Steven Sashen - XeroShoes.com

Welcome to another Founder Wisdom podcast. We have, uh, Steven Sachin for round 2. I had, uh, Steven on about a year or two ago. He's the founder and CEO at Xero Shoes, used on a daily basis to go walk my dogs. They're pretty cool. I even run in them. They're not made for running, but I have a fun time running in them. They're super flat. Yeah, it's not the running ones. I know you have some running ones too. Maybe I should try these as well. But yeah, Stephen's a special individual. You're a next clown, right? Like you use, or magic magician. Like, I forgot which one of the two. And you're just like this wild guy with a bunch of ideas. I love talking to you. So quick re-intro about yourself, Stephen.

Yeah, yeah. Well, after what you just said, I probably need to clarify. So I'm just a guy who's been really, really lucky and almost anything that I have found that I've found interesting, I found a way to make a living doing it. So when I was a kid, like literally when I was 11 or 12, I got into magic and I spent a lot of time. I made a living through college doing kids' birthday parties. I remember my sister called me. She's a couple of years younger. When she was a freshman in college, I said, how much allowance did Dad give you when you were a freshman? I've been working for a living since I was like 13. He doesn't give me an allowance. And she has a long pause and she goes, Don't tell him I called. And then out of college, my day job was doing stand-up comedy, which obviously is technically a night job. Because it was a night job, I went to film school and got a degree in screenwriting and ended up inventing a word processing program for film and television writers. It became the industry standard in that world. And then I've done crazy things. I've taught meditation. I developed a whole course about that. I taught a whole course on what I call applied behavioral economics, basically how to understand money and business in a way that you can use the problems that you think you have as the stepping stones for solutions. I hosted an internationally syndicated television show for a while, and let's do the, the last one, for the last 17 years, almost 17 years. I'm the co-founder and now Chief Barefoot Officer here at Xero Shoes, where we make shoes that let you feel what you've been missing: natural comfort, performance, and health for everything from taking a walk on the beach to climbing mountains to playing professional basketball in the NBA and WNBA. And here we are.

I love this. What brought you to Code? You and I, we talked about you managing some agents, some AI agents at Code at a very junior level. In your opinion, what brought you to this? And if you could give us an update about Xero as well, because last time we spoke, you were under this tariff, you were in this chokehold with tariffs. Right? Like, how's that been going?

Well, the tariff stuff is getting resolved. It's not fully resolved. I mean, we had to pay a very large sum of money when the tariffs, uh, went up significantly for what we were doing. Those have come back down. I don't remember actually if they've come all the way back down to where they were before, um, but we're of course, you know, expecting to get a refund for all that money that we spent. And it has changed our business dramatically, which is something. So that's Well, it's never gone because we know that the current administration can try to impose tariffs of any amount at any time, and they get applied until they're hopefully stricken down by— stricken down, yeah, by the court. So we're, we're living, as are many, many companies, on a bit of a razor's edge about things that could happen at the macro level, um, with inflation going on and how much people— how much disposable income people have for various things and what the the administration is going to do to be helpful or not helpful about all of that. Happily, footwear is kind of recession-proof. People always need footwear. They might not buy as many, or they might be spending a little less money on something, but we, we pride ourselves on being affordable. We started the business during a recession, so we, we know how to weather that storm.

Sorry, um, how does that bring you to codes and like what we were speaking about before we started the call?

Yeah, the way it brought me to code is that, um, in my current role there's less day-to-day stuff that I have to do. And I, I mentioned that I invented, uh, this software for actually for film and television writers back in 1992. I released it and in 2001-ish Some guys were going to buy me out. And it turned out I helped two of them— or I accidentally helped two of them discover the third was embezzling from them. So that deal did not happen. And during the time we were negotiating the deal, one of those people had been a longtime friend and consultant of mine. So he said, while we're doing the deal, why don't you let me run the company? And it was like, great. And then when the deal fell apart, the company also kind of fell apart. And I had to I literally had to fly in and in the middle of the night take back all of the hardware that I had shipped out to let them run things. And in the 25 years since then, I've spent a lot of money trying to redevelop the code, get it more up to date. Back in 1998 and before, there were no industry standards for almost anything having to do with software. And so it just needed a good refresh. But the code is very complicated to do something that looks very simple. But suffice it to say, it's so not simple that in the 34 years since I released the first version of the program, there have been dozens of companies that have come out with similar products, but none of them can do what we did as well as we did, as easily as we did it without any extra anything. Basically, we made a rules-based, object-oriented, context-sensitive database that acts like a really simple word processor. And does very complicated formatting and pagination on the fly in real time. What you see is what you get, including tracking revisions and various other things that, you know, again, no one else has been able to do the way we did it. So about 6 weeks ago, inspired by a number of friends of mine, including one of my best friends who's a very big AI guy, I threw the code into Claude Code and I spent maybe a week going through and having to basically copy and paste a lot of things until the code compiled. And then it was buggy as crap, but it compiled. And then I've spent the last 4 weeks or so, 5 weeks, I'm not exactly sure what day it is, just working. Actually, I moved everything over to Codex because Claude couldn't do some of the things I was asking it to do. It just kept giving me— the AIs are very confident. Here's the solution. And then you try it and it didn't change anything. Oh, sorry, here's the solution. And it didn't change anything. And then it'll say things like, 'Oh, that one's on me.' I mean, literally, that's one of the things it'll say. 'That— sorry, that's my problem.' It's like, 'Yeah, I know, you're the one who wrote the code.' So after days of working on one critical little thing where it— where Claude couldn't figure it out, I moved over to Codex. And Codex has a better integration with the tool that I actually have to use to compile the software and run the software. So I'm— but I'm still using Claude whenever Codex gets stuck. I will have it create a file about, hey, here's everything I tried, here's why I tried it and it's not working. What do you think? And I give that to Claude. Claude, again, very confident, says, here's the solution. It has never been the solution, but it's usually pointed either Codex or me to the actual solution. And the critical part is the me part. And I'll say why. I'm not trying to pat myself on the back, but I think anyone— this is different than vibe coding where I'm creating something from scratch. Where the one piece of advice I can give people is you need to get really specific about everything that you want this thing to do, or it's gonna be, it can end up in a different place than you think. Like more specific than you've ever been. I can't write a line of code, but I'm really good as an architect 'cause I understand object-oriented coding from an architectural level. And I'm really clear about what I want to have happen. And I can do English to programmer translation. Really, really well. So you have to do the same thing with, with AI, depending on the project. Obviously different complexity, different story. But suffice it to say, the major problems that we've run into, um, like if the code does— if, if we're working on a thing like a notes feature and then suddenly there's another problem, then the code is going to keep— the AI is going to keep looking in the notes feature to find, find the solution. And what I've had to do multiple times is identify that the real problem isn't from the notes feature, but some other thing that we never noticed before or wasn't a thing really until the notes feature made it clear. And so once I've been able to do that, then the code is very quick at finding it, find a good solution. But I have one, one issue that I was working on for maybe the last at least 3 days, maybe 4 days, and Finally it hit me and I tried one weird little thing that had nothing to do with any of the things that we were testing for the last 4 days. And it's like, I just found the problem and it shouldn't have been a problem. There's no reason it should have been a problem. It's a basic fundamental thing in the code that somehow got turned off. No idea how, no idea why. And once I identified that, we— the code, the AI solved the problem in maybe 10 minutes. And, and anyway, so that's what got me to it. So the point is, I'm not very far away from having something that has been upgraded well enough that I can re-release it, and it will be so much better than anything that's come out in the 30 years since that I'm gonna upset the apple cart in a similar way that I did way back when. And it's now faster to develop additional features that literally no other program will be able to do because of the design. And it's, you know, it's getting— it's, it's getting exciting. But it's also like going to Las Vegas, because you're, you know, spinning the wheel, spinning the wheel, spinning the wheel, it's getting frustrating, then you hit a jackpot. And it's very exciting. You just want to keep going. And the number of times I found myself, you know, awake at 1 in the morning working on something with the AI, and not knowing where the last 3 hours went. Is very high and I'm working very hard to not just work 24/7 when I have a moment to do it. And I'm also marveling at how— and I'll stop ranting about this in a second. I'm marveling at how I find it easier to talk to the AI as if I'm talking to a person than to edit the way I would talk to a person to talk the way the AI— it's all the AI needs. So I will just be more verbose and be sometimes polite or do whatever, just because that's the way I think if I'm talking to you. And it's easier to just type that out or say that than to do it in any other way. I don't think of it as a person in any way other than, again, a programmer who's a good mid-level programmer, not a high-level programmer, which is kind of frustrating. But the weird interaction, like, oh, here's my favorite one. If I'm halfway through something and it's midnight and I want to go to bed, I feel guilty waiting till the next morning. I feel like I have, you know, I have this thing in my head where it's like it's waiting for me to give the next instruction, and I feel guilty not giving it the next instruction and getting to a stopping point, which I find hysterical that that's how my brain is anthropomorphizing this whole thing.

Isn't there a framework and architecture you can create to reproduce your brain? Sure, there's taste and there's clarity in your brain, but But can't you, like, isn't that the missing piece as well? You've probably followed like all the Clawbots thing and all of that. Like, I follow some guys that have architectures around that. There's like a code reviewer. There's like, if this, it's still if this then that, right? But can't you produce something around that not to continue being frustrated around that?

I'm not optimistic. And the reason why, again, this is a very complicated program with a lot of variables, a lot of settings, a lot of very specific behaviors. And the AI has not demonstrated that it understands how all of those work at a level that will allow it to find some of these little glitches. Again, the fact that I found them— and this has been, this has been admittedly something I've been good at since, you know, 1992 when I started doing this. I'm good at— how do I want to describe this? So I'll give you a specific example, or it's going to be a little vague and specific at the same time. There was some weird thing happening around a page break, and in screenwriting there's very— a whole lot of rules about how an element, someone's dialogue or some action or a scene heading, can or can't break across a page. And there was some weird thing happening that we thought had to do with— I brought the notes feature for adding a note somewhere. And spent like quite a while, like we couldn't figure out how to fix the note problem. But then suddenly I noticed something, even though I had given the AI a lot of screenshots, I noticed something the AI never noticed. And that was essentially that the last line of this one page ended at a certain point on the page, line 57, and after something else changed, it ended up at line 52. And there's no reason that that could have happened having anything to do with what we were focusing on. And suddenly it occurred to me, oh my God, something is happening further up the page where some things that used to be there are— have disappeared or gotten changed in some way. And I scrolled up the page and immediately saw this one thing that had been visible all along to the AI, but actually not to me. Because of having to scroll up the page that it never noticed. It never saw a difference and never called out a difference, even though I had given it instructions for how to call out differences. It never— here's another very— but anyway, so that's on the debugging side.

It—

because I just know the program so well, I will think of doing things to stress test it that no, no one else would think to do, because frankly, because I've had almost 30 years of hearing people report a bug when— and I've had to figure out how to identify that bug before. I know them all. And at some point, maybe from watching what I'm doing, I'm hoping I'm going to give it that suggestion. How can you learn from what I've done? Maybe it will get smarter. But I also have flashbacks to when we had this one guy that I hired to do customer service who would call me every now and then and say something that may as well have sounded like this. Hey, the program crashes if you're standing on your left foot on an alternate Tuesday when the moon is— sun is in, you know, behind the— like crazy, crazy things that made no sense. And then I could reproduce them exactly. So there's just a level where I would argue that the AI, the coding AI currently isn't dumb enough to do crazy things that are really smart.

Yeah, well, I mean, what you're referring to is risk-taking, right? Like taking risks. Is it too conservative?

It's just these are very specific risks. It's like, how would you know to— like, if I hadn't shown it and you want to test copy and paste, which sounds like a very simple thing, there's actually— I used to have a, I used to have a giant chart when we were coding the program originally. Like, if you copy from the beginning of an element to the middle of an element, from the beginning to the end, or from the middle to the end, and you paste it in the beginning, middle, or end of any other element, what's supposed to happen. And it's a very big chart because there's a whole lot of different types of elements and different behavior for each one of those activities. Without having that chart and without having instruction test every possible variation of, you know, what I just described, it wouldn't have thought of that on its own. It's just too weird. So, and again, there's just institutional knowledge that Now that one I could have transmitted and, and still could, but it's not necessary because that was, that was a problem in the original programming, but that original engine is still functioning properly, so I don't need to do that one. But there's things that have been like that where I don't have that model in my head. I'm just trying all the different possible things I could think of that may be related. But this code base is pretty big. It's about, you know, half a million lines. And if you don't understand how every possible thing interacts with every possible thing under every possible circumstance, you're not gonna find some of these patterns.

Right. Well, I mean, there's memory, context, attention.

That's another issue.

AIs are like getting better at this, but what you're alluding to is like, yeah, gut feeling and intuition, which I also think can be taught to an AI. It's just like, oh, this time I had this data and I decided to focus my attention on this memory and pair it with this and that, and I finally landed on that conclusion. If we do that enough times, we can teach an AI to think exactly like us. Do you feel superintelligence is a thing for that reason? Will it happen?

I don't know. I don't see a reason why it couldn't, but I do see a bunch of— I don't think of it as a linear process, and certainly not an exponential process. I think there, I think there might be a kind of point of diminishing returns for a while where, where that evolution slows down. Just again, a about certain weird intuitive things. Here, I'll give you a weird, a fun example. I asked one of the AIs this morning about something— let's go back to footwear— something having to do with, um, with— oh, where did this go? Come on, come on, come on. Um, something about a thing that people refer to in footwear with about— they call it energy return, having to do with cushioning and how bouncy things are and about the amount of energy return you get.. And I asked it— actually, the specific thing I asked is what it knew about the Harvard indoor track, which, which is notorious for people setting personal bests on. And some physicists designed the track, and to sum up what they did, they said they made the tension of this, or the tightness of the spring of the track itself, slightly more tight than the spring of the athlete. And then it went on to talk about foam and the design of modern shoes and how that contributes. But everything it was saying was based on the information it got from just scraping the internet. It was all completely wrong. So it was saying— talked about something where the carbon fiber in this one particular shoe, it's acting like a spring. It's not. Then it said it's acting like a lever. It's not. It's there for functional reasons. Because it's in the middle of two big layers of foam that are called supercritical foam. Supercritical, like on the verge of exploding. If the carbon fiber wasn't there, they would just shear. But it was just regurgitating what it figured out from reading basically all of the PR and all of the bad— all the bad reviews from people who don't understand the physics of some of this stuff. So there's— there's— this is the part that I'm very curious about, is I mean, it's not quite garbage in, garbage out, but I mean, look, back to this software that I'm redeveloping. If you ask it about this software, it won't mention my name, and I'm the guy who invented it. But I— but it's in part because that happened back in 1992, so it doesn't have all that history to look. And it just— then it just hallucinated and said that the people who started the company were two people who actually worked for a different company. So I'm, I'm, I'm, I'm always mildly optimistic. Let's just say I'm globally optimistic and locally annoyed. I can see a possibility, but I can also see places where it's going to be problematic.

I don't know, maybe it's just like the— its knowledge about shoes, maybe it doesn't have enough good data. And yeah, of course, if it feeds from PR and non-technical pieces, the question is, I don't think there's a lot of technical pieces out there about shoes, right? Uh, last time we—

there's actually quite a bit.

Oh yeah.

But the interaction between the technical pieces of shoes, like how— so for example, any piece of foam, it doesn't matter if it's shoes or not, is effectively tuned to a particular weight and speed. So it'll compress and decompress optimally based on the mass that's hitting it and the velocity at which it's being— it's, uh, the, the two objects are interacting. And, and actually the characteristics of the object as well. You know, a, um, sort of like for a steel ball is different than a ping pong ball, for example. So when it comes to footwear, there's a whole other set of complications because you have all the interactions of the musculoskeletal system in the runner that are, that are, uh, called into play. So not only how fast you're running, what your cadence is, how many steps per minute you're taking, how much you weigh, how bouncy you are, which is basically as you're— when you, when you are off the ground, your hips are slightly higher than when you're on the ground. So basically you're falling a certain amount. So that amount of falling and that amount of weight change things as well. Put all that together. And then of course, um, how well— let's use your feet. There's a thing called the windlass mechanism. If you pull your toes up, it stretches the arch, the, the tendons under your foot. Um, basically it's what makes your arch be as much of an arch as it can be. Even if you have flat feet. So that's the Windlass mechanism. Pull your toes up, creates some tension, which pulls your arch in. So if you're doing that, if you're pulling your toes up, that's going to be a different interaction with the foam than if you're not doing that. If you're automatically doing that because you have a shoe that has a thing called toe spring, so it's pulling your foot up, that's going to change things as well. So then there's a whole bunch of other interactions. Um, this is inspired by a video that I watched this morning from a guy— oh gosh, um, I just looked him up, um, it's got another Steve, and I'm going to remember his name in a bit. Anyway, he did this whole thing about golf balls and golf clubs and just showing the interaction between a springy thing and a not springy thing, and just what happens. I mean, everything's a little springy, so he made this great model showing the optimal way a golf ball and club could interact. And that just reminded me that if you— the golf club is the easy one. That's going to be the shoe and the foam that's in the shoe. The golf ball is a very complicated biomechanical thing called your human body and all the various joints that are involved in running effectively and handling force properly. Can you build a model to do that? Probably. What's that going to do to footwear design? In my mind, it means that every runner is gonna need a different shoe. And even more comically, if you're running a longer distance, you really should be changing your shoes in the middle of the race as you start getting tired and your biomechanics change. So what's it really gonna do? I don't know. Can you build a really cool model? Absolutely.

I was listening to, uh, the latest record holder of the Appalachian Trail yesterday. I'm dissecting a lot of ultra runners. I'm an ultra runner myself, as you know. And was talking about like her— some of the weird techniques, um, she used, right, to complete the record. So they were comp— they were competing while she was competing against male counterparts, right? She beat the—

her—

the record of the guy by 13 hours, right? Males have like stronger muscle mass, and during the Appalachian Trail, which is like 3 weeks or some crazy shit like that if you speed do it, by the way, because if you walk it, it's like 8 months or something. Um, she— her muscle mass would get less depleted, but she was competing against a Ferrari. She was like a Honda body, right?

Yeah.

So the analogy that AI is used— because I use AI to deconstruct— it was like she would be a Honda but with no brake. That's how you beat a Ferrari on a racetrack, right? And then she would get all these hallucinations. She would do some, uh, right some dirt naps to, to counter these hallucinations that can get dangerous because you can just walk off trail, right?

Yeah.

Um, and she did the last, I think, 46 hours running because she knew that she was gonna break the record, which is fucking crazy. Uh, so all kinds of techniques like that. So she was some, some kind of, uh, she was very technical. I also studied Jim Walsley, right? Uh, do all of these athletes study the biomechanics of running that deeply? There's so much technique, right, in raising your knees and, and stuff like that. Or do they go with their bodies and they realize that even if they study it, they can't really modify their core technique?

Well, it's an interesting question. Um, so there's a researcher named Dr. Ben Onyegbu who has a thing that he— I think he calls it the preferred movement theory, which is basically that you have a particular way that you like to move. Now, the way this has been misinterpreted by some shoe companies. So they go, everyone's a unique little snowflake. You each move differently. We need to make something for you. What he actually said was, you have a preferred way of moving, and if you artificially change that, it's gonna lead to injury. Now, I said to him, are you telling me that if someone switches from a big, thick, padded motion control shoe to barefoot, that they're definitely, that their gait will change, or their gait will or won't change because, you know, preferred movement. He goes, oh, no. Your gait will definitely change. Because you don't have the shoes on. I said, are you telling me then that that's a natural or unnatural change? Because if you're telling me it's an unnatural change that would cause injury, I'm not actually seeing the evidence of that. And he said, no, no, no, your gait just naturally changes to adopt. Then you have a preferred pattern that way. Right? So that's his idea. Then there's a guy named Dr. Nicholas Romanoff who teaches something he calls Pose Method Running. And the basic idea is that in any complicated movement, um, that's a repetitive movement, there's going to be a position that you need to get to at a certain time, an optimal time. And if you're not in that position, the pose, in the right time, you're not doing the movement optimally. This came out of when he was growing up in Russia and they were doing work with the Russian ballet. And when you're doing ballet, you're doing all these exercises, you're doing pliés, you're doing all these very slow movement things that are very specific so that you're learning the pose. And so that when you go to then take, just, you know, take a plié instead and do a leap into the air, a leap starts with something like a plié, and then you're propelling yourself into the air. If you're not pliéing properly, you're not propelling yourself properly. So similar thing with running. There are preferred movement patterns. There are preferred poses that you need to get to in the right place at the right time when you're running. If you look now, the other point that Nicholas will make is as you look at people in any activity, ballet, running, bowling, whatever, the better they get, the more similar they become because they are finding this optimal pattern. There will be little idiosyncratic things. So, for example, the former world champion 100-meter runner, Otto Bolton, he ran with his right foot turned out. And people said, oh, because he grew up playing a lot of soccer.. And then some people would say, well, imagine what would happen if he turned his foot straight. It's like, who the hell knows? But the point is, other than the right, the turnout, everything else about the way he runs looked like everybody else he was running against. If you look at slow motion video of Usain Bolt, and then in a race, and you look at the other runners in that race, other than the last couple of steps where sometimes people get a little crazy, you'll see that they're all basically doing the exact same thing. It's just that for whatever reason, when he did it, it was a little better. And there's arguments about, you know, what that is. So the answer to your question is depends on who you are as a sprinter. Some people just kind of naturally do the right thing and then they get a little better. And some people, they need less, they need some kind of training, some kind of feedback to get to be that more optimal thing. Sometimes they just need a cue. There's a great cue that someone gave me. So I'm a competitive sprinter. I'm a Masters All-American sprinter.. And my friend Doug Adams from a company called RunDNA, he said, what I want you to do when your foot comes off the ground, I want you to pretend, like, literally do it as if what I'm gonna describe, you're actually doing it. Pretend you're kneeing a soccer ball to me standing in front of you. So don't just imagine it. I want you to actually knee a soccer ball to me with every stride you're taking. And that little cue changed my gait noticeably.. And I went from like, you know, feeling good to like, oh my God, it feels like I'm flying. And in fact, back to Romanov, where what it did, it got me into that pose that he mentioned, uh, essentially— oh gosh, like 3 hundredths of a second faster, which sounds like nothing, but that's something. Maybe 4 hundredths of a second faster, because he showed me— he, he videotaped me in super slow motion, like 240 frames a second, and I was just, you know, a few frames late before I learned this cue. And then once I got this cue, right on it. So, um, it's a bit of a— actually, there's a guy named Ralph Mann. He was a, I think, silver medalist in the 400-meter hurdles back in Munich, became a biomechanical engineer, and he evaluated 600 of the fastest people in the world, had video of them, and from that came up with a stick figure model of a perfect sprinter. And so What he does is he videotapes you. He makes a stick figure out of you, measures your joints and joint and, and limb lengths, makes a stick figure of you, takes a video of you, and then shows the stick figure of you doing it perfectly. And the goal is to then be the stick figure. I've argued with him many, many years ago, like 16 years ago, that the way he's currently having people try to learn to be the stick figure is not the optimum way to learn to do that. But the point is there are some ways of moving that are just biomechanically better, that are— the physics is better. And if you can do those, there's a good chance it's gonna be an improvement. If you are a complete freak of nature and you don't need to do those and you're still, you know, world champion, why argue with success? Because, you know, that's just too risky, frankly. And it could be— and Ben Onig could be right in that situation that making an— and a change because you think it will make you better, straightening out Otto Bolton's foot, could screw things up for, you know, maybe until it becomes natural again. Don't know. But if your living depends on doing that thing and you're already, you know, making a great living doing it, you know, don't mess with your paycheck.

Right. Do you feel that barefoot running will withstand the test of time? Seems there's deep science. There's also, like, counterpoints to barefoot, right? Wait.

Hold on.

In my case, I've seen, like—

Hold on. You're asking the wrong question.

Okay, but let me finish it. I didn't even ask the question yet. In my case, in my— I mean, I use your shoes, right? And I've been using, um, yet like Vibram shoes for quite a while, and I use them for training, not for actual races. For actual races, um, I'm using Hokas, which could be in between, right? Like, I mean, the, the, uh, the Speedgoat ones, right? Like, not the Ironman's, uh, the cushion ones, but Speedgoat is good for a non-injury standpoint. There's like, there's not extreme questions in there, and these will last me like through 100+ kilometers. So what do you have to say about this blurb?

Okay, so let's back up. What you're asking is, is what human beings did for 99.99% of human history, which is not—

they didn't run 100Ks for—

hold on, hold on, hold on, hold on. You're— what we're talking about is using your body correctly. So what you're actually— where you're stepping now is into what I call the reverse naturalistic fallacy. It's like, well, we didn't used to do this. That doesn't mean we couldn't, just means that we didn't. So there was no reason, necessarily a reason to. We don't have any evidence that people weren't doing it. But if you don't— but if you're living somewhere where you need to run 100K to get food, you're gonna move. So let's leave that alone for a moment, but I will answer your question. But I just wanna highlight the whole barefoot shoe thing is not even about shoes. It's about not natural human functioning. It's about letting your feet do their job properly so the rest of your body can do its job properly. What does that mean? Mostly it means don't overstride. Don't land with your foot too far in front of your body. Don't land with your ankle too far in front of your knee or in front of your knee at all. Land midfoot or forefoot so you're activating the Windlass mechanism properly and you're using your muscles, ligaments, and tendons as natural springs and shock absorbers. That's the basic premise. So, the research could not be— every bit of research that's coming out is getting more and more clear that there's value in letting your body do what it's made to do. Not elevating your heel and messing with your posture, not squeezing your toes and making it so your feet can't— you don't do push-ups with your fingers squeezed together. You do push-ups with your fingers naturally spread. Same thing with toes. It's not better to reduce the amount of stuff you feel under your foot so your brain doesn't know what you're stepping on or in. Now, to your point there, and I'll do both of these, there's 3 things that you brought up. You did so much in one simple question. There are times and places where doing, quote, what's natural is not the thing you need to do. When you're running 100K or 100 miles, you do whatever you need to do to get to the end of that race. End of story. If you're an Olympic athlete, you might have bad form, but it works for you to win that race. But, you know, maybe you'll be done at 30 and you won't be able to walk again after that. But who cares at that point? You've done what you needed to do. Race walkers, the way race walkers walk is really, really weird. It's the most effective way of walking based on those rules as quickly as possible. So what I say to people, like if you're running in a pair of Hoka, is if you're happy in them and it's working for you, awesome. But as soon as you're done, get off the trail, get off the whatever you're doing, put on a pair of our shoes. They let your feet bend, flex, and move and feel. That movement is going to help with circulation. That's good for active recovery. And there's research showing that just walking in our shoes builds foot strength and— or walking in shoes like ours builds foot strength. And building— there's research showing building foot strength reduces your risk of injury no matter what shoe you're wearing after when you're running. So run in the HOKAs, get out of your HOKAs, put on a pair of Xero Shoes, walk around, build some foot strength, build some better balance. That's going to make you more injury resistant when you put on your HOKAs yet again. So, in other words, there's always an opportunity where what we're doing, getting out of the way and letting your feet do what's natural, to be valuable. Now, to the research. More and more research is coming out every year where people are examining this because they're— excuse me— now getting the funding. For the first, Jesus, 12, maybe 13 years after the whole barefoot thing became popular or started to become popular, there was no one putting money into research because who's gonna spend the money on it? The big companies, the big shoe companies, they're not gonna spend the money because, frankly, the big shoe companies, they all know this is for real. They— look, Nike had a very robust barefoot shoe program that they just never launched because how are they gonna tell the story of barefoot and the story of the exact opposite at the same time? So 2 things about research. One, more coming out. For example, A year-long study done, sponsored by the Centers for Disease Control, took elderly people, put them in our shoes, and over the course of a year— because they did a very slow transition, much slower than necessary— their feet got stronger, their balance got better, their risk of tripping, falling, and dying was reduced. This is an $80 billion insurance problem right now, and who knows how— what the cost is to the families that have to deal with this. There's research showing that treating plantar fasciitis, one good treatment is walking around in shoes like ours. There's more and more research showing the benefits of letting your feet do what they're made to do, which can make them stronger, more resilient, more better feet, better response to feedback. Now, the research that's saying that we're full of shit is very interesting to me because again, we're not talking about footwear, we're talking about form. It's just that footwear changes or informs your form. If you wear a shoe with a big thick heel, the odds are good if you're a nat— if you're a typical runner, you're gonna overstride and land on your heel. Interestingly, even if you aren't overstriding and landing on your heel, it's getting in the way of something natural. If you have that big thick heel, and I see this, we have a lot of really good runners in my neighborhood, you're not letting your Achilles stretch all the way. It's not coming all the way to the ground or as far down to the ground as it could. So you're not getting this literally free elastic energy return from using your Achilles fully because you've gotten in the way of it. But, again, the biggest thing is this— we're talking about form. In the research that is, quote, anti-barefoot, without exception so far, I, I might be wrong, but I have never seen one study where they looked at gait, where they were looking at form in a barefoot situation versus a non-barefoot situation. I did this in the lab with Dr. Bill Sands, who used to work for the US Olympic Committee. He had a lab out here in Colorado, and we saw that almost, almost everybody, if they switch from whatever their fa— well, what he would do is he'd have you come into his lab with every shoe that you ever used, and he showed that every shoe that they wore changed their gait in some way, and they didn't know it. And, um, now the only difference is we saw is like sprinters and middle-distance runners, like 400, maybe 800-meter runners, where they were so dialed in you could put bricks on their feet and nothing really changed. But I even saw people who were, quote, accomplished barefoot runners who actually spent most of their time running in a pair of five fingers. And when they were in the five— when they were barefoot, their form was great. And when they were in the five fingers, they suddenly started overstriding and again didn't know it. So the gait problem is thing number one. That they're not analyzing the thing that's actually what those of us in my side of the world think is the important part. But the other part is this: the study design is usually we took some runners and we let them get acclimated to running barefoot by training either barefoot or in whatever shoe that they were in, which is not necessarily a barefoot shoe. I'll come back to that one. But we had them acclimate to barefoot by spending 5 minutes on a treadmill. That ain't going to do it. Different people take different amounts of time to get comfortable with this new gait pattern from running barefoot. And again, there's a whole bunch of footwear that is sold as, quote, barefoot, that isn't. Too much cushioning, some heel lift, some stiffness, some thickness, you know, various things that make it from what Dr. Irene Davis calls either minimalist, which is truly barefoot, or partial minimalist, which she says is the worst of all possible worlds because usually it's extra cushioning, so you're not getting the feedback that you would need to make that natural gait change, and it's not giving you enough cushioning to handle you're overstriding and landing on your heel. So that was my long answer to your seemingly simple question.

I appreciate it. Um, took a lot of mental notes here. Stephen, thanks for, for being on the pod again. I love, uh, talking to you. I love, uh, the breadth of your answers, right? Like, it's good, like, uh, to, to think deeply about these things and to, to brainstorm, I would say, these ideas. Um, But yeah, I'm definitely a proponent of barefoot running. By the way, I'm probably going to go crunch a long run with my Vibram. I think it's still the best training to adapt the feet even if you're going to run in more cushy shoes, it's going to strengthen them. But yeah, where can people find out more about you, Steven?

Not surprisingly, XeroShoes.com, XeroShoes.com, or if you're in the EU, XeroShoes.eu, or on social media. @zero-shoes or slash zero-shoes, wherever you happen to, at or slash.

That was, uh, Stephen and Charles for