AI is turning sketches into real-world products

This episode is brought to you by Accenture. When your advertising operations fall out of sync, everything else follows. Spotify and Accenture are working together to reinvent the rhythm of ad sales, using automation, analytics, and smarter workflows to simplify campaign delivery and access better data across the business. The result? Less time spent on operations, more time connecting brands with the moments and fandoms that matter most. Learn more at Accenture.com/splotify. This episode is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome? That's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50-page restoration block, or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it. Ready to make anything online makes sense? There's no place like Chrome. Check responses set up required compatibility and availability varies 18 plus. There's like build a bear for bears, like we can be the like build in anything store. Like I want to be able to like go somewhere and just build like almost like, what's that thing from Star Trek? Captain Picard says, "To your great heart." That beverage has not been programmed into the replication system. Sketch, sketch again, mold it, reshape it, trash it, and then start all over. This is how product design used to work for cars, shoes, clothes, and more. But one company is changing all that. Viscom is a new design tool for designing the real world, turning dream sketches into three concepts in minutes with AI, and companies like New Balance, Ford, Canon, and more are already using it. This week on Access, you'll hear the full story from founder Jordan Taylor's first jobs at Honda and Nvidia to building a tool for designing the future of, well, everything. Let's get into it. It's a funny time to be chatting with someone at the center of both hardware and software design. We have to talk about the Ferrari Luce, and we need to get your thoughts ASAP. What do you think? Oh wow. First, how do you say Luce? Luce? Luce? Luce? Luce? Luce? Luce? I think it's Luce. Like a talent. Yeah. Like an Italian, like kind of. For light? Yeah. Yeah. No, you know, it's interesting. I actually like it as a product, right? Like, now, is it a Ferrari? That came off like a serious burn. [laughter] Okay. Like, the brand, I don't know, like Ferrari is the thing to represent that as a product, maybe not, but like, it's standalone, no badges, like, I'm like, oh, this is great. This is like a great, like, product. I think just when you put the, like, Ferrari brand on it, it just sets such a, like, expectation or, like, preconceived notion that you're kind of like, it confuses you. I don't know. It's like an illusion, but I like it. I'm a fan. I'm actually a fan. I love the interior the most. I love the interior as well. I got to see the interior in person with Johnny actually a few months ago, and it was crazy because it was the first time that a car had been unveiled like this, where they showed the interior to everyone before the exterior, and then they had another event before this that I didn't go to where they just talked about the drive train, or, I don't even know what you call that with an EV in the software, so they were staging it out. And when we saw the interior, it was like, I don't know, we talked about this on a past episode. It was exploded out, like, iPhone parts on a table, right? And you could, like, sit in, you couldn't sit in it, but you could see part of the cockpit. So it was really strange. They were showing, like, part of it, but no one could really infer what the thing looked like. And I guess it's because they knew that the exterior was going to be so controversial. I don't know. It was a very unique experience, seeing the interior of the car. That's all. Yeah, I saw the, I know exactly what you're talking about. Like it's kind of like almost like floating, like an exploded view. Yeah, I saw images of that. Yeah. No, it's, it was worth celebrating. So it was, it was really cool. Well, I know you've worked it. Honda in the past. You've worked in car design. Now you make an app for designing cars and other products. Do you think they'll process of shipping? Something as controversial as that is going to change. Or at the end of the day, it still just comes back to getting executive sign off on something crazy. Yeah. Like, do you think it, do I think it's going to change? Like they're going to go back and like redo it or something like that or. Yeah. Yeah. I don't know. They're not redoing it all. Come on. Yeah. You know, like $650,000 car. It's, it's an expensive change and it's not like the sonic movie where you can just like go back. Remember when the sonic movie first came out? Oh, yeah. I don't know where like reaction and now I think they're just going to roll at the punches. It's like, it's controversial now. But I think it'll age very well. I think maybe like our grandkids might be like, oh my god, you were there in the loop. Like, whoa. I don't know. Who knows? Why do you think it will age well because it looks like like, yeah, I don't mean to insult your former employer, but like a, like a Honda concept car meets a Tony Stark Audi or something like that. This doesn't look like Ferrari at all. I know. I know. I guess it kind of depends on what the catalog of products will be, but to kind of like support it. But like, is it just like a one off thing or is it kind of like a grander, like, are we going to see an SUV follow up and look like this, right? Like, that definitely matters a lot, but it definitely has this like staple and like the tail of the times. I feel like the cyber truck had a similar like, oh, wow, that's more of like a tail of the times as much as it is like a product that's meant to make money in some way. It was like more of this like expression. I don't know. I just think it's like one of those things will like be able to like point to and be like, oh, yeah. That was that moment in time for this brand and like, it's iconic because of that more. So then like, the way it looks and then things like that, yeah. I don't know. It's cool. I think the story around it is just interesting. So one more question before we take a step back and talk about the origins of vis-com. I know every AI model has its preferences when you start riffing on ideas about a car with vis-com. Is it pro tactile buttons or pro screens and software buttons? All that. Oh, man. Oh, man. I mean, it really comes down to the intention, of course, but like, I'm someone personally when it comes to cars, I do like the buttons, not going to lie. Like, there's just something about when you're driving and like, I'm just like trying to change like one thing, one thing. I got to like go through the menus and stuff. It gets really hard. Now, the argument is a wolf. The car is driving itself. If you don't need to, but I don't know. I just think there's something about it responding to you in some way that feels tactile. Like tactile. Like tactile. Yeah, yeah. There we go. Luchin. That I don't know. It doesn't matter. Sactile. Yeah. Yeah. Tactile. I don't know. But essentially, yeah. I think it, yeah, it comes down to the intentions, but I'm pro buttons. We have enough screens. Yeah, we got this thing. Yeah. It was interesting. I think it was with Cleo Abrams. Johnny, I was giving an interview with the, for the final unveil of the exterior and talking about multi-touch. And he said something when I was there for the junket, for the interior as well about how multi-touch is just not designed for a car. And I mean, he helped work on multi-touch and invent multi-touch. So he knows what he speaks of, but the idea that you're interacting with interface and touching a screen and this kind of flat UI that you really have to visually look at does not make sense for a car at all. In fact, it's quite dangerous. And so I guess if you were going to approach an interface and a car from first principles, you would go pretty analog, wouldn't you? Yeah. Yeah, yeah. It's all about, you know, like foundationally multi-touch is not a bad idea. I think it's just like the context in which you're putting that is matters a lot, right? Yeah. It's a bad kid, role-blocks, daycare, like, okay, like yes, that makes sense. But like, okay, I'm trying to like get home safely, like, yeah, like it's just the context of that is just a little bit odd in the situation. But yeah, it's, I think they, they found a really good, oh, I love how you can like rotate the screen by the like handlebar. It's like, it's like, yeah, it's so cool. I think that, I think. Another thing Johnny was showing us was he was obsessed with events. The events are aluminum and they have this clicking noise. It's very satisfying magnetic, almost clicking noise when they open and close and he would just like, he kept like opening and closing and never brought us as the most, uh, Johnny I think ever. Of course. The subtext is that tactile things are more satisfying. And I think that's one of the fun things about doing what you do, Jordan, making software for the real world, as you say, is that whether it is the ridges on a clog that you're making like a celeb, benberry, you know, clog or whatever it is or the knobs in a car, it's like human beings enjoy touching and twisting buttons and pressing things and turning things and doing their turn signal or whatever. And I think in some senses, you know, Tesla talks about, uh, how much simpler it is, but I feel like that's just the disguise for how much cheaper it is to not have any of the physical buttons. Yeah. That's how they made the car so cheap. Yeah, I know sometimes you can lean into the, the consequences is like a feature, but, uh, I don't know. There's just something. If you go back to when we're like kids and babies, like, babies for some reason they learn, you learn through touch. Like, I, I, I tend to like, Montessori School for a bit when I was a kid and like, we learned so much through play, which was just like a fascinating thing. I don't know. There's something very innately human about like physical interactions, especially in today's time, which we can, I could talk about forever, um, but yeah, it's an interesting thread to, to go down. Well, switching gears, take us back a little bit. How did this come come to be? I know there was a viral Reddit post. I know you built an app. Instagram account that had tons of followers for concept design, take us back to where it all started. - Sure, sure, sure, sure. So let's see. So yeah, so my background's in car design, went to school, art school. And at this time, you know, there's art center in Pasadena, kind of like our URL is. We have that in the Midwest called CCS. And there's like these design schools all around the states. And I think Behance was kind of the only place where you could see what other designers were doing. We're always curious like, oh, what are other students doing? So I created this community called Designers Pen on Instagram, purely around car design. And that's kind of how I originally started, I guess like this call me. I came at it from the community first approach. So fast forward to like 2020, 2019 or so. I started working at Nvidia at the time, like the state of the art models. This is like super dating me. But like Bert was like this language model back then. That was like state of the art. It could like tell you the price of a burger and it was called style gan as well in the image side, right? So just these early inklings of just seeing like what computers can do in this untraditional way. I was like, wow, like, so you're telling me this can generate a human face of someone that like doesn't exist. It was really bad. The quality was like literally like a Nintendo game way advanced kind of situation. But I was like, oh, wow, like this. I had a pretty Will Smith eating spaghetti. Oh, yeah, yeah, yeah, this is pretty like even. The idea of generating a video was still like very sci-fi. It was like, oh, that's 20 years out or something. But you know, that little inkling of just seeing it do what it could at the time. I was like, oh, I gave me enough conviction, but it also was like the signal of, oh, this is the worst this will ever be. So fast forward COVID happened that I was like, okay, well, now we're all working from home and I have some time to kind of do my own thing for a bit. So that's when I made that Reddit post. I trained a cycle gan, which was this convolutional neural net on my own artwork from school to see if it could generate a car design of sorts. And it could, and it was really bad. And it was meant to be something that was a starting point for designers. I never meant for it to be something to like generate a perfect car. I was like, oh, this is going to be something where people draw on top of to further develop. So I made a Reddit post about this. It was a really bad react app. But then I woke up and I was like, oh, well, like a lot of people actually resonate with this. Like let me just keep developing it. And then my co-founder who I've known since fifth grader, so it was just kind of crazy. And he was then missing it. And I was like, you old dude, so I got this extra room and mountain view, like all my roommates went back home because of COVID, like you should come and move in with me. And we'll work on this thing. Like I don't have a name for it yet, but we'll like, we'll work on it. It kind of like renders your drawings for it. And he was like, oh, it's kind of cool. So Caitlin drove out. Then we just kind of like granted that out for, we had, so I left and then living off my savings for quite some time. And we lived by a Costco. So shout out to Costco, man, those hot dogs came in clutch because they're like $1.50, you get one for one. - That's the peak, is that not peak product, is that it? - Oh, definitely. When it comes to pricing and packaging, like the Costco hot dog is like, that's, we should aspire to get there. - Yeah, so that's kind of the short end of the story, but that's kind of how the originally came off the ground. It was just like having the foresight, but then also the insight into how designers work, which was myself. I was just like building this for myself. And then I wasn't like trying to like build a company. I wasn't like, you like, I gotta be Mark Zuckerberg, like I gotta build a company, like that will did my intention. - But you didn't have the house in the South Bay. - Yeah, I did. - You were zucking it a little bit. - Slightly, slightly, I wasn't mountain view. But like I was like, I wasn't like my intention at all. I just kind of like, having an accident. But yeah, that's a little bit about how it started and the problem space, but. - Yeah, shout out Costco, the only product, the Costco hot dog that resists inflation, I think. - Yeah. - Yeah, that is a unique aspect of that that I always enjoy about Costco. - Yes, and the rotisserie chicken as well, which is a money in the pizza. And the pizza, all of it. - Glizzies, the perfect shape, the perfect product design and tactility, warmth, all the things humans need. - So how many designers are using this account now? - Let's see. - And do their bosses know? - So definitely in the hundreds of thousands, I don't know the exact number, their bosses do know. And it's because it's very much like a bottoms up thing where they're usually bringing it to them versus the other way around. Because we put a lot of emphasis on students, especially. And I think when students graduate or get internships, they usually bring this come with them as well. So it's, I would say, you know, even when we first started talking LSD, the zikie, sort of the sentiment around AI, I would say AI has had probably one of the worst PR things in the last, like either eight or 12 months. I don't know what change, it's kind of always had a rough thing, but I don't know. - And what changes they said, all your jobs are gonna go away. That's what changed. - Yeah, like when that turns out, people don't like that very much. - It went from, we're gonna make a robot move and we're gonna make Will Spitz spaghetti to your jobs are going away. That's what happened. - Yeah, I know. And it's played in our favor in the sense of being able to stick out even more in that way. But yeah, it's fast. You know, I think one thing about Vizcom, even from the get-go, I was always obsessed with this idea of the drawing. Like the drawing is a starting point. And there's something about a drawing that is just so innately irreplaceable. It's similar to like language, right? It's like we're always gonna be talking or communicating to each other verbally through audio in some way. But I think a drawing way, and that's where the name comes from, like visual communication, like Vizcom, like. So in a world where we're leaning on the human capabilities and in a world where it's like automation versus acceleration. I'm very much in this camp of like accelerating how designers already work and how they already think versus like, you know, we're gonna create this like overlord thing that's trying to like get rid of all of it. Very early. I think that's not the right approach in this space, but that's how I think about it, personally. - You say that like in like a sarcastic way. Like, yo, we're gonna create this overlord, but like people are literally saying that these days. Like it's not even a joke. Like what was it yesterday? Alex, like the Anthropic people were saying-- - Oh, yeah. I mean, your newsletter, they're like, we're building God. - Yeah, I mean Bill Gurley has this great thing. He calls the Dr. Frankenstein Theory of Anthropic, which is that they actually believe they're building this crazy all-powerful deity. And I actually do think many people there believe that. I don't think they don't view themselves as like a normal tech company. A lot of the AI researchers do believe this. And so then it's interesting to use those models to build a product like what you're building. And I was gonna ask this earlier, but it's I think even more interesting now. Like I'm curious how you use the models. And when you started, were you using a specific model? Like was there an insight with a model that you were specifically using that you said, "Oh, I need to build a product layer around this." And then how are you using the models now? - Totally, totally. So back then, the first model we were using was called StyleGand2. And this came from Nvidia. And at this time, there was no idea of like, what's called like text conditioning. So you couldn't like use text to like create something. You just, it was domain-specific. You give it images and you just press a button and you get something randomly. Like there was no control per se. - What was this 2023? - This is 2021. 2021, I can see you guys an old demo video of like what our first pitch looked like. But yeah, 2021, and then fast forward a few years, stable diffuser came out, right? So I don't know if you guys remember Stability AI. That was like a huge thing. And then once the fusion models became kind of the new standard, we just moved over to that because I was like, wait, so you're telling me, not only can I give it a drawing, I could also give it like text description to help guide that drawing in some way. So that was like a massive unlock. And then that's when controls started becoming super important 'cause it's like, okay, we're not going from this random shotgun approach thing to this more, hey, my intentions can really come through. And then kind of fast forwarding till today, it's like now we have these like consistency models that can like really understand what you're trying to do. Like that, that intention translation is no longer a problem as much as it was before. So this is kind of now a collection of like, you know, a lot of open source models, but like some of the big models, of course, like Claw to help us like translate, okay, like what's the intention from the user? How does the model best understand that? And then some of the off-the-shelf image models, you know, you got like the nano bananas of the world, the twins, things of bad nature, to then kind of help, you know, actually distill those images. So yeah, that's kind of how it works today. We're more so focusing on the, what are the capabilities these models unlock, rather than kind of like, okay, let's start from the tech first as far as like, how do we invent something brand new thing? Let's kind of focus on the problems and then really understand what are the user pain points and then kind of ask ourselves, okay, like how does the model fit in that situation? (gentle music) - This episode is brought to you by Accenture. The result? Learn more at Accenture.com/Spotify. - This episode is brought to you by Google Chrome. That's new. It can help you with practically anything on the web Like restoring a vintage motorcycle from a 50-page restoration block or finally break down that long article You've had open for weeks. Check responses set up require compatibility and availability various 18+ So I remember back when we worked together. I don't know. I was at a couple years ago at this point You said that a sketch is better than a text prompt for AI, or maybe it has more character and more Elements of your individuality. Do you still believe that? Why is that? Yeah, I Just think There's this like I don't know you guys are working out. You've heard something someone say like all of the muscle-to-mind connection or like When you're working out and you're actually thinking about it, you can feel it almost more and you're like Where's when you're kind of like autonomously like I'm just here to like get it done? Like there's just something about like the mind and like your hands and the intention you can get across with that like Connection versus trying to describe like like right now if I try to describe a green cube in Like this like igloo room like I could probably get something but in my head I have a very clear and I have the ability to sketch like what does a green cube look like in an egg Glue exactly how I drink Taylor wanted I think as far as capturing a creative intention a drawing is just like Maybe not the end all be all, but it's like a piece to the puzzle that is necessary in order to really capture that So I think we're just inherently very visual creatures too, and so I just think me personally is well-hatched slow I love drawing and it's just I can get a lot more across there than try to communicate But what if you're like me and you can't draw oh Man, I mean if you can't draw. I mean that's there's definitely ceiling. I think there's still capabilities You can do purely with text but still like even just simple annotation I'm not saying you got to be like Michelangelo here, but even just like the idea of like oh wait It's like what did you mean excited? I was like no no this guy over here was supposed to go over here and then do this thing Like I don't know if you guys remember it's kind of like spilling some beans of what we're doing But like when you were in we're in middle school high school your teacher would do like the English teacher would put like Three lines under your the first letter that meant like oh, that's supposed to be capitalized or they would do that like P with the double lines It's like I meant it's a new paragraph and you like she didn't have to tell you that but like that little Hyroglyph was enough to kind of communicate a whole thing, right? So I think there's maybe net new ways we can invent new visual ways to communicate, but Yeah, I don't know. I just think it sits very interesting. Yeah speaking of visual communication I mean, we used to say this at snapchat all the time that like our our friend Jacob and Drew Alex who Where is he now? He was VC and now he's the head of clippy He's the head of copilot. Yeah, so one of the things he once said when he was a young man at snapchat He said humans are IO constrained and I was like what the hell are you talking about? And he's like well you have to translate all these ideas and feelings in your head into words and that's texting and no wonder There's so many miscommunications and whatnot thus the idea of snapchat where you send pictures and videos It's just more natural to the human being But you also think well the words give the AI is so much to work with in terms of structure that made LLM's work. How does that actually work with the AI's and like your doodle? Yeah, is it a similar mechanism or it's it's very similar. It's very similar But it's like it's a combination of the two. So there's like this idea. I'm gonna kind of Trust and fly this so it's like there's noise and then there's like a text description and then it kind of like Basically is trying to like distill that noise and your input as like a way to kind of like Generate something that's net new to some degree, but guided by the text you can almost think of it's like guardrails Like you're introducing more guardrails like Texas like you know You've probably heard of like context like a million times like in the last month But that's and that's one way of setting context I think a sketch or an image just like another way of giving context to the model to like reason through It's like similar to us, right? It's like if we have the right context like I don't know There's something also about like memes. I love memes and the reason I think they're so genius as well And I think you're almost needs to be like an economic class around like Memology or something like this, but like did you see the Nicholas Cage the Nicholas Cage interview where he explains Memes just thing we call memes. Mememification. I coin the word. Memification. What is a meme really? I mean, what is that because the memes were not unlike the dreams anyway, sorry He definitely he had an early like 2010s like meme run like I don't know if you have to go those old formats Or it's like the baby with the fist and like the old like we we've evolved so far since then, but I think it's a visual language Right exactly exactly like that if I send you Arthur's angry hand That communicates almost just as much as if I were to like give you a huge paragraph of like how I felt about a situation as we also used to say and all of our snapchat decks A snap is worth a thousand words. Yes, and the other guys just go, oh Yeah, no snapchat was so ahead of its time like I haven't used snapchat in a long time, but Yes, I'm certainly super unique and it's in its own way. There is a problem with the stock price Jordan, which you just said Yeah, so uh, I didn't realize until we just been chatting, but I've actually been vis coming With my daughter I recently finally got an iPad for our access live show a couple weeks ago And I got an Apple pencil pro because I can expense it to myself and we've been doodling in the Apple Notes app And of course my five-year-old discovered that there's actually an image gen built into the Notes app now And if you draw like a little plant and you do this rainbow Like pen in Apple Notes circle the plant that you drew. Let's let's say it's a flower and then you write flower into the tax and it generates you a flower Uh And that's kind of it, isn't it um with that said I do not really have much faith that Apple is actually using the doodle almost at all Yeah, I love Apple Notes. I always default back to Apple Notes in a lot of ways like I was like a heavy notion And like I just went back to Apple Notes. I don't know like I love Apple Notes, but that's essentially it like the experience we're kind of creating you're describing there. That's one aspect of it Absolutely loves it. Yeah, um, I mean to have you had yeah Apple Notes is also beautiful for just how dumb it is and how simple it stays despite how much software is getting more than that Yeah, oh my gosh, dude the new foot we're gonna a little bit off top. The new photo gallery is terrible Like is it just me? Oh, yeah, I don't know how you guys feel about this, but like I can't I It's it's I can't I hate it. I don't like I'm like where did this photo just go that I saved or took like I don't know I don't know. Maybe I'm just I'm entering what the kids call on status now So So Jordan we've been talking about the inspiration for viscom a little bit about how it works And I get that I'm curious though How is it being used? What is being made with this? Can you point us to concrete? Oh I'm setting up your quote-unquote concrete examples of things that That viscom is made. Yes, wait, okay Can you guys give me like to I'm gonna go I can graphically really it's like right off either door. Yeah, go get it This is something. Okay, I really want to show this off. This is really cool You're breaking here. He's got a pretty good bookshelf game there You could fill in the middle but Okay, so before I show you the objects so So the the flow I guess are kind of the spatially focused on so going from paper to production, right? That's that's kind of the the area of focus This is what industrial designers do footwear automotive all these people they sketch they render Create 3d models and then they either 3d print mill create prototypes and then mechanical engineering makes it real So where does this add the most word? Where is this working? You'll lock the most value here. So footwear companies Let's say someone like a new balance right or a Nike they they're using our tools to help Visualize and physically understand maybe how their early concepts might look like much earlier in the process So you can think of it's like these guys up until this point They never had like a debugger like a programmer can like write something and be like oh, yeah, that doesn't work That doesn't work. There's so the ability to fail faster here is kind of a new concept, but just crazy So so in the shoe world right we we worked with these guys called so zeller failed right? I don't know if you guys they make this is a shoe. This is a 3d printed shoe Which is a crazy design of these like it's this idea of like two hands coming together to kind of alien meets a easy Yeah, it exactly So these are super neat and this is like so zeller fell down in Austin, Texas They have these custom 3d printers for shoes and exact and it's it's actually really soft and like you know You can actually like bend these and stuff It's almost like a crock material and then these those are super cool too So these are called the catapills these are like these caterpillar 3d printed shoes So these started out as a drawing in this calm and then actually became something you can hold And this is like you know like the whole like make it real story, right? It's like not only producing things that stay on the screen, but like stuff you can actually physically hold So that's kind of totally be seen wearing those around those filous on a weekend So I actually got the celeb bembery crocs And. (laughs) My wife said that they looked like human flesh, and they were also kind of like a pinkish color, and so I had to return them. And I was up, I was up, clicky, clicky, clicky, waiting for the drop, trying to secure it. - She made you return it, I mean, they did look, they did look ridiculous. - This is very unsettling the whole by the way. Like it's quite literally hands-on feet situation here. - Yeah, yeah, it's really cool. - It's like alien on a easy. - So did someone just sketch that shoe, put it in Vizcom, and then AI made the 3D model that was then printed? Is that what happened? - So they sketched it, they used some models to visualize it first, and then they kind of iterated. So we've created a lot of very novel interfaces to kind of help you explore forms, color materials, height with things of that nature, and then we're able to then take that, and then generate a 3D model from this. This is like a mesh that gets exported into like an STL file. And then that gets put to the actual printer. There's some processes in between, like something called a slicer that kind of helps you set it up for the printer, et cetera. But yeah, a lot of these are straight up generated 3D models that then got prepped for manufacturing and just printed. - And for people who don't understand the complexity of this, like me, how long would that process have taken Vizcom? Like would someone have to really draw every part of the shoe? What has changed in the process? - The rate at which you can fail. So before one even understanding and knowing this is what you would have wanted to do, would have just taken much longer. So before you're spending Photoshop, like time and Photoshop, quite literally coloring and kind of visualizing like a flat 2D line drawing. And then you're going into like a rhino, a blender, moto, like these 3D, like Maya, for example, to then try to understand that. First of all, then you got a first, okay, first I'm going to sketch this from multiple views because you don't have multiple views at all. You just have a side view, maybe. So now you got to like, I got to draw that out. And then you got to build that manually and first interpret it first in 3D because there might be some loss in translation here. So that's like another thing, right? The designers' intention and clarity of his thoughts now are able to stay way more intact versus like before all of this. So you see the same thing happening in software, right? Now the designers can kind of like push PRs. You're like, oh, there's no more of this like pixel correction back and forth situation. The same thing is happening in the physical world. Like the designer is now able to kind of really carry his like torch, I would say, much further in the process of going to production versus before it was kind of like a multi handoff process of like everything I just mentioned before required maybe seven different professions. Now it's like these are all starting to converge into like one thing similar to this like, this PM versus engineer versus designer conversation you see happening in software. It's like that same thing is starting to happen in like physical product creation. So. And there's an argument to be made that yes, that's kind of taking jobs away. You described a bunch of jobs that are getting kind of pulled into one. But at the same time, I think what you're building shows how AI can help humans be more human, do things more efficiently. It's an interesting counter narrative to a lot of the doom and gloom out there about AI just wholesale replacing people. And maybe you didn't have a strong key of you on this when you started 'cause you started like pre-all this, you know, narrative that we were talking about earlier. But I'd be curious to hear how you're hearing, I'd be curious to hear the reception you're getting from people in the design world to this. And I know there's a lot of fear about AI in every creative field. And I'm sure there is in this design in the design world that you're in as well. - Yeah, I hesitate to say the words like taking away. I almost feel like it's evolving jobs, right? Like a videographer existed 50 years ago but they look very, if you showed him a videographer today you're like, "Way, is that even my?" What do you mean that's my job? That's that looks very different from what I'm doing today. So I think just the makeup of a lot of these jobs are just gonna change fundamentally, right? Like this exact situation here would have been a news broadcaster 20 or 50 years ago. Now it's like, "No, we're not podcasting." It's like the same thing, just maybe a slightly different title. Now we have four deployed engineers, but I think that was maybe like consulting back in the day possibly, so I really do think it's an evolution. And so to kind of go back to your question. So I think from the beginning, we were always in this camp of authorship. Like how do we instill authorship? And you kind of lose this sense of authorship when you lean too much on these models doing things for you. And you kind of feel this with Claude and like LLM's a little bit where you're like, "Yes, it can do these long running agentic tasks." But I'm like, "I feel like I need it to like intro." Like I don't know, like I feel like I lost my sense of authorship slightly. And I don't know if that's like a nately human. I'm still trying to figure this out. I don't have the answer. But I don't know if it's us trying to seek like a sense of purpose still or if authorship is something that's just something that we truly just need as people. But that's something we've stuck to from the very beginning. I did not know at the start though that the kind of public zeitgeist would evolve into what it is today and have such a negative. I saw some like Google exact dude giving like a college talk thing. And like he's like, "Guys, like graduate and usually I." And then I was like, "Boo, tomato tomato." - Oh yeah. This is Eric Schmidt. - Some Google exact guy. Eric Schmidt. - Yeah, Eric Schmidt. I was like, dude, I was like, "Oh man, this is bad." Like, he was a no, he was a no, yeah. - Well, so what do you think changes in modern shoe design as a result of tools like this? Am I going to be downloading models and 3D printing my own shoes in a few years or do things get more experimental? Do more people become designers who may not have been? - Yeah, yeah, yeah. So the last two where, you know, I don't think everyone will, you know, we had a moment we all had laser jet printers in our house, so I don't think I have one still 'cause I like printing stuff and collages, but I feel like that's not as crazy as it once was, but so I don't think that will happen, but I do think it's gonna push the envelope of what is actually possible. It's like, like these kind of ideas I'll just show you guys, it's like, well, like, okay, like this is almost a reflection of the tool just as much as it is the person themselves, like the reason cars in like the 70s and 80s looked so boxy wasn't because no one couldn't think of curved cars. It was more so a constraint driven by the tools, but the moment alias came out that allowed for a curvature in 3D to be a lot more accessible, then we started getting these more like jelly bean, like early 2000s, PT cruiser kind of like looking cars and stuff, and like we didn't just all the sudden get more creative, it was like, oh, the tools enabled that. So I think with generative AI, it's like, man, we're gonna see some really funky products potentially. Like the next IKEA catalog in like 2030 is gonna look very fascinating, I believe. Alex, did you know that the car I drove around in college was a silver PT cruiser with flames on the set? 'Cause that on-brand or is that an aberration? - That's incredible. - You said flames on the side? - Yeah, they were flame magnet decals that my grandma got me and I just like, I just like tore around in Arbor, Michigan, slipping and sliding all over the ice in my front wheel drive PT cruiser, yeah. - The hamburger, I love it. - Yeah. So Jordan, tell us a story of the team member who built a concrete keyboard from scratch. - The concrete keyboard, man, I wish I can go grab that. It's upstairs, but yeah, so Amel Lucas on our team, so it was kind of a collab between me and him. I have like a little doodle and to like quick this commenter, I was like, yo, like, we have this concrete keyboard. That'll be so funny. And he's like, wait, like, wait, I think we can actually do that. Like I actually have cement that I could pour and like try to do this. And I was like, oh wow, okay, cool. So he took it all the way. He generated 3D mold for it. And then we sliced it in half poured some cement in there and then built like a functioning keyboard. It was kind of wild to see. And like that's an example of there where it's like, I would have never thought of or even wanted to put the effort into visualizing that initial cement keyboard to want to then be like, okay, let's make this real now. Like, so getting to that conviction much quicker. But yeah, that was a wild situation. And then holding it in your hands is like, it's really satisfying to use. I actually use it like pretty much every day on my desk. It's like a, it's really chunky, but it's cool. I love it. - And people say mechanical keyboard bros are annoying. This is a whole new level. But I mean, this is beautiful though. I mean, it's just like, you know, having a kid, I mean, it's one thing if your kid is gonna vibe code in app. But it's another if they're gonna like draw something and then make it somehow, whether it's with the 3D printer or otherwise. I mean, I know some dads who play with 3D printer stuff all the time and that, that, that seems like a net good for humanity in a very obvious way. You know what I'm saying? And yeah, I mean, how does it feel to be working in the hardware world versus so much of tech that's making like digital stuff, you know? - Yeah. It's interesting where we're definitely at the intersection. Like a lot of the stuff we do is so very software as far as what we're providing. But to then see it manifested physically is like, really cool. Like, I just think in general, like things that are super high fidelity and digital and stay digital in our screens, the signal of effort is no longer there. Like, it's like, you seem like it's super cool on the internet. And now you're like, okay, like, that's cool. Like, it doesn't have that same like, like 10 years ago, you see like a really nice rendering. You're like, whoa, like, first of all, how did he do that? - That person knows their shit. - Yeah. - Yeah, spend nights on this. - Exactly. So I think the next signal is like, okay, than how we're doing. did he make that thing? That's like, okay, the physical becomes the new signal of effort in my opinion. So just kind of really leaning in in that and being a part of it is super cool. And seeing even these shoes, it's like, okay, well, this is like, there's something here. We're still in the early innings of this in my opinion, but it's cool to be a part of the early signs of it being real. Yeah, I'm curious where you see this going. When you've raised a decent amount of money and the business is growing, you said, you've got hundreds of thousands of designers on the platform. Are you coming for all of these legacy creative tools, the Adobe stuff, maybe even something like Figma? Where do you see this ending up with what you're building with SCOM? Yeah, I think it's really focusing on that paper to production pipeline. So understanding to the best of our abilities, how physical products are made in the real world, and then just keep going down that value chain. Now, I always jokingly say internally, like, hey guys, if there's like build a bear for bears, like we can be the like, build an anything store. The, um, uh, that like, uh, fabricator thing. Yeah, yeah, yeah, yeah, exactly. Like, Captain Picard says Earl Gray, hot. Yes, exactly. Like if we can do like, I forgot this is someone will, someone will comment it. But like, like, if we can do that, like that experience of like, can I really hold my ideas? Like now, what does that look like as far as farm factor? Do we get into the world of hardware? Like do we have to start inventing net new ways to like build things because of what's happening upstream? I don't know. But I do see us being the ones to unlock the kind of like instantaneous, like, you know, I upload a drawing and next thing you know, it's in the mail about the other day or something like that, right? Like, like some kind of like physical thing. Do you think that process gets to a point where someone like me who doesn't really understand this workflow at all, especially how it used to be, could create a physical product in the same way that I can now code with cloud code work or code acts without knowing that I'm coding. Totally. Do you think that eventually I'll be able to make something physical with this? Yes. Totally. Like, and I kind of, I try to stay away from the word easy and more so unaccessible because easy to me almost like degrades the, the craft of something or like, it like, it makes something seem like it's less valuable. Like, it'll be easier, yes, to do these things. Now, will you be able to make something as good as like a designer from like new G or like T engineering? Maybe not, but like, you will be at least able to play the game that they're playing to some capacity, which is kind of what these like LLMs have done for like a wide spread of folks who are. I do intend. I am as good as Fouca Sava and the Moji products he's made, but I cannot yet prove it because I can't make stuff yet. Yes. And that's where the whole like taste conversation comes up, and which I'm like, and I'm still kind of like trying to understand like what that means sometimes, but like, yeah, that's that's where like your judgment and that kind of stuff comes in, which is, which is really hard, which is really hard. I am a believer in that. I mean, you know, I am a writer by trade, but as you saw when we were together, Jordan, it's like I have, I have the heart of a product designer. And I think of words as part of the product design of like helping people experience the entirety of a product, you know, and a lot of times those words can in turn inspire the products in mind. And people have all sorts of experimental ideas in their mind. And I feel like historically, it's just been gate-capped by whoever could could make that stuff. Like words are your like brush or medium to do something of sorts. Like I think your start to see is a little bit more where physical outputs of sorts are becoming like the medium of like whether it's marketing, like you'll see someone kind of like do like, I saw someone do a cassette tape release for an album, which is fascinating. I was like, oh, wow, that's interesting. Or what's the one space movie that came out recently? You could scan a QR code down with a 3D file as part of the promo and then a project tell Mary probably. Yeah, yeah. Project kill Mary. Yeah, yeah, I heard that movie. I haven't seen it yet. I heard it's great. But there's a lot of good movies coming out recently. But anyway, it's like, yeah, it's tough like that as far as like using physical things as like another form of expression. I think it's going to become a lot more accessible. Yeah. So I think one of the things that informs your taste in judgment, we don't talk to very many founders from Detroit where I'm also from how do you feel like your your background has informed how you how you build differently. Oh, man. And you are and you are in the valley. So yeah. Yeah. Yeah, the Midwest man. Oh gosh, I like, I didn't realize garage bands was like a Midwest thing. Like someone told me that it's probably a tweet or someone's like, Oh, they're trying to cosplay. We don't even have garages here. Exactly. We even have garages of say with basements and stuff, right? But I think like being from Michigan, one, you get all seasons, but the most prominent one is winter. And when it's winter time, like you do not really want to like go anywhere. Like, I don't know, like even the idea of going to warm up your cars like this, like, oh my gosh. So you inherently just spend a lot of time indoors. So that just naturally calls me to like play on computers, like wood shop stuff with my dad and the garage. Like, so I think just being in that environment, kind of ushirt those kind of things. And then also, I don't know, there's something about the Midwest where the like career thing, being your personality, kind of like force not that not being a thing almost like forces you to like just try a bunch of stuff. Like, I was into like magic that gathering. I was like, I tried playing guitar. I just like, I don't know, I tried a bunch of random stuff. Um, I think. Dude, why were we not hanging out? I was playing magic and going to the Detroit auto show every year. I'm sure we brush shoulders at that event. Yes, we probably crossed concepts like seminal shit for me, seeing all the concept cars back then. Oh, yeah. Like, that was the future. That was the future, like seeing the Detroit, it's sad what they've done with the now, but back then, like the auto show, like pre internet getting really like algorithmic social media stuff. It was like, whoa, like I'm here in like game and former magazines, Nintendo power. Like, I was really into this stuff. So like, I don't know, maybe it's just like the time period I was in the Midwest. I'd be curious what someone 10 years younger than me thinks. But all of these things, kind of like is what, I don't know, I played a part into how I've I've developed my mindset. Jordan, considering that you worked with Ellis at one point, I like to ask every guest who's done that and we could maybe end it here. Um, was Ellis actually helpful to you? What did he come up with? What were his shower thoughts? How did he translate into a meaningful venture dollars that you raised? Yeah, I was, I was actually, so I think Chris was the one that originally brought this up and I was relatively skeptical. I was like, well, what do you mean like words? I got words in my head, I can do this the refund on Chris. I was like, Chris, bro, we have GPT now, this is like three point something. I was like, Chris, we can figure this out. He's like, no, Jordan, like trust me, like I know you care about stories. Like this is something I can help you. I was like, okay, great. So yeah, it was, it was awesome. Like the whole like make it real statement. Like what do you see on this hat here? Um, that that really stemmed from that exercise, right? Ellis, I don't know if you remember that. Like we, we've went through a many different iterations and just kind of like really getting to the core ethos timeless nature of like what it is we're trying to do. Like in a world where the mission is now the mode, right? Like I don't think it's no longer like technologies. Well, the mission is the mode that's going to go on else's website. You got to be careful, Jordan. No, but like, that's what we say this. I say this a lot internally because I'm like guys, like the mode is not what we're building for say it's the reason why we're building it is more so the important aspect of it. So Ellis really helped at least us to establish that direction. And it's pretty timeless. Like even the make it real video we released almost two years ago now, you watch it today. You're like, oh yeah, yeah, we're still trying to do that. Like that is still holding up today. So yeah, that's that's that was a fun exercise. And yeah, cool for us. I'm imagining Ellis like smoking a big blunt with no shoes on. And he's sitting cross-legged with you, you know, in the park and set up and he's just like make it real. Yes, but what's exciting. Let's how like romantically you think about it, Alex. Like, I'm literally sitting right right at this very spot. In a cave with no natural light. I'm not sure what you could do out in the Detroit winner. I know, I know. All right, last but not least predictions for industrial design for tech. I know we talked about general monster before before you came on. We've got the new loom lamps that are kind of stainless steel robots that do your laundry. What do you think is next for for for product design and tech? Oh man. I don't know. Who's going to have the best glasses that people actually want to wear? I know it was just reported this week. Apples working on it as well. Yeah, yeah, yeah, yeah. I think opening eyes really are surprised people with their dime thing. Like I just think that's really going to shock folks. I haven't tried it yet, but I've heard many great things. I just think that there's something there potentially, but I love the Sunday robot. I - Actually, I love battery. - A lot. - Shout out former guest, Tony. Another else client. - Oh, no way, really. Oh wow, okay, that's awesome. I love there's something about just like the personable aspect of it. So, but, - Yeah, I don't know. - Metal has a hat, which makes all the difference, just like us. - Genuinely, there's something about the similarity. Like now, do robots need eyes? I don't know, like he really want to like see that, but yeah, I don't know. I think the OpenAI hardware device is really gonna like blow people's minds potentially. Like, I wouldn't underestimate that. - Ooh, that's expected to do. - Yeah. - Yeah, what do you know? (laughing) So, you know something. You wouldn't be saying that if you didn't know something. - No, I just think it's gonna be something very surprising. Like, whoa, like I didn't maybe think of things could be that way. Or like, similarly like when GBT drops, you're like, oh, whoa, I didn't even expect to even be talking to my computer. Like, I didn't expect this to be a way to interface with this stuff. - Sure, and here's what I think this is a semi-informed take and I want to get your reaction. I'm expecting ambient, contextually aware, lots of sensors like a puck or, you know, almost like a walkman that you could wear, put in your pocket when you need it, set it on your desk, set it on a table between you and someone else. But it has like a personality, it's proactive. Like, it'll interrupt you, make conversation and say, it'll actually know your books, you can't do that or something like that. It's like almost this other being. I think they may tweak that and maybe, you know, there's probably contention internally about how much of a being to make it. But that's my bet is that it goes in that direction, which would be, like you said, it'd be pretty, pretty unique and pretty new. React to that. - Maybe, I mean, that sounds really cool. I think, you know, one thing I like to relate these two sometimes is like imaginary friends. Like, when you're a kid, you have your imaginary friend and like, it feels so real and like, he doesn't like, he only reacts when you react. Like, it's not maybe as proactive, maybe, I don't know. I don't know. I honestly don't know. I can go down this rabbit hole forever, but I do think there's something there where people, I love the, this thing called the little guy or something. Like the little, it's like a star. - And then they're farming guests. - Oh, real kids. - No way, okay, no way, no way. - Jordan, do you really not know that else did that too? That Daniel said that. - I did not, I swear to God on my life. - Well, the Daniel said that I didn't do any of it on the show. - Oh, that's true, yeah. - There's something about this, like, Sunday, little guy. There's just something, I don't know, I was talking to Chris about this the other day. It's like, and somebody posted this really funny thing, I'm sorry, I'm going to load over time, but like, some of these, teenagers nearing taste slop or like aesthetic slop. And I was like, oh, that's, I get what you're trying to say. Like, they're almost like a victim of their own success. Like, them being so good and like, people being able to be like, "Huh, my favorite product is iPhone." And teen generic, like, it almost like, but it's like, yes, we know that. Like, we already, we can all agree that that's good. So, yeah, I don't know how to explain this kind of like, thread between like, Sunday, little guy, like that world. But, I don't know, there's something novel about it still to me. Sorry, it's kind of like a word salad. - That's all good. What's important to me, Alex now has proof that I do have taste. And I am the taste maker that I aspire to be. - Make it real. - Yeah. Well, thank you, Jordan. We will let you get back to the workshop. - Yes. Thank you. - Thanks for coming on. - No, thanks for having me and, uh, what you guys are doing and this is great. So. - Thank you. - Yeah, thank you. - And that's it for this week's episode. Stay tuned after the break for Alex and my reactions. (upbeat music) - This episode is brought to you by Accenture. Spotify and Accenture are working together to reinvent the rhythm of ad sales, using automation, analytics and smarter workflows to simplify campaign delivery and access better data across the business. The result, less time spent on operations, more time connecting brands with the moments and fandoms that matter most. You think you know a browser, but Gemini and Chrome, that's new. It can help you with practically anything on the web, like restoring a vintage motorcycle from a 50-page restoration block or finally break down that long article you've had open for weeks. Ready to make anything online make sense? (soft music) You know, we need more people in the tech industry like Jordan. We think about AI is not something that's replacing people but empowering them. And I think if more people building this talked like Jordan, I think you wouldn't have politicians railing against data centers and Bernie Sanders pitching a sovereign wealth fund for AI companies like you did in the New York Times the other day. I think this rhetoric about AI is really a result of the way that leaders, especially if the big labs have been talking about it. And then you've got people like Jordan who are actually building tools that hundreds of thousands of creators are using and sees it I think correctly as an empowerment thing and also just a natural evolution of how these jobs change with new technology. - I mean, it sounds so silly but like to be solving a real problem which is that you have to draw manually every single side of the shoe and how much longer it takes to see if things are working or not, you know, that's a real problem. And I mean, that's why we didn't get super into it but that's why like you have companies like new balance, Ford and Canon. I mean, I assume it's a lot harder to get your tool into Ford than into a new tech startup that is always in the market for trying new tools, right? - Well, it sounds like he's winning on product. I mean, he said it's bottoms up and you know, the designers are bringing it into their workflow and into their companies and that's how Slack grew back in the day. That's how Figma grew. That's how all these great kind of generational SaaS, you know, creative platforms, collaboration platforms grow initially. So it seems like he's approaching it right. And yeah, to go from, you know, kid in Detroit, works at Honda, goes to Nvidia. We didn't say this in the interview but gets shouted out by Jensen during a GTC keynote a few years ago and now doing his own thing. It's cool, it's very cool to see. - I'm also hopeful that we will have more characterful products in the world as a result. I think at the end of the day of these companies still want to make something that is mass market appealing. But you have to hope that if you could present your idea better that it has a higher chance of getting through executives and other barriers internally. And I mean, you know, we're in a time with a lot of visual consolidation, I think, especially in tech. I mean, former guest of the show Karl Pay, you know, his whole mission is to make tech fun again. And I'm wearing the nothing headphones now, even though they don't connect as well as Apple AirPods and they aren't as seamless and they don't have quite the same Apple TV spatial audio experiences. But I'm wearing them because they rule, you know? - Well, it's because you will look cool and talking to each other on Downtown LA. - Tokyo town. Yes, exactly. And so I do think there is a desire for a lot of this stuff. And as Jordan alluded to the teenage engineering design that tech has kind of the alt tech hipsters had been following for the last 10 years. Maybe a little bit played out and we have to see what's next. But you're already seeing like a lot of interesting innovation whether it's in shoes or glasses or some of the new cars, you know, as much as you might hate something like the Ferrari Luce or even the Cybertruck. They're at least putting things into the zeitgeist that are so different that maybe they give someone the freedom to try something new and experiment. - I also just want to acknowledge how we're thinking about the show. You know, we're about what? September, we're about nine months in. Time is a blur, 30-ish episodes. And we've experimented a lot with the kind of guests we have on the format, how much we do news, whether we talk at the front, whether we talk at the back. And I think what we're doing this episode and what I think we've decided we're going to do more of is really explore these more earlier stage interesting companies and founders that, you know, I think it's going to be a mix of us putting you onto things that we think are cool and unique. And then also maybe some names in the early stage world that are a little more known to start. But I think we're going to mix it up and make it interesting. And yeah, I feel good. I feel good about the direction we're going. I think we bring a good balance to these conversations where we're not going to glaze the guests too much. But, you know, we're also very genuinely interested in how these interesting startups are getting built. - How are you thinking about it as opposed to your newsletter? I mean, I feel like it's been a little bit blurry who's going to be interviewed where? I'm sure that way a little bit. - Well, the newsletter is such a fast-paced thing. It's very on the news and very quick. And a little more business-focused, right? I mean, I'm very interested in this stuff. We're talking about the show, but on the show, but I also have a lot of interest in kind of how AI is impacting the financials of these companies, the economy, the palace intrigue of the executives, where they're going, all of that. And you're definitely going to be getting that in sources. And then, you know, sometimes people can only talk for 20, 30 minutes. And I think this show is a good example of a format where, you know, you really get to know someone more. It's a hang. And I think a newsletter isn't a hang, and that's something that I'm more cognizant of lately. I think people want -- they're looking to newsletters for something to get out of it. And not that you don't do that with podcasts, but I think there's this hanging out aspects of a good podcast that we're trying to achieve here, and it just -- it doesn't lend itself to the fast-paced, quicker world of the newsletter. Please send us feedback if you like the new format, or hate it, or have ideas for how to improve the show. You could find us, where can you find us? You can feel free to email me, [email protected] if you made it this far, you deserve to know my email. Yeah. I'd never get my email. How would you like to be found? I think given my email in here ever either, I'm AlexetSources.News. I'm a little scared to say that aloud. I've got an alias on my website that I filter, but you know, the real heads who are listening at this point, they can actually email me. Yeah. You don't want to -- that's the beauty of creating a new email address is that you're not yet in the PR wankster email database, which you will never ever get out of. Yeah. And you know what? Just saying that means that an agent has likely crawled this transcript, and it now is in the PR database. And that's it for this week's show. Thank you to Jordan for coming on. Yes. If you like this show, don't forget to like and subscribe everywhere you get podcasts. We are access.show online, and you can find us in video via access pod on YouTube. And if you really like this episode, we'd love a five-star review. I thumbs up, and if you really feel like it, share it with a friend. You could find me @hamburgeronTwitter and itmeaning.com for your storytelling needs. You can find my newsletter at sources.news. Access is part of the Vox Media Podcast Network, and the show is produced by Hook Creators. See you next time. That's nice. Go on, book it. It's easy. Booking.com. Booking. Yeah. I want to date with Rawls. Cardi says. Rawls. Ruffer asks. This is the love story of real-hinged couple Cartian rocker, written and read by me, Nicola Dynan. Listen to the free audiobook now.