The Shintaro Higashi Show - Peter Yu Returns! | The Shintaro Higashi Show

Episode Date: June 30, 2025

Dr. Peter Yu returns to the podcast to share exciting updates on his post-PhD life, including his move to Silicon Valley to work on autonomous vehicles. The conversation dives into the state of self-driving technology, Peter's role as an applied scientist, and how his research background plays into cutting-edge developments. The trio also explores the biomechanics of judo throws, surprising parallels with golf, and the challenges of life transitions after academia.Join my Patreon for:✅ 1-on-1 video call coaching✅ Exclusive technique breakdowns✅ Direct Q&A access✅ Behind-the-scenes training footage🔗 Subscribe & Support Here: https://www.patreon.com/shintaro_higashi_show Links:🇯🇵 Kokushi Budo Institute (The Dojo) Class Schedule in New York, NY 🗽: https://www.kokushibudo.com/schedule🇯🇵 Higashi Brand Merch & Instructionals: https://www.higashibrand.com📚 Shintari Higashi x BJJ Fanatics Judo Courses & Instructionals Collection: https://bjjfanatics.com/collections/shintaro-higashi/David Kim YT/Instagram:  @midjitsu ​

The Shintaro Higashi Show is sponsored by Hakuin AI. Hakuin AI helps you measure, predict, and solve customer churn. Visit Hakuin AI and start your free churn audit today. Judo TV, your premier destination for live and on-demand judo coverage. Never miss a throat. Higashi brand.

Train hard, live strong, wear Higashi brand. Hello everyone, welcome back to the Shintaro Higashi show with David Kim and Peter Yoo. Peter Yoo is back to visit us. What's going on Peter? Nice to see you man. Yeah, it's great to see you. I haven't talked to you guys in like so long, I feel like. Yeah, it's been a while. Yeah, it's been so... But yeah, I'm in California now. I know there's some like time difference issue. We had to go over compile I thought you know, I had to come and visit you guys on this podcast. I've been watching I've been watching David's been making some good improvements Yeah, yeah looks better, you know sounds better

You finished your PhD? Yeah, it was a quick five-minute update You finished your PhD. Yeah, I'm a... Give us a quick five-minute update. Yeah, so I defended and... Man, I can't remember now. May. At the end of May. So, it's Dr. U for you guys now.

Oh snap! Yeah. The doctor. And I took a couple of weeks off, just hanging, packing a little, played a lot of golf actually. I forget. Very cheap to play golf in Michigan. I went to, of course, I did judo with Christian, if you know. Okay, good.

Then I started my new job in the Southern Valley. I'm in the autonomous vehicle space now. It's a company called WAVE, W-A-Y-B-E. So it's a relatively smaller company than like Waymo or Tesla obviously. Is it a public company? No, it's still a private startup. We just raised Series C last year.

We? Yeah, well, I didn't do anything. But yeah, so this is my first week. I started last Monday. So let me ask you a question about the autonomous vehicles. Now that we're all Asian in here, we just nerd out and talk about science and stuff. I just watched an Elon Musk interview about why he chose to go against LIDAR and Waymo uses LIDAR. What about Wave? What do they use? So we don't use LIDAR right now, but I think our approach is that we're trying to just kind of take whatever sensors that are available. If a lidar is available in the car, because we're not, we don't make cars.

We are trying to sell our software to other OEMs. I call it the car companies. OK, yeah. If the car company has a lidar, we're going to do, we're going to handle that. If the car company wants radar in it, we're going to handle that. If the car company wants radar in it, we're gonna handle that. But cameras are primary. So in that sense, we are pretty close to Tesla. We actually, my company actually started the first, that was the first company to kind of try this, what they call the end-to-end approach before Tesla. Yeah. Interesting. Interesting. Yeah. So what do you do there? So I'm a, I'm a, my job title is Applied Scientist and... Right, side that's so present. I know right. But I mean yeah, I work on the

the brain of the autonomous vehicle system. Yeah, like basically, how do we make it smarter? Basically that's what I'm trying to do. So where are we in the whole autonomous vehicle thing? Like Waymo's already out there, has 1500 cars and fleet or something like this. Tesla just did robotaxi release in Austin. It's been going for four days. I myself have the new Model Y.

It has full self-driving. You don't, where you pay for that? Where is it right now? Did you? Boy, it came with it, the new Model Y, Juniper, the limited edition one came with the FSD, and I have it, and it's absolutely amazing, but it does make some mistakes that are kind of scary.

Yeah. Uh-oh. Yeah. I mean, it's hard to say like what the current status is in the sense that like, I don't know what, like what kind of, what your reference point is, but it does a lot of like, it's, these problems are always like this,

it's a power law, like 80% of like the, 80% of the work takes 20% of the effort and then the last 20% of the problems take 80% of the work. And that's where the battle is. I think we just said something like that the other day, right? Yeah, Pareto principle, right? Pareto principle. Shintaro brought it up on his own, unprompted.

It's not power law, right? It's Pareto principle, Gini coefficient, whatever one of those things. Power law, I mean, yeah, but same idea. Yeah, yeah. Long tail. Yeah, long tail, long tail. So that's the thing. I mean, so I think all the autonomous vehicle companies

are trying to solve the long tail problem, which means all this unexpected thing, unexpected situations that you cannot always, you can't collect a lot of data off. Right. So how do we do that? So, yeah, I'm kind of trying to help out in that regard. Yeah.

So are you actually using your field of research in this work, or is it more just... Yeah, yeah. It must be to some extent, but like, is it directly hitting the sweet spot of your research, or is it something sort of adjacent? Yeah, so I can't comment too much on it. It's all like, you know, right. It's a very competitive field spot. I'll say, yeah, I was hired thanks to my dissertation research. On the Model Y, right? Yeah.

Yeah. You know, like in New York city, you're not allowed to take a right on red. Yeah. Yeah. Yeah. It takes a right on red. Yeah. Probably because.

Can your software solve this? So that's, I'll just say that's one of the challenges that's left like kind of like all this like reason capabilities or like generalization like kind of reading the rules, knowing the local laws and then reason that in New York City there's a law and then you pop. Oh, because I think the problem in Manhattan is that everywhere there's no right turn or right. Yeah. Yeah. But they don't even bother putting up the sign, right? So sounds like prompt engineering to me. Yeah, so I mean that's like the whole thing. Yeah, like can we get some of the abilities from LLM back into self-driving cars?

Now how far are we away from having like all right so the rollout for the rover taxi happening and I follow this Tesla thing closely that gets expanded and then all of a sudden if you have a hardware for a model Y like all of a sudden that thing can go and you know drop me off in the city and the worst part about driving to the dojo this is one of the selfish I look into this, right, is going to the city, get into the dojo and then me looking for parking. Yeah. I dropped myself off at the dojo. It's like, all right, go around the block and then pick some people up, make some money,

and just come pick me up after the judo's done. How cold are we that that's, that's all I care about. Just me selfishly. That's what I, that's what I want to know. This is why we brought you back, Peter. These are the questions that people want to know the answers to. I can't really make a prediction on that. I don't know. I don't think anyone knows. Give me a couple of months, years. Years, I can't give you a timeline because I really don't know. I mean it's like we probably are a few breakthroughs away from actually I think they're like levels defined by the highway agency the level one two

three four five right? Yes. We have level two which is like basically a lane following kind of thing and we're like footing with level three and four. I think Waymo is the one. I'm gonna mess up. Let me look it up. But I think level four is kind of like Waymo is in San Francisco. What level is? But yeah, I'm probably,

like there's some nuance to it. People can look it up. Yeah, you can look it up. But we're level'm probably like, there's some nuance. People can look it up. Yeah, you can look it up. But we're level five is like, we don't know. Yeah, it's like fully capable. You don't need anything. And man, we're on level five. I don't, I mean, it's probably ways away.

And at the same time, it's like, again, that's, we are supposed to level five, right? We humans, but we still make mistakes and we get into accidents. And then, you know, I think car, it's like one of the deadliest thing in the world. Yeah, for sure. Yeah.

So in that sense, that's the risk as a society we're willing to take. And, you know, some countries have lower risk tolerance. So it's hard, they make it harder to drive. America has, as a society, decided that we're gonna have a very high risk, so we let 16-year-olds drive after taking a month. Crazy, like my son. Yeah, it's crazy.

I think that's absolutely nuts, but we've built our society around that, so I don't think that's the right way. Driving fast and rock and roll, man. That's what America has built on. So that. So I don't think that's the last way. That's what America is built on. So that's the thing. I think now we cross into this more societal issues. It depends on what we decide as a society.

I think Detroit tried to like, I think Michigan built a self-driving lane and I-94 or something, but I've been on that lane. It's silly. It's silly. But I don't even know what they're trying to do with that. Anyway, but yeah, those are some of my thoughts. I mean, it's not flashy. I can't really answer your question. Well, you've only been there for a week. Also, yeah. Autonomous vehicle is kind of like a new field to me too, because I wasn't, that wasn't my... You didn't write your dissertation on...

No, so what does work like? Like going back to work after all these years just kind of grappling, hanging out and getting your PhD, like what's that like now? Like you gotta go to work every day. So it's, I think it's more... It sucks! I wouldn't say it's relaxing, but it's more enjoyable in the sense that, like, I think it's more I wouldn't say it's relaxing but it's more enjoyable in the sense that Like I think the worst part of the piece is that you could never rest

you yeah, you are always in the Cycle like you have in the there's no end no end Like you have to make a conscious effort to shut off and rest in the... There's no end. No end. You can't really like, you have to make a conscious effort to shut off and rest. Cause this, it's kind of almost like running your own business in a way. Like if you don't do it, it's not gonna, nothing's gonna happen.

So... You did it for six years, right? Six years? Five years, a little bit less than five, like four and eight months. Yeah, four years and eight months. I mean, that's the standard. And then if you wanna become a professor,

I wasn't like the best PhD student. I don't have the best publication records. But if you wanna become a professor, you need like a stellar publication record. So people will take an extra year or two to publish more papers. And on top of that, to post up year to publish more papers and on top of that to post up to do more research and then like it's very competitive.

You knew you were going to industry. Yeah I mean that was my original intention and then I quickly realized that yeah the people who become professors they're like going back to the kind of like the sports analogy I used in the peer review episode, they are the superstars in like superstar athletes. Yeah, if you become like their superstar coaches, superstar athletes. So I am I'm just I don't know your I got some medals at continental or something. some medals at continental or something. Like, yeah. Yeah.

So work, you can kind of, there's some clear separation and I think there's more support for you because as a PhD student, you kind of have to do everything but yeah, it's a little nice. And then I can start thinking bigger. That's another good thing. Do you have to go into work physically every day or do you have remote days or what are that?

All this stuff. I think we kind of have a, you have to come into the office a certain amount, number of days, but I prefer going in. Because you have an empty apartment, that's why. Yeah, that's also, that's a big reason. I don't have any furniture right now.

I just gave it up. And your wife moved there with you or she's gonna come after in a little bit? So she came here for a week to help me settle in and then, but she had to go back because her contract runs until September in Michigan. But she does have a job lined up here in October. Of course she does.

She got it all together. I mean, she hates it when I say this because it is still hard, but her job search was so much easier than mine. Like it's MDs, man. They got like boys and girls go to medical school. No, you're just destroying so many dreams, Peter. go to medical school. Destroying so many dreams. Peter, have you decided on like a Jojo to train already or so?

I did some preliminary research, but you know, you know, Jason, you know, our biggest patron, he goes to another Shintaro here Shintaro Nakano's gym and then looks nice yeah so I might pay them is that close by it's actually very close yeah it's very close to me so if I five to ten minute drive or something it's perfect I'm gonna send you guys both an email and connect you guys when you guys can you know what is it with you and Shintaro's that's a little creepy oh you know that of course that bad and what are you kidding I saw a couple weeks ago out of judo tournament oh I mean Shintaro is like the bread pit of judo so I have a funny story about that one of his students was texting me thinking that he was talking to another

Chintaro. That's so funny. He's like, at the end of it, I was like, who do you think you're talking to? Chintaro's like, yeah, you know what? What are you talking about? What tournament do you expect me? What is your name? Let me live up to your membership on my body.

And he gave me some randoms. Like, dude, I don't think you're my student, dude. That's so funny. He was like, yes, I am. I'm like, that guy's like, man, Shintaro's a real asshole. I just saw him at training the other day. Yeah, we're going back and forth, and then finally we figured out, I was like, wait a minute.

You know you're talking to Shintaro Higashi, and you're like, oh, shoot, I thought I was talking to Nakano. He knew who he was talking to. That's so funny. He knew who he was talking to. Yeah. That was all the ruse. Man, Peter, you look great. You look healthy.

You miss being on the podcast? Of course I do. I mean, it's like, you know, it's always a, I think, you know, when you grow older and then you have all family obligations, I think it's important to just kind of make this type of regular hangout sesh.

Great idea for a podcast episode, David, by the way. David was like, we gotta get Peter on. We gotta get Peter on. I'm like, oh, I would love to talk to Peter. I haven't spoken to him in a couple of weeks. I mean, I'll definitely come by here and there. Like, I think, you know, just that, like, yeah,

I'm glad that I could make it on here. Like, if there's a time difference, and then, you know, I work pretty long hours right now, especially because I just started. Can they reach out and be like, listen, man, you gotta get this done by Wednesday? So So I think if there's a deadline, but because I'm on the science team, I kind of shielded from it. Like, I don't. Yeah, my stuff I work on, you know, it doesn't go out to production, so to speak, right away. So it's a more like fundamental research. Yeah, stuff. I mean, not fundamental because I'm still a applied scientist. So it's like, Oh, yeah, fundamental to the, yeah, exactly. So, you know, you don't have to, how dare you, David, how dare you fundamental

the systems I make will have to go through a lot of rounds of testing before it can get out. So it's okay. Yeah, that's that's. And then, I mean, man, and, and then if your company keeps asking, like if you keep firefighting, then something's wrong with your company. Like you gotta work on your process. Yes, it's a sign of weakness. That's right.

It's a sign of weakness. Yes. David, what was the last time you had a real job? Me? I have a real job right now. What are you talking about? No, no, you're the last time you had a real job? Me? I have a real job right now. What are you talking about? No, no, you're the owner, though. You're the guy that's running the show.

You could do whatever you want in your fancy house. I'm talking about like a real job. Probably 12 years, something like that. Oh. Damn. You missed that? No. No. I'm a little bit envious sometimes. What are you talking that? No. No.

I'm a little bit envious sometimes. What are you talking about? You are living that life right now, Jantaro. You are the owner. No, I know, but I don't have a real job. So it's not... I don't know, when Peter's like, I get up in the morning, I get a coffee, I go to the office and I see the guy, what's up, man? How's your weekend?

I don't have that. Yeah. I think it all depends on where you are in life, who you're dealing with, and what industry you're in. I mean, for me, like I had no, I guess I probably to a fault, I just can't deal with corporate BS. I see. Yeah, I think for me, I actually prefer this structure, I think. I think that's probably why I didn't really like thrive in PhD because PhD is kind of like running your own business. It's not a real job, so to speak.

So you guys want to talk a little Judo? Yeah, Judo. We haven't talked that much Judo. All right, so you did Judo recently. you guys want to talk a little to talk to any judo. All right. So well you did judo recently. So I was thinking this, you know, so I been playing a lot of golf, right? And I realized that all these all the sports that contain this turning motion, the swinging. I think evolutionary wise, I think we are best at throwing things. And I think swinging actions like golf, tennis,

judo throws, baseball, I think they're all related. And I was like, as I was like, kind of like trying to perfect my golf swing, I realized that I was doing something wrong with my own judo forward throws, trying to perfect my golf swing, I realized that I was doing something wrong with my own judo forward throws, like turn throws. From golf. From golf, yeah. And then I think it kind of goes to that full debate about kuzushi, what kuzushi means and like do you gotta, like when you, is it like, when you do your uchimara,

do you have to pull down or you have to really like see your watch and stuff like that? You know that whole thing? No, yeah, I do. I think I kind of have my own theory and I got to only experiment with it once or twice But I think it works better. So that's my current judo thought. So what was the epiphany? Like what was the change? So in in in golf what they say is that Arms only move up and down Because that's where that's the What arms are designed to do up and down it you know you just have to be turning yeah exactly

It's really we are really good at accelerating our arms to up and down But we're very weak this way because we're using like the small. Yeah. Cause you're like, we're using deltoids. No, no, it's you're a lot stronger pulling things down. You've said this yourself, Shintaro. Yeah. Yeah.

When you pull high and you look like the classical watch, you're using a lot of the small muscles, as opposed to pulling down to the side like Hanpan TV. They're making that argument when you pull down to the hip. So I think that's what it is. And then in golf too, they say drop your hands and then because your body, your hip is turning, the the arc is created the whole circular motion they really emphasize that because golf the really biggest point is the swing speed because you want to hit far and then it goes to I think with the judo too I mean if you want to throw you can't watch the yeah you have to so so the, it's really the pulling motion,

pulling down motion that really throws a person. And then you just jump with your legs, your turn, your hip, you know? And then what that means is kuzushi, this whole thing, you do have to pull up a little, but that's just to create enough room for you to pull down you know what I mean? Oh that's interesting no you tied all that together I had no idea where you're going with this but now you've tied that all together you know what I mean so like so the

yeah so the one thing that wasn't working for me was the right-sided borote seoi nage I used to be so good at it but for some reason I lost it and I could only hit the left side. I was thinking about it and it's because it wasn't working. I went back to the classic way of like what to get the watch and all that. So I couldn't really move the person because I was using small muscles. So what I did was I just focused on pulling the person down and jumping in. And it was just like, there was nothing blocking me into it. It was very freeing. And so that's when I realized and even Uchimana too, I just

focused on pushing the, I mean, you always tell me to push, push the head down, push the head down. I didn't really get that and I thought I just had to add more rotational energy. But it's more that I really have to push down like the person into the ground. But when you just do the whole leg kicking move, the spinning move, it everything kind of works together mechanically. So I realized that I could really finish you know the Kenken Uchimura is better. Like even if I don't get the unbalancing correctly, timing correctly, I could really force it down and then really finish the throw. So that's I'm gonna try to work on that more. I don't know if it makes sense to everyone but yeah that's what I've been working on.

That's what I judo thought. Interesting. I can't wait for you to do judo out there and see how that goes. Yeah I know it's a yeah I got a little target in my back, especially like. I wonder when you walk into these dojos, if people will recognize you from the line. I've already gotten like some people reach out to me saying, oh, like inviting me to their like BJJ jams and all that. I mean, why would they not want you?

You know, they would do every dojo is gonna want you. Oh, you're nice to. I don't know why it's. Shit. But yeah, so that's my general thought. Yeah. Well, this is definitely something I'm going to follow up with you about because we still have that whole idea on the physics modeling. Yeah, I know.

I don't know if we'll get to it tonight, but we're definitely going gonna follow up on this because it's, you know, this concept has been burning a hole in my mind for probably 12 years and I need a PhD's technical and ordered mind. You know, a guy once told me, you know, like, you want a PhD because it's the way they think. It's the way they think. And he knew I wasn't a PhD. So I was like, what the hell are you talking about? Yeah, I mean, I guess, yeah, that's kind of the point,

I guess. But I'll get like, I would love to have that talk, talk to you about that. But the caveat is that it's not exactly my idea. I knew it. Yeah, no, you know, well, help out as much as I can. No, but I think the idea would be more just, we can talk about like the feasibility and just, you know, what could come of it. Yeah. I think it's more, that's more interesting anyway. It's like, what could you, let's assume we can do it.

Let's assume we can make the $6 million man for the people who are old enough to remember that show. You know, what could we do with that technology? And I think, yeah. yeah, see? Yeah, that's my man. What show? It was Steve Austin, you know, very old show. There was a bionic man and the bionic woman. They're everything moving in slow motion because they couldn't they didn't have the special effects to make it go fast. I think, you know, after I start going to a gym regularly here, yeah, I could like talk about the scenic, the California judo scene too, I guess, a little bit.

Yeah, that'll be really cool. Yeah, definitely. Definitely do that. I can tell Shintaro really misses you, man. I can tell. Look at him. He really misses you, man. I miss you too, man. I miss Shintaro too. But I'm sure everyone's so happy to see your face again Peter. And again, once you started hitting up the judo scene over there and you have something good to report. Yeah, that's right. Cool. Thank you Peter.

All right, see you guys. Peter, nice chatting man. Nice chatting David.