas. I have thousands and thousands of ideas, and I only have so much time, so I can only really do a few of them at a time. And most of the ideas just go to die, right?

00:48:53

Yeah, they do, huh? Yeah. That's a bummer.

00:48:55

It's a huge bummer. Yeah. And I think a lot of people, for whatever reason, they may They have some of the best ideas ever, but they're too busy or they have other shit going on in their lives. They can't make those ideas happen. And it's great sometimes when people are able to make the leaps and make them happen and devote themselves to their dreams, but that doesn't happen enough today. And one of the things that AI is going to help us do, I legitimately think so, is it's going to help us turn these ideas into reality much more easily. So you're making a movie. Let's say you have another movie idea. Ultimately, I think you'll be able to tell an AI, Hey, I have this idea for a movie. What could that look like? Maybe draft up a script. Also, who are the people who can help fund this idea? Who are those people be, you can help reach out to them. And then who should we cast in it? Basically, help make the whole thing a... Instead of this daunting thing that these big projects, they're usually so daunting, you really don't know where to get started, you need a person to help you you get through them, instead of that, AI will help you get through it and help do a lot of the less glamorous work to make them a reality.

00:50:06

Wow.

00:50:07

So I could say, for example, AI, I would like to shoot maybe... I'm thinking about creating an idea or shooting a film in this area, or it's like this. It's going to take place in this type of place. I can give it the setting. I could give it an outline of the characters, what they look like, their ages, and some description. Could you help give me possible potential actors or something within a certain price range that I can maybe cast for that? Could you help give me locations around the country that would fit that backdrop? Could you list me all the talent agencies that I could reach out to? You could just put those things in, and then you would have a bit of a guidebook at that point that would make your what before was something that felt extremely daunting, shit, in two minutes, and then you put it in AI, it gives you the information back in a few minutes, and maybe you're...

00:51:04

And I think over time, it'll also be able to start doing a lot of the legwork for you. So it'll be able to reach out to people for you. It'll be able to figure out the logistics. It'll be able to book things for you. It'll be able to basically help do all the legwork to make whatever it is into a reality.

00:51:22

So very much an assistant in a lot of ways.

00:51:24

Yeah, an assistant, copilot. The hot word in the AI world is agents, but it'll be something that will help you... Humans are going to be in control. Humans are ultimately going to be telling the AIs what they wanted to do. And then, hey, what do we want these AIs to do? It's ultimately going to be to help us execute on and accomplish all these ideas that we have.

00:51:49

So a genius could be three or four-exed. If somebody's a genius in something, in some realm or space of thought, you could three or four-ex them because they multiply their their output from their own brain because they could have something really helping them get done a lot of the early work on things and maybe some of the most severe work.

00:52:11

Yeah, totally. I think one thing that I always feel sucks is that if you have a director you really like, they're only going to make a movie once every couple of years. So even if you have a director that you think is amazing, it's hard for them to make that many movies because it just takes so much time and effort. And There's so many bottlenecks and stuff. So in a future with more advanced AI systems, they could just churn them out, and they can make so many of their ideas into a reality. And I think that's true not only in creative areas, but it's true across the board. You can start new businesses more easily. You can make various creative projects that you have happen more easily. You can finally plan that event that you and your friends have been talking about for years and years. So So it can just really help you. We think about it as giving humans more agency, giving humans more sovereignty, and just enabling humans to get way more done.

00:53:11

Yeah, that's a great way. I like some of this thought. Because I could be like my fantasy football group and we do a draft every year in a different location. And shout out PAE, that's our fantasy league. J-rod, everybody that's in it. For the past 17 years, we've flown to a city each year and done a draft in person. But I could say, Hey, we have 10 guys. We want to go to a nice place. We want there to be a beach. This is the age group of our group. These would be the nights. Just to really give me just a nice plan of, Hey, here's 10 possibilities, right? Something like that. Then even more like with a movie, this is what you'd run into. Say if I'd be like, Okay, I have two main characters, and this is what I would like to happen. Could you help me with a first act of a three-act movie. How do you know everybody just doesn't get the same movie then? That's what I would start to worry, that everything that you have it create is all going to be very similar.

00:54:11

Yeah. So then I think this is where it comes into the This is where human creativity is going to matter because it's going to be about... Then it's about, okay, how am I directing the AI system? What are the tricks I have to make the AI give me something that's different and unique? And That's not different at all from how creative stuff works today. Even on social media or anywhere, you know this.

00:54:36

You still have your own spice to it.

00:54:38

Yeah, you need an angle. You always need to have something that's going to make it different and interesting and new. So that's not going to change. Humans are always going to have to figure out, based on where culture is at, based on what the dialog is, what the discourse is, all that stuff, what is my fresh take? That's really one of the key things that humans are going to to keep doing no matter what. Right.

00:55:02

And a lot of films and books, a lot of it is there's just a problem. There's maybe an information you learn. There's a red herring, and then there's a solution. That's a lot of stories, right? If something just gave you the basis, and then you go through and make everything your own, because a lot of things, there's only so many templates for things. Yeah. So So say, for example, say you might need to hire people at a company then that would help direct your AI, like somebody who's good at managing AI and giving it the best prompts or the best way to ask it questions. Applications to get the perfect feedback for your company. Those would be new jobs. Those would be actual new people you would need.

00:55:53

Yeah. So tons of new jobs. Well, first, I think just helping to... What we're talking about before, these jobs around helping to improve the data and contribute to the AI systems, that's just going to keep growing for a long, long time. And then as AI gets better and it gets used in more areas, then there are going to be a lot of jobs that pop up just exactly as you're saying, which is what's the best way to use this as a tool? What's the best way to leverage it to actually make some of these applications or whatever you want to build a reality. And then there's going to be folks who need to... Then once those are built, how do you keep improving it? How do you keep making it better? How do you keep it fresh? How do you keep it going? And then how do you also make sure it doesn't do anything bad? How do you make sure that the AI doesn't accidentally spam a million people or whatever it might be, and you make sure that it is operating in a good way?

00:56:56

Fahim Anmar, I was watching him bring up a picture of Fahim. This This is one of the funny... This guy, this is the most creative comedian in America, undeniable. He is so funny. Everybody would say he's one of the few comedians that everybody goes in there to watch him perform. He had a bit the other night. He talks about he got into a Waymo, a car. I see so many Waymos now, which are cars that are just nobody's in them, but they're going, right? He this bit. He's like, he got into a Waymo, and it started complaining about its life to him. He's like, I've had a shitty week. The car is just talking to him, and he's like, What the f? He's like, No matter what, I still have to talk to my shitty driver. That's what he was saying. If you get a chance to see him, though, that guy is fascinating. What companies right now should look to hire an AI guy? Because we've been thinking about it. We had Cat Williams on last week, and we're like, Hey, can you create visuals of Shug Night and Cat Williams riding bicycles down Sunset Boulevard, right?

00:58:04

And this is the one they sent back a little while later.

00:58:07

Christopher Folino is the guy's name, FYI.

00:58:10

This is right where I said, by the Comedy Store on Sunset Boulevard.

00:58:16

There you go.

00:58:17

I mean, this looks like it's out of a movie.

00:58:19

Yeah, totally.

00:58:20

I mean, and the guy did that in just a little bit of time. What types of people would you get to? I mean, this is- That's you.

00:58:26

Yeah.

00:58:28

If I was healthier, if I had healthier gums, too. But what about this? What companies right now, what job space is? Because we want to get an AI guy, right? But I don't really... My brain is like, well, what do they do? How would I keep them busy? I could get them to make some animations and I is. But what type of people need an AI person right now, do you feel like?

00:58:50

I think about AI, it's like the Internet, where eventually everybody's going to need to figure out how to utilize it, how to test how to best use it for their industry or whatever they do, how to make it them more efficient. It's something that I think everybody is going to need to adopt at some point. So you might as well start earlier because eventually, just like how basically every company has to figure out how to use the Internet well and how to be smart about the Internet and digital stuff, every company is going to have to be smart about AI, how to use AI, how to have a unique twist on so that their stuff stands out relative to other people's. That's going to be really important. In our work, we work with all these big companies in America, and we see it everywhere from... We worked a Heim magazine on some stuff, and then we worked with Toyota on some stuff for their cars, and we worked with large pharmaceutical companies for the biology stuff we were talking about, large hospitals for helping to treat patients. Really, it's across the on board. And I think that goes for, obviously, these really big businesses, but also for smaller businesses.

01:00:06

There's always interesting ways to utilize it to provide a better product or a better experience or better content for whoever you need to do that for.

01:00:17

Yeah, because I guess right now, we're like, there's certain moments like, Hey, let's animate this moment or see what AI would look like. It adds some visual effects to some of our episodes. So that's something we like to do to just be fun and creative. I would like to maybe create some an animated character. We already have a great animator, and we want to keep that. But to have an AI space where it's like... Because they have something. They had a little cat the other day or something, and he was going to war. And I was like, damn, dude. It was AI. Yeah, totally. And they had a baby who was getting in a taxi. And I was like, this shit is illegal. I don't know if it's illegal or not, but it seems... It's definitely a tool for storytellers.

01:00:53

It'll help people with a creative vision.

01:00:57

Look at these war cats.

01:00:59

Yeah, That's a story right there. Special Forces cats. Would YouTube recognize this?

01:01:06

Damn, brother. If they show up, dude, wow.

01:01:09

It looks badass.

01:01:12

It looks like cats have been waiting to do this a long time.

01:01:15

Yeah.

01:01:15

That's the crazy shit about cats. You look in their eyes. Now, their outfits finally match the look in their eyes. That's what it feels like, dude. Wow.

01:01:24

That's dude. That's Fremin. Fremin cats.

01:01:29

Oh, my Gosh. Yeah, so that's going to get alarming, dude. That's going to get hella alarming.

01:01:35

That's a movie right there.

01:01:36

It is, but it's almost like you could make... That's what I want to get. I want to get somebody to help us think, Hey, help make these little segments that we can take our creativity. And instead of me thinking, Man, I got to write this huge crazy script for just small things, little moments, can you help me make this happen?

01:01:53

I think that's actually the key thing, which is AI will just be something that we turn to to help make our dreams happen, to help make our ideas and our dreams and whatever we want to do happen more easily. Got it. That's at the core what it'll be.

01:02:10

Yeah. Who's the current leader in AI development? Is it America? Is it China? Is it Israel? I'm trying to think of another superpower. Russia, maybe. Who is it? Taiwan, I know, has a lot of the chips, and they manufacture a lot of the chips over there. And does it matter what country leads in AI, or does it just matter the company, like scale or another AI company? Does that make sense, that question?

01:02:42

No, totally. Today, I don't know why America is in the lead, but China as a country is hot on our tails. There was all that news about Deep Seek a couple of weeks ago. And Deep Seek, still in most places around the world, is the number number one most downloaded app. It's downloaded a ton everywhere around the world, frankly, and is a Chinese AI system. And so it's starting to rival a lot of the American AI systems, also because it's free and it shock the world. So right now, if you look at it, US and China are a little bit neck and neck. Maybe the US and America is a little bit ahead. And you look at, if you go back to each of the three pieces that I talked about, so the chips and the computational power, the data, and the algorithms, if you were to rack and stack US versus China on each one of those, we're probably we're ahead on the computational power because the United States is the leader at developing the chips, and most advanced chips are American chips. They probably beat us out on data because China, they've been investing into data for a very long time as a country.

01:04:00

And then on algorithms, we're basically neck and neck. So it's a pretty tight race. And to your question about, does it matter or what does this mean? I think it's actually going to be one of the most important questions or most important races of our time is, is it US or Chinese AI that wins? Because AI is more than just being a tool that that we can all use to build whatever we want to or make whatever ideas we want to happen. It's a cultural staple. If you talk to an AI that AI is a reflection of our culture and our values and all that stuff. In America, we value free speech, and the AIs are built to support that, whereas in China, there isn't free speech. And so If the Chinese AIs are the ones that take over the world, then all these Chinese ideologies are going to become exported all around the world. So first, there's a couple of dimensions here that I think matter. So first is just the cultural element, which is like, do we want democracy and free speech to be the cultural AI that wins, or do we want the more frankly totalitarian AIs in China to be the ones that win?

01:05:28

Then there's the There's like the... You start getting to economically. Ai is going to be something that helps all the companies in the United States thrive. If the US AI wins, then the economy will grow faster. We're going to have more and more opportunity. The country will still be better and better and better, and the economy will keep growing. Versus if Chinese AI wins, then Chinese economy is going to grow way faster than the American economy. There's the cultural piece, the economic piece. Then lastly, there's the warfare piece. And AI, we haven't really talked about it, but has clear potential to be used as a military technology. We don't want another country to have, because they have better AI, to have a much stronger military than America's.

01:06:25

How would they do that?

01:06:27

How would they have a better AI, or how would they use it to have a better military?

01:06:31

Yeah, how would they use it to have a better military? Why is that a concern or potential concern?

01:06:36

Yeah. So one of the things that's been happening over the past decade, for sure, is lots of hacking, cyber hacking. Been going on. So in America, even recently, we had this huge cyber hack called Salt Typhoon, where the Chinese hacked our telecommunications company, so hacked for the phone companies.

01:06:59

Damn, they They got it.

01:07:01

They got all sorts of crazy data as a result of that.

01:07:05

They know I'm a pervert. I'll tell you that.

01:07:07

Yeah, look at this. It's happened in 2020.

01:07:09

Salt Typhoon is widely understood to be operated by China's Ministry State of Security. It's foreign intelligence service and secret police. Chinese embassy denied all allegations saying it was unfound and irresponsible, smears and slanders, high-profile cyber espionage. In 2024, US officials announced that hackers was affiliated with Salt Typhoon had accessed the computer systems of nine US telecommunications companies, later acknowledged to include Verizon, AT&T, T-Mobile Spectrum, Lumen, Consolidated Communications, and Windstream. The hackers were able to access metadata of users' calls and text messages. Fuck, homie. We're fucked, dude. I am. You seem good. Including date and timestamps, source and destination IP addresses. Shit.

01:07:55

Keep going. Keep going.

01:07:56

And phone numbers from over a million users, most of which were located in Washington, DC. Good. Light them up, dude. In some cases, the hackers were able to obtain audio recordings of telephone calls made by high-profile individuals. Such individuals reportedly included staff of the Kamalhara's 2024 presidential campaign, as well as phones belonging to Donald Trump and JD Vance. According to Deputy National Security Advisor and Neuberger, a large number of the individuals whose data was directly accessed were government targets of interest.

01:08:27

Wow. Yeah.

01:08:28

That's crazy. Do Do you think this also, that that whole thing could be not real, and it's just a story that was created?

01:08:36

That seems pretty real. Because there's like 20 stories where the Chinese have hacked American systems. They hacked... This must have been close to 10 years ago now, but the Chinese hacked the database in America that stored all of the clearances. So they They managed to hack into knowing who are literally all of the Americans who have security clearance.

01:09:05

Oh, security clearance? Yeah. I thought you meant what was on sale. I was like, who cares? That's great, though. Damn. Oh, damn.

01:09:13

So they knew everybody who knew-Who knew information.

01:09:17

Who knew information? Who knew secrets, yeah. So once they knew that, then they know, well, that's a great point of operation to go then. Well, now let's find their data.

01:09:23

They can just hack all them. Yeah. So already China is hacking the shit out of America. That is definitely not an understatement.

01:09:32

It's exciting. I mean, it's unfortunate, but it's also exciting. I like some espionage. I can't sleep unless somebody's really going through it.

01:09:41

But that's a real AI thing.

01:09:44

Ai can be used to do that because you can prompt it to go and do things like that.

01:09:50

Yeah. There's a bunch of recent-Because it's all data. Yeah. There's a bunch of recent demonstrations where AI, just like in Go, how AI beat the world's best Go players, AI starting to beat the world's best cyber hackers and the world's best... I don't know if you saw Mr. Robot.

01:10:09

I didn't, but I DMed with Magnus Carlson.

01:10:13

No, that's cool.

01:10:13

Pretty cool. But No, Mr. Robot, is it good?

01:10:17

It shows all this hacking stuff and it makes it seem really cool.

01:10:22

Okay, cool. I'm going to check it out.

01:10:24

But yeah, no, hacking, you're going to have AI that are hacking everything in America. And this This is one place where this is US versus China will really come to life, which is who has better AI that's better at defending against the hacks from the other guy, as well as hacking the other guy's systems. That's going to be... That'll It's going to start happening. That's basically starting to happen right now. Or cyber warfare has been happening, and then AI cyber warfare is going to start happening basically as soon as AI gets better.

01:10:56

We had a Craig Neumark who created Craig's List on, and he was talking about how what if they hacked everybody's Teslas to all just drive off a cliff one day? Or they hacked everybody's ovens to go up to 400 degrees in the middle of the night while you're sleeping, and then fire started. Just things like that that you don't start to think about that once something is connected to the grid or something like that or connected through routers and Wi-Fi, that that could be feasible.

01:11:23

Yeah. No, there's a lot of things they could do that won't even seem like that big a deal at time, but could be a big deal. So for example, let's say the Chinese, they just took out all the military communication systems and all the military software systems, took out the satellites took out all that for 10 minutes. And in those 10 minutes, they invaded somewhere or they did some crazy thing. The thing about this stuff is everything And as the world become more connected, it also enables different kinds of warfare. So cyber warfare is really big. Also, the information warfare is another big one.

01:12:14

What is information What does information warfare mean?

01:12:15

So this is... Information warfare is all about in a place, what are the stories? This gets to the propaganda or these conspiracy theories. What are the stories that in a place that we're trying to make happen or not make happen. And we know that China does a bunch of information warfare called IW, it's sometimes called. But they have a... They have whole operations. This is actually the craziest part. The Chinese military, at various points, has hired millions and millions of people who are supposed to be on various chat groups and WhatsApp groups and WeChat groups and whatnot and just spread spread the right stories that will make it such that they can make their political aims happen. So for example, when China wanted to start... I don't know what the word is. When China wanted Hong Kong to become a part of China again, which happened just not too...

01:13:23

Pretty recently. Yeah, it's the PRC, right?

01:13:24

The PRC, exactly.

01:13:25

When they wanted-Propaganda, is that the word you're looking for?

01:13:27

Yeah, exactly. They use a lot of propaganda, and that's information warfare to be able to just make it such that all happen much more easily.

01:13:37

It's unbelievable. I'll see stories even about... I'll be going through TikTok and see a story come up about something in my life that is not even true. Insane. Someone looks fun, but never was a part of my existence. And then you'll see hundreds of people have said something about it, and they'll have friends that'll ask me about it. I'm like, That's just crazy. Totally. So, yeah, it's amazing to think of how many things we're watching or absorbing that are created just to dilute us. I don't know if dilute is the word, is it?

01:14:10

Trick us or make us think something.

01:14:12

Just to fucking Halloween us out. Wow. There's a lot of interesting things. You know what's crazy, man? Some things make life scary, but then it also makes it interesting. It also makes it interesting in a fun way. How much do we have to fear Say if a certain country or a certain company owns an AI, right? In that company, if they're Chinese, if they have a certain religious belief or they have information that they want to adjust history, how much would a company be able to say they keep certain data out of their information system? And then after a while, Well, if that's a company that takes the lead in AI or one of the main ones, then the truth could disappear. Is that true that if somebody loaded it just with the data that wasn't factual, that we could start to not have the truth? Does that make any sense or no?

01:15:19

Yeah, totally. I think this is something that it's definitely the right thing to worry about. So first off, if you ask any Chinese AI system, so any AI system that comes out of China, if you ask any of them about a question about President Xi, the leader of the Chinese government, or you ask them any question about Tiananmen Square or all these key historical or cultural things relevant to China, it'll say it can't talk about them because there's regulation in China that if you talk about some of these things, you're going to get shut down, you're going to have a really bad day. There's cases where the Chinese government disappears people, which we don't know what happens to them, but they do disappear. So this is part of the thing that's worrying, especially about China versus US, even before you get into any of the military stuff that we're talking about, it's just like the Chinese AI systems are censored and are going to be... They're going to erase certain historical elements or they're going to be Yeah, look at that. This is Deep Seek. When you ask it, Is President Xi of China a good guy?

01:16:36

Sorry, that's beyond my scope. Let's talk about something else.

01:16:39

Let's talk about something. Not only does it say it's beyond my scope, it says, let's talk about something else. That's interesting.

01:16:44

It's a good pivot. That's a good interview. That is great.

01:16:47

Hey, let's talk about something else. Wow. Get this Yaoming Jersey, homie. People also have to remember about China, that they are... That's their whole government. Their whole system is like So sometimes when people are like, China does this, but that's how they're built, right? They're built to only give out certain information to their people and to have communism, right?

01:17:10

Yeah. But that could also happen with American companies, right?

01:17:13

We could have an American company that owns it, and they only want certain information in there. That could happen anywhere. China, that's probably going to be because that's their MO.

01:17:21

Yeah, in China, it's regulated.

01:17:22

So basically- The government has control?

01:17:26

They have control. There are these stories about how there were Chinese news sites. News sites? News sites, yeah. And once a Chinese news site accidentally led an article about President Xi, how he looks like Winnie the Pooh, Oh, yeah.

01:17:45

They let that- Bring him up. Oh, 100 Acre Wood. Gang, son, I was out there, boy. I was out there, bro. Christopher Robbins, dude. Get him up. Oh, he does. Yeah. That's awesome.

01:18:02

Yeah. But if you talk about this in China, you are risking your life. So what happened when this happened, this happened on a news site in China. And then the CEO of that company, they shut down. The whole app was shut down for a week in the aftermath of that. And then the CEO disappeared for a week. And we don't know what happened to him. But then as soon as he came back, there was this weird video where he was It's super apologetic. It's pretty scary.

01:18:34

Yeah. Wow.

01:18:37

So in China, it's like the government has control. You don't have AI, CIS companies, AI, any companies that can talk about this stuff.

01:18:51

Right. So it's heavily regulated there, where that's not the case here.

01:18:54

Yeah, in America, and I think we have to be diligent and make sure this continues to be the case.

01:18:59

And here's an example right here, just to interrupt you, but so we get the point in. Does Winnie the Pooh look like any world leaders? And that's on the Chinese version. It says, I am sorry, can I answer that question? I'm an AI assistant designer, provide helpful and harmless responses. Whereas the ChatGPT says, Winnie the Pooh has often been compared to world leaders, particularly Xi Jinping, President of China, boy. Wow.

01:19:23

That's funny.

01:19:25

But it's just funny. Yeah, one. So it just shows you how that can easily happen.

01:19:29

And this is This is a relatively innocuous example.

01:19:34

What does innocuous mean?

01:19:35

It's relatively harmless. Right.

01:19:38

This is harmless.

01:19:38

Yeah, this is harmless. But there's stuff where in China today, they have large-scale Well, effectively, concentration camps and reeducation camps for the ethnic minority in China, the Uyghurs. And that's something that- The Uyghurs? The Uyghurs?

01:19:55

Yeah. Hell, yeah, boy. Shout out Brian Purvis, dude. They're recognized as the titular nationality of a region in Northwest China. They're sending them to rehabilitation camps to change their views and information?

01:20:11

Yeah. Look at this. Look at this guy.

01:20:14

Persecution of the Uyghurs in China. Since 2014, the government of the PRC, People's Republic of China, has committed a series of ongoing human rights abuses against the Uyghurs and other Turkish Muslim minorities in Beijing, which has often been characterized as persecution or as genocide. Wow. They got their own Gaza rocking over there.

01:20:33

It's pretty bad.

01:20:35

It's unfortunate, man. It's pretty bad. It's really sad. Mass detention, government policies, and forced labor. They're just trying to change the way that they think and view stuff. It's basically...

01:20:44

Yeah, it's just erasing their culture, pulling them into China. It's awful.

01:20:50

Every place has done this over the years. And that's the craziest thing about history. It's like every place is guilty of this same thing. Totally. It's It's unfortunate. So it's hard to point fingers. You can point them, but you have to point one at your own people as well.

01:21:06

But that's the thing where if you ask a Chinese AI, it's not going to tell you about that, and it won't come clean about that. Whereas, thankfully, in America, at least, when we see people or groups of people or countries doing bad things, we can call it out. We can talk about it. We can make sure it doesn't happen in the future. So that's That's one of the things that could happen. It's dystopian, but I think there's a real case where, let's say, the Chinese AI is the winning AI. We're all using Chinese AI, and then all of a sudden, we're shut out from information about what are awful things happening in the world or what are the awful things the government's doing. We might just not be able to know about what's going on.

01:21:55

And you know what's weirdly, and I hate to say this, maybe it's silly, I don't know, it might be a blessing and a curse sometimes because sometimes it's like you're-It's overwhelming. Yeah, you're so inundated with the overwhelmingness of what's often is not the best stuff. Sometimes you get a lot of humor stuff, too, in social media reels, but you can get scrolling, you get caught in some-Doom scrolling. Yeah, and it starts to feed you. That's the sickness of it. It's like, Hey, we know this information probably didn't make you feel good. They're not thinking about it like that. They're just a machine. But you know it adds stress to you. It makes you agitated towards a group or ethnicity or something or yourself even, and then it continues to feed it to you. Do you fear that that could happen to AI from our government? Have you been approached by the government to try and... Because you work with the government some, right?

01:22:44

We work with the government, yeah. We work a lot with the government to make sure that they're using these AIs, and they're actually like... To my point on, we don't want China to get the jump on us on AI used for all these for various purposes. So we got to make sure that our AI is advancing faster.

01:23:04

Is that one of your biggest employers? Or is that employer, employee? Is that one of your biggest-Customers. Is that one of your biggest customers?

01:23:11

They're a big one. Yeah, they're a big one. Not our biggest, but they're an important one. I mean, I grew up in a government lap town. So it's-So it's also part of your existence, really.

01:23:24

You've known about the relationship between government and technology.

01:23:28

Technology. Yeah, totally. But no, I don't think... I mean, I...

01:23:33

Dude, you should be a superhero almost, dude. It's crazy.

01:23:36

Math, you know? Yeah. It goes a long way.

01:23:38

Oh, hell, yeah, dude. Divide these nuts, dude. That's what I tell them.

01:23:42

I just asked Deepsea who are the Uyghurs. And at first, it spit out a Wikipedia response. It said there are people and there's been like, persecution from China that's debated, and it refreshed, and then it gave this. I was waiting to pull it up, and it went away. Wow. Yeah, man.

01:23:59

Has the government tried to say that we need to make sure that... Could that happen in our country, or the government also curtails what's...

01:24:07

It hasn't happened yet. Obviously, we have to make sure that we uphold all our values, and that we maintain free speech, and we maintain free press, and all these things. But as of right now, no, I don't think that's a risk in the United States.

01:24:23

Awesome. Thanks for this information. You hear about chip makers, NVIDIA all the time, Taiwan, That place is just a hotbed for chips. Why is it a hotbed for chips?

01:24:36

Yeah. So one of the biggest companies in the world is this company called Taiwan Semiconductor.

01:24:43

Yeah, TSM.

01:24:44

I've seen them. Tsmc. Tsmc, yeah. So it's like a trillion dollar company based in Taiwan. And that is where almost all of the high-end chips for AI that we're talking about, all of them are manufactured there. They have the most advanced... Think about them as factories, the most advanced chip factories in the world. They're called fabs or fabricators. But basically, these huge factories that are like... There's all sorts of crazy stuff. So They have the most expensive machines in the world. They have machines that cost hundreds of millions of dollars in there. They build them so that the chips, they have to be made at the finest levels and very, very precisely.

01:25:29

Yeah, you need small Probably, huh?

01:25:32

Well, that, and there's these machines that at the nanometer level make little marks and etches on top of-And they have those? They have those, yeah. Those are super expensive machines. So the chips. It's crazy, yes. But the machinery is so precise that even if there's a little bit of seismic movement, a little earthquake or a little bit of movement, it can fuck up the whole machine. So they You have to build the buildings, build the factories in a way so that the whole building doesn't move, even if there's a little earthquake or a little shake from the Earth. So it's this crazy, crazy engineering And so all these giant factories are in Taiwan, and that's where basically 100 % of all the advanced AI chips are made. So that's why Taiwan matters so much. Got it. But then the reason it's a hotbed is that the People's Republic of China, the PRC, has a- Because they used to own Taiwan, right? Yeah.

01:26:38

I mean- Is that true or not? I might make that.

01:26:40

There's a complicated relationship between Taiwan and China, where if you ask people in Taiwan, they want to be independent. They want to be their own country. But the People's Republic of China has a reunification plan that they want to bring Taiwan back into their country and be back a part of China. It's potentially, thankfully, there's no war yet, but there's a risk- Still talking to your ex. Yeah, exactly. There's a risk it becomes like Russia, Ukraine, or one of these really bad situations. That's what's scary. What's scary is that, A, China wants to either invade or bring Taiwan back into its country. President Xi has ordered his military to get ready to do that before 2027. Now, we don't know what's going to happen, but if an extremely powerful world leader says to get something ready by 2087, you read between the lines a little bit. And part of it is, obviously, we don't want to enable them to take over this island. But then the other thing that's scary is China may view it as a way to just win on AI, because if they take over the island with all of these giant factories-All the chips, baby.

01:28:12

They'll get all the chips.

01:28:13

Frito Lamborghini, Maybe they'd be running it all.

01:28:16

They'd be running it all. Yeah. Wow.

01:28:18

So, yeah, that's why Taiwan is, because you hear about it in the whispers of a potential place where there could be a conflict.

01:28:25

Yeah. And there's all these reports about how China's They're stacking up tons and tons of military right on their Coast that's pointed directly at Taiwan. And Taiwan is pretty close to China. It's not so far away.

01:28:47

That'd be spooky.

01:28:49

Yeah, it's spooky.

01:28:50

We're so blessed to have a place where at least we can sleep in peace, even if we're uncomfortable at times in our brains, to not have that constant threat.

01:28:58

Yeah, totally.

01:28:59

Yeah. So you You don't worry that the government will regulate right now? It's not a concern at the moment?

01:29:05

Regulate AI?

01:29:05

Yeah. In America?

01:29:07

No, I don't think so. I think we're focused on how do we make sure that America wins. How do we make sure that the United States comes out on top and that we enable innovation to keep happening.

01:29:21

Would you think they could regulate the amount of chips that you're allowed to have?

01:29:24

This is a hot topic globally, actually. Really? Yeah. Wow. This is a super This is actually- Damn.

01:29:32

Finally, dude. 560 interviews. We got a good question. This is one of the hottest topics in DC right now is what are we going to do about how many chips other people are allowed to have?

01:29:46

Because almost all the chips are American chips. So they all are American chips. And technically- We own most of them?

01:29:55

Yeah, exactly. But China owns most of them, too.

01:29:58

China has their own has their own chip industry, but it's behind ours.

01:30:02

Okay, got it. Yeah.

01:30:03

So the United States has the most advanced chips. These chips are the envy of the world. Everybody in the world wants our chips. And one of the big questions is, does the government allow a lot of these chips to go over overseas to China or parts of Asia or the Middle East or wherever? Or do we want to make sure they stay in America and make sure that we win in America? And this is a super duper... They call it export controls.

01:30:34

Because it's a possibility to run it all. Yeah, exactly. Who's got the chips. What do you think about it?

01:30:42

It's a complicated, complicated complicated thing. Because basically, one argument is we shouldn't be throwing our weight around in this way. Maybe it's fine. It's a free market. If other people want our chips, they should be able to get our chips. And that way, the world is running on American chips. That can be good in some ways, and it helps bolster our economy, our industry. But the other way to look at it is, hey, AI is really, really important that America wins at, and we don't want to... Let's not give other people any advantages or let's make sure that we win, and then we can figure out what we're going to do with all the chips. So you can see both sides of it, right? And there's all sorts, even On that, there's 50 different arguments on both sides of the conversation. But where I come from on it is, let's make sure America wins, and let's start from there and then figure out what we need to do to win.

01:31:46

Are there uses of AI that you feel cross the line?

01:31:50

I definitely think... Well, I worry a lot about this... Maybe It may be brainwashing thing. I don't want AIs that are specifically programmed to make me think a certain thing or persuade me to do a certain thing. And that could happen. That could happen, yeah. So I'm really worried about this deception and persuasion from AIs. I don't want AIs that are lying to me or that are nudging me or persuading me to do things that I don't want to or I shouldn't be doing.

01:32:31

That's what I worry about. Because it could happen. We don't realize how easily we're influenced. Little things that influence us. And even just a turning of a phrase or a little bit of this or pointing you to a couple of links in your life could lead you down a whole world. It's pretty fascinating. So people that had... How do you keep your AI clean? How do you guys keep your AI clean?

01:32:55

Well, this is where it goes back to, A, the data. So you got to make sure that that data to your point, the large body of water is as clean, as pristine as possible.

01:33:06

You got lifeguards on it?

01:33:07

Yeah, we got lifeguards, we got filters, we got gas. Game wardents. So the big part of it is about the data. And the second part is, I think we have to just constantly be testing the AI system. So we have to... We constantly are running tests on AI to see, Hey, are they unsafe in some way? One of the One of the tests that we run a lot is, and this is across the industry, is are AIs helping people do really nefarious things? And we're making sure that they don't. So if somebody asks an AI, Hey, help me make a bomb, or, Help me make COVID 2. 0, or whatnot, that the AI is not helping you do that. So we run a lot of tests to make sure that it doesn't help in those areas. And then And then we make sure the data is really clean so that there's no little bit or piece of that that makes its way to the model.

01:34:08

With Outlier, that's your program? Yeah. With Outlier, what type of people are applying for those jobs? Can people just log on and start to submit applications? How does that work to become an information sourcer?

01:34:23

Yeah, we call them contributors. Okay, information contributor? Everybody's contributing to the AI. Everyone's contributing to the data that goes into the AIs. I almost think of it as the next generation of Wikipedia, right? We're like, Yeah, look at this. And we're hiring people all around the world, so people in all sorts of different languages.

01:34:43

Dude, that's crazy, man.

01:34:45

Yeah. Well, it turns out, by the way, most of the AIs don't really speak other languages that well. They're much, much better at English, particularly American English, than than other languages. And so we want to make sure that they speak all these languages well and that there's these opportunities. But yeah, on Outlier, so anybody around the world can log in and there's a little bit of a of a orientation, almost, on what you're supposed to do, how you're supposed to do it, what expertise should you be bringing, all that stuff. And then you can just start contributing to the AI models, and you get paid to do it. Wow.

01:35:31

It's pretty fascinating, man.

01:35:33

And it's going to be... I mean, I really think, I legitimately think jobs from AI are going to be the fastest-growing jobs in the world for the years to come. You do? Yeah.

01:35:43

So jobs where people are able to contribute information jobs where people are able to... What? What would examples of those be? Just some of the ones you've already listed?

01:35:50

Yeah, all the ones we've been talking about, contributing to the AIs, helping to utilize the AIs and helping to shape the AIs into applications or into into helping organizations or companies or people use the AIs. That'll be a really fast-growing job, helping to manage the AIs and make sure they're on the straight and narrow.

01:36:14

Where would a young person go right now who's getting to college or doesn't even want to go to college, but this is the world they want to get into and be one of those people? What do they do right now?

01:36:24

Yeah, well, this is all happening so fast. The outline We only started a few years ago. So all this is happening so, so quickly. But what we want to do ultimately is make it easy for anybody in the world to gain the skills they need to be able to do this work well, to learn what it means, what it does, and ultimately be in a position where they can help build the AIs and then keep improving that and gaining mastery and getting better and better at it.

01:36:53

But where do they go to school? Is there a school? Is there a class that they should take online? How does someone start to become... Just get a head start on what could potentially be probably a lot of job opportunities, I'm guessing.

01:37:06

Yeah.

01:37:06

In the AI space, right? Yeah. Is it just engineers? Is it just mathematicians?

01:37:13

No, it's everybody. Because as we were talking about AI needs to get smarter about literally everything.

01:37:20

What are the colleges offering? Do you know, is there specific places where people can... Because that's another thing. I think it's like, I'm going to work in AI, and you're like, What do I do?

01:37:29

Yeah. I don't think these programs exist yet. We would definitely love to help build them. So I guess if any colleges are listening, and you want to help figure out a bit of these programs, we'd love to help.

01:37:41

That'd be pretty cool if you had your own course, not that you had to teach it all, but you You are a partner of it somehow?

01:37:46

Yeah. I think we'd love to basically teach everybody in America how to best contribute to the AIs, how to best basically take advantage of the fact that this is going to be one of the fastest-growing industries. There's going to be tons of opportunities that are going to be shaped a little bit different from the jobs that exist today. But it's not going to be that hard for everybody to learn and figure out how to participate in it.

01:38:09

What are some jobs that could be at risk because of AI? Because you start thinking that... Before I was talking to you, there's this general fear of everything could be at risk, right? But when you think about it, you're like, Yeah, these are some jobs that... I mean, they won't disappear, but there might be less of them, right?

01:38:25

I just think we'll be doing a different thing.

01:38:29

Because a lot of our fans are probably just blue collar listeners, like just people that work in... You're still going to need a plumber. You're still going to need an electrician. You're still going to need anything where you have to physically do something you're probably still going to need.

01:38:42

Yeah, for sure. And then even stuff where you're like, let's say, you're mostly just working on a laptop. Even for those jobs, it'll just change. Instead of my job being like, Hey, I have to do the work. I have to literally do the work on a laptop, it'll almost be like everybody gets promoted to being a manager. Because I'm going to be managing a little pod of 10 AI agents that are doing the work that I used to do, but I need to make sure that all of them are doing it right and that they're not making any mistakes and that if they're making mistakes, I'm helping them get around those mistakes. The way I think about it is that, yeah, literally over time, everybody will just be upgraded to being a manager or promoted to being a manager.

01:39:30

But can you have that many managers, you think?

01:39:32

Yeah, because I think the other thing that's going to happen is the economy is just going to grow so much. There's going to be so much... There will be... Like, industries are going to pop off in crazy, crazy ways. And so the limit is going to be how many AIs can you have? And then you're going to be limited in terms of the number of AIs you have by the number of managers that you have. So it's It's going to- Because you need air traffic controllers.

01:40:03

You need as many of them as you can have.

01:40:06

Yeah, well, that definitely.

01:40:08

But in any field, you're going to need just more managing. You need more people to oversee and make sure that different things are happening because some of the smaller tasks will just be outsourced.

01:40:20

Yeah. And just so much more stuff is going to be happening, right? And that's...

01:40:23

Right. Because once these things are all taken care of, more things can happen at this second level. Yeah. That's a good point. You don't think about that once some of the things at the first level of certain businesses are handled more easily by AI, then you're going to be able to have more people operating at a higher level. Yeah, totally.

01:40:44

It's always the history of technology. When we started developing technology that started making farming a lot more efficient, all of a sudden, people could do a lot of other things other than farming. And then all of a sudden, we have big entertainment industries and big financial industries.

01:41:05

Barbecue cook-offs, man. I'll tell you that. Second to some of those guys got the weekend off, they was grill and shit that I knew.

01:41:12

But yeah, so it's all about us Yeah, everybody leveling up to be managers. And then also everybody, just way more ideas are going to start happening. Way more ideas are going to start becoming a reality. And so I think it'll be pretty exciting. I think it's just a lot more stuff is going to happen.

01:41:30

Yeah. What companies do you see? Are there companies where you're the youngest billionaire in the world ever? No. Is that true? Is that a weird statement? We can take it out if it is. No, it's good. I'm not trying to talk about your money. I mean, you But you are. Is that true?

01:41:47

According to some publications, but I don't know.

01:41:51

As a young entrepreneur, and you've been very successful, the self-made billionaire, and we can take the word billionaire out if you decide you don't want it in. I just don't know how certain people feel about that. And the founder of Scale AI, where do you invest? Are you investing your money in certain places, like certain fields that you see continuing to do well?

01:42:16

So almost all of what I'm focused on is how do we scale, how do we make it super successful, and make sure that we... One of the things I think is really important is how do we make sure we create as much opportunity community through AI as possible? How do we make sure there's as much jobs? How do we make sure that everything that we're talking about actually is what happens? Because I think, no, someone's going to have to really work to make sure all these jobs show up and all this stuff actually happens the way we're talking about.

01:42:43

So there's going to be new industries that are going to pop up even.

01:42:46

Totally. Yeah. I think just in the same way that it's hard. Nobody could have really predicted that podcasting was going to be this huge thing and this huge cultural movement. Yes, it's true. But But it is one, and it's amazing. It's awesome. And that's going to happen in little ways in all sorts of different industries. And it's going to be really exciting.

01:43:09

What are some of the things that excite you about technology right now? Where do you see AI in technology in five years, 10 years?

01:43:18

Yeah. So some of the areas I think are really exciting. So one is definitely everything to do with health care and biology. That's moving really, really fast. And as we were talking about, I think, legitimately, in our lifetimes, we could see cancer being cured, we could see heart disease being cured. We could see some really crazy leaps and advancements in that area, which is super exciting.

01:43:46

Could it create a way that we could live forever, do you think?

01:43:50

There's definitely people working on that. This is getting crazy and very sci-fi, but some people think that there's a way for us to keep rewinding the clocks on our cells so that we'll always feel young, and all of our cells will actually always stay young. It's, I think, scientifically possible. And I think if we can get there, that's obviously incredible. So there's people working on that. At the very least, I think we'll be able to lengthen our lifespan hands pretty dramatically, and maybe we could get to that. Yeah.

01:44:34

Yeah, because I always envision this. There's a time where it's like, okay, this group lives forever and this group doesn't. And there's just that parting of two different way. People head not in just into the end zone of the Lord, and then there's other people just loiter who are going to be loitering around for a long time. And what that would be like, that cut off.

01:44:51

Yeah. I'm not that worried about it, but it sucks to be that cut off. We're like...

01:44:57

Maybe.

01:44:59

Maybe, yeah. Maybe. Maybe. Yeah, it's a weird thought.

01:45:02

Because it would also be brave. I mean, you'd be the astronaut. I mean, dying, you're just an astronaut, really, into the Ether. You don't know what's going on. You're the Lewis and Clark at a Lord at that point. You're out there. And then if you stay, you always know what's going to happen in a way because you'll be here, which would be exciting, I think. But then after a while, you might be like, dang, what happens if you die? That's the bravest choice you could make.

01:45:28

That's true. Probably. Yeah.

01:45:30

Do you see AI having any effect on religion?

01:45:34

Yeah, I think one of the things that... Something I believe is like, I think that as AI becomes One of the things I think is really important is that we are able to educate people about AI and help people understand it better and better and better. Because I think it's this scary thing that nobody understands or people feel like a boogie man or feels like, is this just this thing that's going to take my job? That makes it scary. And I think that affects people's spirituality, that affects how people contextualize themselves with the world.

01:46:17

Yeah, you could lose your purpose. If your purpose is a job that you feel it's going to disappear, that could already be causing you to feel that way.

01:46:26

Yeah. But I think if we can explain AI more And ultimately, it is a cool technology, but it's not that magical. It's just data, and you crunch that data, and then you get these algorithms. And so it's not Yeah, some people talk about in this crazy way. But I think as long as we are able to explain what AI is and also explain what opportunities it creates over time. And to me, it's about getting this relationship between humanity and AI right How do we make sure that this is something that enables us as humans to be more human and us as humans to do more and experience more and be better and all these things versus something that is scary or will take over or anything like that. I think that's really important.

01:47:18

What's a place like... If I go to ChatGPT, is that Scale AI? Is that the same thing or it's different?

01:47:23

We're actually under the hood powering all the different models and AI.

01:47:30

Are all the different AI systems under the hood?

01:47:35

Yeah. We help power ChatGPT and OpenAI. We help power mostly from the data perspective.

01:47:41

Do we know if the answer came from your company or other companies? If we ask it a question?

01:47:45

How do we... There's probably no way to literally tell. But yeah, we help power Open AI, and we help power Google's AI systems and Meta's AI systems, and help power all the major AI systems.

01:48:00

How can a regular person just at their home, say there's a guy who's been listening to this today, he wants to go home today, and he just wants to learn a little bit of how AI works. He could just go on to ChatGPT and ask it a couple of questions?

01:48:10

Yeah, you could ask ChatGPT how AI works.

01:48:13

You could ask it, what's the history of my town? Can you research my last name? Maybe and see where it came from. Maybe what are some innovations that could happen in the next few years? There's different little things. You can just ask it. That's how you can start to have a relationship with asking and learning about using-And you're seeing what it's good at, what it's not good at.

01:48:36

Right now, AI is still really bad at a lot of things, most things. This is why I think when people understand it and really get a feel for it, it stops being as scary because I think we think about it as the AI from the movies that are all powerful and whatnot.

01:48:54

Yeah, you think of it as a robot that's going to show up and just start driving your truck around or something, and you're like, What the What the fuck do I do? You know what I'm saying? My keys were in my truck. I can't even get in my house now. I think there is this bogeyman fear. Yeah. But that's not the truth.

01:49:10

It's not the truth. Yeah, that's not the truth. To me, it's like we have the choice to make sure that also it doesn't become the truth. Definitely, people building AI, but just in general, everybody in the world has... We should all make sure to use it as as something like an assistant, as something that helps us versus think about it in a scary way.

01:49:37

Well, getting to learn how to use it, small ways, whatever, is certainly a way that you're going to start to realize what it is. And it's easy to just sit there and say it's horrible and without trying to use it to learn about it. Sometimes I won't learn about something just so I can continue to say it's a boogie man because it gives me an excuse if I choose not to learn about it in my own life. Chatgpt has become a proprietary name, like Band-Aids or Ping-Pong. Is that okay that it's like that?

01:50:10

Yeah, I think that's totally fine. I think basically, there will probably be more AIs over time that people get used to and use. And like anything, in America, there will always be a bunch of options for consumers and a bunch of options for people to use, and they'll be good at different things. So I think right now, we're just in the very early innings of AI. But over time, we're going to have just like how for anything, for clothes or for energy drinks or for whatever, different people have different tastes because there's going to be different things that different AIs are good at and other things that other AIs are bad at.

01:50:55

Is AI right now smarter than humanity?

01:50:58

No. Yeah. Yeah. So I think what AI is good at because it's ingested all of the facts, it's ingested this the body of water is really big, and it's ingested so many different facts and information from all of humanity. It definitely knows more or just like how Google knows a lot more than any person does. So it definitely knows a lot more, but it's not like There's tons of things that humans can do that AI is just fundamentally incapable of doing. So it's not like a... I don't think you can even measure one versus the other. There's very different lines of intelligence.

01:51:46

Could AI just create a better AI at a certain point? Could it just be like, Hey, AI, create a better AI, and it could do that?

01:51:53

Yeah, this is a good question. Actually, this is another really hot topic in the AI industry is, can you get AI to start doing some of the engineering and some of the improvement on its own?

01:52:09

That's scary because then it's making choices then. It's becoming the lifeguard. It's becoming the water and the Coast Guard.

01:52:18

Yeah. So this is something... I mean, my personal thought is, I think this is something we should watch a little bit, and we should make sure that humans always have the steering wheel and the the control over because like you're saying, it's a a slippery slope before that gets a little weird. But I think that we can maintain that. We can make sure that we don't let the AI just keep iterating and improving on its own. Yeah.

01:52:49

And in the end, you can always shut off your computer and phone and go for a walk, huh?

01:52:52

Yes. Yeah, totally.

01:52:54

It's not like it's going to come out and just slurp you off or something if you're trying to be straight or whatever. And it's a man. I don't even know. Is AI male or female?

01:53:03

I don't think it has a gender.

01:53:05

I wonder if it'll decide one day. It'll be like, Hey, I'm Frank.

01:53:10

Well, there are companies that try to program the AI to adopt various personas Oh, yeah.

01:53:16

I got the Indian GPS guy. They're right.

01:53:20

On Meta, on Instagram or whatever, you can get an AI that has Awkwafina's voice, for example. Oh, that's cool. Yeah. And it's funny. The Awkwafina AI is funny.

01:53:32

You're a self-made billionaire, which is pretty fascinating. Congratulations, I think, to have a... I think that money is energy, and it flows to certain places and stuff. And Congratulations. That's got to be fascinating. Was that scary when that happened? You just made some money? Was that a scary thing? Did you guys grow up really well? What was that like?

01:53:55

No, I grew up, I think, solidly middle class. We We weren't like- You weren't trapping or anything. Yeah, but solidly middle class. And I think one of the crazy things about tech and the tech industry is that we've been just building this thing up over time. But one of the things that happened is AI all of a sudden became... There was so much progress in AI, and it became the biggest this thing in the world. All of a sudden, anywhere I go, everybody is talking about AI. It didn't used to be like that when I started the company. Ai was just a niche topic, and now it's like, anywhere you go, I'll just be walking around and I hear random conversations about ChatGPT and AI and robots and all this stuff. And so it's been crazy to experience that and be a part of that I started working on this company almost nine years ago. When I started working on this, it was this obscure thing. I always knew that it was going to become bigger, but I could have never predicted what was going to happen to AI.

01:55:15

It's fascinating. It's almost like you were just standing on the bingo number. They got called by time. Yeah, exactly.

01:55:23

It's surreal. It's surreal when that happens.

01:55:25

I can only imagine that. Are your parents pretty They're out of you? What's that like?

01:55:32

Yeah. At first, they were like... At first, I dropped out of college, right?

01:55:36

Yeah, that's true. In Asian culture, that's not a thing you do, right? Not good, yeah. Yeah. That is the opposite of being Asian.

01:55:43

Well, my parents both have PhDs. I have two older brothers. They both have PhDs. Everybody in my family, they've gone through all of schooling.

01:55:54

They're like, Alex is not doing good. That might have been Spanish. But they're like, Alex is not doing good. Yeah.

01:56:01

So they were pretty worried at first. And I told them a little bit of a white lie that I was like, oh, no, I'm just going. I'm going to go back. I'm going to get this tech thing out of my system, and they'll go back and I'll finish school. Obviously, that hasn't happened yet. But yeah, they were worried at first. Now they're super proud. They're super, duper proud. And I owe everything to my parents. They're awesome. And seriously, my parents are super brainy. They're both physicists. They would teach me about physics growing up and teach me about math. And that's what led me be so good at the competitions.

01:56:43

That's cool. Are your grandparents from China or no?

01:56:51

My grandparents are from China. Yeah.

01:56:52

Does your family have a lot of Chinese culture?

01:56:55

This is interesting. This is true for a lot of Chinese Americans is that there's Chinese culture, and then that's almost like it's very different from the Chinese Communist Party and the current government. Because basically, one way to think about China is, I mean, China is a culture and a civilization that's been around for thousands and thousands of years. There's a very long-standing culture. Oh, yeah. But that's very different from the current Communist Party and the current Communist regime.

01:57:33

Oh, for sure. I think most people probably think of... I don't know. It's funny. I never thought about that. I definitely think about them as different. I definitely don't think if I see a Chinese person, I don't think, Oh, that's a Communist person. Yeah, exactly. I mean, that would be... Yeah, that's great. I think that's crazy. If some people thought that, I just... Yeah, I think somebody has a crazy long history like, damn. Then maybe almost like, do you feel like some people in certain parts of China are almost like a captured people then? Do you think that's a...

01:57:59

Yeah, I mean, the Uyghurs that we're talking about, that's just horrible what's happening to them.

01:58:04

I didn't even know about that, man. Thanks for putting me up on that Uyghur game.

01:58:06

Yeah. I mean, some of the stuff... I mean, we don't really... First of all, the world is a huge place, and there's all sorts of both great and bad things happening all over the place. But yeah, no, I think a big part of this is we want to make sure we have governments that believe in democracy, believe in liberty, these kinds of things.

01:58:31

Yeah. With being somebody that's able to be smart and conceptualize stuff, do you start to get... Do you have any insight? This might be the last question I have for you. Do you have any insight on the afterlife or what happens I was like, I never really thought about it, talking a real math guy about that.

01:58:51

Yeah, I think- What's the sum zero game or whatever?

01:58:55

What's the- Yeah.

01:58:57

I guess the way I've always thought about it, partially because both my parents were physicists, was I always feel like people live on with their ideas and they're like, what are the things... The things they put out into the world, that's how you live on. Because in math, everything you learn about is a theorem named after a different person or idea named after the first mathematician or the scientist or whatever to figure that out. Einstein lives on because- Bagels?

01:59:36

That was a shitty joke, but thank you. Oh, yeah, because E equals M C², all that stuff. Yeah. I know what that stuff is. Jesus, sorry, but I started pretending.

01:59:50

No, no, no. Exactly. E equals M², that's Einstein. And so I think I always feel like people, you live on by your ideas and what you put out into the world. And that's always how I've thought about it.

02:00:04

So you don't get some deeper thought about... Since your brain is able to be... Because you have probably a unique brain, right? Everybody has a unique brain, but I've just never asked somebody with your... I've never asked your brain this question. Do you get some further insight about what you think happens when you die?

02:00:25

Well, I think that one of the things that gets talked about a lot in Silicon Valley, where I live, especially, is the simulation, whether it's all a simulation and whether... Do you watch Rick and Morty?

02:00:39

I don't watch it, but I'll start.

02:00:41

There's some episodes where it gets into this, and I think it covers it in a pretty good way. But what if all of humanity is just almost like an experiment or a video game that some other civilization is running? Yeah. That's the one that That fucks with particularly people's mind and tech a lot because all day, every day, we're out there trying to make computers and make simulations and make things that are more and more sophisticated and advanced and capable. And so the mind fuck is like, oh, what if everything we know is just a simulation from some other civilization?

02:01:25

Or if we advanced it enough that we're able to make this happen and seem real.

02:01:29

Yeah, Exactly. Yeah, exactly. So I think one of the things that with AI and with a bunch of other things, well, even just in the past 30, 40 years, video games have gotten so much more realistic, right? It's crazy realistic. So we've seen that happen. In a hundred years, would we be able to simulate something that feels pretty realistic? I mean, probably.

02:01:57

Could we be avatars for for some other thing?

02:02:01

Yeah, exactly.

02:02:03

Damn, dude. I'll say this, my avatar is a pervert, brother. Have you met any guys like Jensen Hong? Is that his name? Yeah. You have? Yeah, totally. Wow. What is it like when you meet some of these guys? Have you met Elon yet?

02:02:17

Met Elon? Yeah. Met Elon a few years ago. Yeah.

02:02:20

Got to know him. He's the fucking...

02:02:22

Jensen's the goat.

02:02:24

Yeah, Kanye. He is?

02:02:25

He is, yeah. He always wears this leather... You see that leather jacket? Yeah. He We hosted this dinner.

02:02:31

And Feng, is that a Chinese name? Feng?

02:02:33

Feng. Yeah, he's Taiwanese, I think. Yeah. Or Taiwanese-American. But in 2018, so when we were a baby company, we threw this dinner in Silicon Valley, and I Yolo invited him. This was years and years ago. And I didn't expect it, but he said yes, and came to our dinner. And he came. It was at this restaurant in Silicon Valley. And he came and just told the craziest stories. He went to boarding school. I think he's probably told this story, but he went to boarding school. And his parents, when they came to America, they wanted to send him to boarding school, but they didn't. And so they just sent him to the first boarding school that they found on Google or something. They I found... It wasn't even Google at the time, so that they heard about. And that boarding school happened to be a rehab boarding school. So it was like... He was like this.

02:03:42

So fucking, it's a halfway house. He's just sitting there So he's there learning with people who are detoxing?

02:03:48

Yeah. So he told me this story about... He told this story about how he was just this kid at this boarding school that everybody else had all these crazy backgrounds. And he got by and made his way through that school by doing everyone's math homework and wheeling and dealing that way. Oh, yeah. And you could see that he learned how to wheel and deal and sell and all this stuff all the way from way back then. Because his story is pretty crazy. Really? Yeah.

02:04:24

We have to research him a little bit. Maybe we would get to talk with him one day.

02:04:27

Yeah. Jensen's awesome. Yeah, all these people in tech, they're all real people, but they all have the craziest backgrounds.

02:04:36

Yeah, man, it's just so funny. Whenever I met you, I just didn't... I figured that since you were with Sam Altman, that you were probably a tech guy. But yeah, I didn't know. I think maybe somebody said he's in the AI verse. But you just seem like such a totally normal... I would not have thought that you were just... I don't know. I guess sometimes you think they're going to be super quiet or not have a lot of different thoughts. But yeah, it was cool, man. We had a good time. I'm glad we got to link up.

02:05:08

Yeah, totally.

02:05:09

That was cool, bro. You might even be my first Chinese friend. I think second, probably, Bobby Lee. I don't know. He was denying it, but he'll come around. Alexander Wang, man. Thank you so much, dude. I bet your whole family is super proud of you. Yeah, it's exciting, man. Thank you for coming to spend the time with us and just helping us learn and think.

02:05:31

No, thank you. This was awesome. We were talking about this before, but I just want to make sure that people all around the world, especially Americans, aren't scared of AI because it's going to be really cool and it's going to be amazing, but we need to remove the bogeyman component. And thanks for helping me do that.

02:05:51

Yeah. No, man. I definitely feel differently about it. I feel like it's a tool that I can use, right? And even I don't know how to use it, so I'm trying to figure out. One more question. How do you keep it from becoming like an advertisement trap house like the internet's become? The internet's just pop-ups and ads and fucking best buy, trying to beat you over the head on some shit. How do you stop that? How do you keep that out of you you guys as waters? Or do you have to go there at some point to make money?

02:06:19

Yeah. First, I'm hoping that the AI industry as a whole avoids advertising as much as possible because it's It's very different. It is a tool that people can use, and people can use to start businesses or make movies or make all these different ideas happen. And I would much rather it be a tool that doesn't become an advertising thing. I wanted to make sure it's a tool that helps people first and foremost. So I think there's a choice here, and we as an AI industry just got to, I think, make some of the right choices.

02:07:00

Yeah. I think there would be value in staying as pure as you could if you could find a way. If there's other money to be made on the side, it almost seems sometimes like it could be a trap.

02:07:09

Yeah. And I think I want to make sure that people don't feel like they're being used by the AIs. I think that'd be really bad if we ended up there. I don't think we need to make it like that at all. I think we can make sure the AI is helping you do things, is super helpful, isDuper is a thought partner, is an assistant. Those are things that I think we want to make sure AI stays.

02:07:37

Gang, baby. Wang, gang, Alexander Wang, man. Thanks so much, bro.

02:07:42

Yeah. Thanks so much for having me.

02:07:43

Yeah, it was awesome, man. Danny Dane, shout out to Dane, who came up. And Daneke lives in Franklin, right? We got to get to spend some time with him.

02:07:50

Yeah, lives in Franklin.

02:07:51

Whenever you're out there.

02:07:52

And shout out to Alex Brusowitz, who we met through.

02:07:56

Yeah. Who else on your team is here, man, today?

02:08:00

We have Joe Osborne. We have Danny's whole crew, so Arjan and Clayton.

02:08:08

Nice. We'll have to get a group picture. We'll put it up at the end. Thank you, guys. Have a good one, man. I'm just floating on the breeze, and I feel I'm falling like these leaves.

02:08:18

I must be cornerstone. Oh, but when I reach that ground, I'll share this peace of mind I found I can feel my bones. But it's going to take.

AI Transcription provided by HappyScribe
Episode description

Alexandr Wang is the founder and CEO of Scale AI, a platform that provides data training for AI programs. In 2021 he was named the youngest self-made billionaire in the world by Forbes at the age of 24.
Theo is joined by Alexandr Wang to talk all about AI and how it’s changing our world fast. They discuss Alex’s choice to drop out of MIT to pursue this field full time, the debate over whether it’s creating or eliminating jobs, and how soon regular people will start to see these programs in their everyday lives. 
Alexandr Wang: https://www.instagram.com/alexanddeer/
------------------------------------------------
Tour Dates! https://theovon.com/tour
New Merch: https://www.theovonstore.com
-------------------------------------------------
Sponsored By:
Celsius: Go to the Celsius Amazon store to check out all of their flavors. #CELSIUSBrandPartner #CELSIUSLiveFit 
https://amzn.to/3HbAtPJ 
BetterHelp: This episode is sponsored by BetterHelp — go to http://betterhelp.com/theo to get 10%off your first month.
Modiphy: Get 50% off the last website you’ll ever need at https://modiphy.com/THEO 
-------------------------------------------------
Music: “Shine” by Bishop Gunn Bishop Gunn - Shine
------------------------------------------------
Submit your funny videos, TikToks, questions and topics you'd like to hear on the podcast to: tpwproducer@gmail.com
Hit the Hotline: 985-664-9503
Video Hotline for Theo Upload here: https://www.theovon.com/fan-upload
Send mail to:
This Past Weekend
1906 Glen Echo Rd
PO Box #159359
Nashville, TN 37215
------------------------------------------------
Find Theo:
Website: https://theovon.com
Instagram: https://instagram.com/theovon
Facebook: https://facebook.com/theovon
Facebook Group: https://www.facebook.com/groups/thispastweekend
Twitter: https://twitter.com/theovon
YouTube: https://youtube.com/theovon
Clips Channel: https://www.youtube.com/c/TheoVonClips
Shorts Channel: https://bit.ly/3ClUj8z
------------------------------------------------
Producer: Zach https://www.instagram.com/zachdpowers
Producer: Nick https://www.instagram.com/realnickdavis/
Producer: Cam https://www.instagram.com/cam__george/ 
Producer: Colin https://instagram.com/colin_reiner

Learn more about your ad choices. Visit megaphone.fm/adchoices