Is atomic design dead? with Brad Frost === [00:00:00] Hi, this is Emily, producer for PodRocket. And we want to know, how did you find us? Did you see us on Twitter and a newsletter, or maybe we were recommended by a friend? Let us know by sending an email to me at emilykohannakettner at logrocket. com or tweet at us at podrocketpod. You can find all these links in our show notes. Thanks again for listening. And here's the show. Hello and welcome to PodRocket, a web development podcast brought to you by LogRocket. LogRocket helps software teams improve user experience with session replay, error tracking, and product analytics. You can try it for free at logrocket. com. I'm Noel, and today having Brad Frost back. He's here to cover his latest conference talk, Is Atomic Design Dead? What a title he's a web design system consultant, web designer, speaker, and writer. Welcome back, Brett. Thanks for having me, Noel. Of course. Yeah, I'm excited to get in. I think it's been like a year since we spoke last. It's been a little bit. But yeah, some stuff's kind of changed. The, [00:01:00] I don't know, the developer ethos around design systems has changed to be sure. I think so. I'm excited to dig in a little bit. Could you give us kind of the On what you, how would you, how you define like a design system? Maybe someone who isn't thinking about this a lot, like just kind of frame it for them so they can follow along Yeah. Yeah. It's such like an unfortunate name because it's design system. So it's like, oh yeah, this is the thing for design. But it's really, there's that parable, the blind men and the elephant, which is just kind of, you know, they all approach this elephant and they're asked to define what an elephant is and the person, you know, touching the tail is like, oh, it's like a snake or A Yeah. Whatever. branch and oh the person touching the leg is oh, it's like a tree trunk or whatever and that really holds true for design systems because basically it gets into You know, all these different disciplines, everyone responsible for creating the user interfaces [00:02:00] for websites and apps and stuff like that. So everybody kind of has their own different lens to it. But with all that wind up, I tend to define a design system as the official story of how your. organization designs and builds user interfaces. And so a design system really kind of contains the shared assets, the kind of solved problems, the settled science that are meant to be applied to an organization's you know, sort of software landscape, right? So oftentimes That boils down to the assets of a design system represented in design in code and documented in some form like a reference website like material. io or stuff like that. But but yeah, at the end of the day, it's usually the you're kind of common, fair, kind of boring, Low level UI components, right? Here's like our Lego kit. Here's our buttons. Here's our accordions. Here's our tabs Here's our form [00:03:00] controls and things like that we use to assemble to create digital products. gotcha. Gotcha. So in, in the talk, you kind of give an overview of how design systems have morphed alongside tools and technology over the past several years. I think a lot, there's been a lot more focus on tooling to make this process easier. I don't know if all of that was kind of just done with Let's make the design system really good, you know, like the, this, the pipeline really good here, or if it was just kind of by virtue of people, like seeing acute problems, fixing them, and then design systems kind of being the framework with which we think about and structure these things. But I guess, regardless, can you kind of, can you cover a little bit about how that kind of relationship has evolved recently? Yeah, Yeah, there's definitely a chicken or the egg kind of thing that you're kind of alluding to which is you know, how much is design systems driving the evolution of different tools and technologies [00:04:00] versus, you know, how much are design systems adapting to, tools and technologies. So like a good example is, you know, All the way back in the day, back in the GeoCities days, it's like you were building websites. If you were designing websites, you were just in HTML, probably copying and pasting some animated GIFs somewhere, right? And, And that was the extent of web design in the sort of like 90s. Then kind of creating killer websites came along. People were like, Oh, you could use Photoshop. And here's this like slice tool and image ready and sort of stuff like that. And so. You kind of had this like first generation of let's just chop up a picture and graft it onto a website effectively. And then tools like Sketch came along that are like, Hey, this is more of like an interface design tool. And then Figma came along and evolve that a little further. And kind of along with that, they kind of realized Oh yeah. designers actually, believe it or [00:05:00] not, work with other designers. And it's important for them to kind of share the same assets. So things like libraries, and design tools, it like became like a really important piece of this puzzle. Cause up until then, it was just kind of like this. vague, either a brand style guide, or you might have a UI, like sticker cheap, but it wasn't connected in any meaningful way. So, so that evolution on the sort of design track has been important. But then similarly, like on the dev side, we have this evolution from, again, DO cities, just HTML, really nothing. And it was like kind of in the like 2010 ish era that sort of this idea of even a pattern library kind of came to be tools like bootstrap started to emerge where it's just Oh, here's the thing. You check the CSS file in the head of your document, match these classes, and you get this look and feel for these user interface components. But [00:06:00] again, like not terribly connected. It was like, you know, kind of pseudo connected a little bit. And so it's really with the emergence of things like ES modules and stuff like that, that you're actually able to bundle up. Reusable bits of code and share them and manage them in that way. So what we've seen over the last number of years is this kind of ES modules. And then of course the exclusive rise of frameworks and stuff like react becoming super popular where you're able to actually, Oh, we can create. A react based component library and then all of our different react applications can consume that similar and you know, whether it's angular view doesn't really matter. It's the same sort of concept and then sort of what we've been seeing over the last, like last four, almost five years as we've increasingly been building. Design systems and helping our clients build design systems using web components, right? So like web components as a standard part of [00:07:00] the web platform interoperable with React, Angular, Vue, WordPress, Drupal, the rest, right? So, so what we've seen is this kind of evolution of the technology getting a lot closer to yes, we could actually create. a true source of truth for our front of the front end code, right? The HTML, CSS, presentational JavaScript, like kind of intentionally dumb UI building blocks, but that could travel to all these different places so that the downstream teams don't have to worry about corner radius or, you know, is the... Is this the right hex value for this button? You're just able to literally import these components and get to work kind of wiring them up and breathing life into them. Yeah. Yeah. We, I think we talked about this quite a bit last time, kind of just like design systems role. In bridging, kind of all three of those gaps where there was like [00:08:00] disparity , on the design side and tools disparity on the product side and tools. And the problem, like bridging that gap between product and like the design system itself. Or I guess product and design itself, the system kind of being the answer that is, or the tool. Or maybe the mental framework that is trying to solve that problem. But again, we could rehash that all day for sure. So let's like get into the meat a little bit more here. What is atomic design? And how does it help people think about this? So atomic design is really a mental model for thinking about user interfaces in this kind of interconnected hierarchical way. And it was really, again, those 10 years ago, I mentioned it was kind of born in that era where it's just like, This notion of kind of just vaguely components are a thing or it's like here's a button or here's like an alert or whatever was just kind of starting to emerge and it's you can take these pieces, these components and build things out of them. And that continues to [00:09:00] be true today, but it's not particularly nuanced. And what I was doing those 10 years ago was like kind of stratifying it, kind of putting a finer point on it, adding some nuance where it's just okay, if we were to like really explode any user interface into its atomic pieces you're going to end up with these very like low level. primitive things, right? A radio button you know, a badge, a text node an input field or something like that. And all of those things have their own properties. They, you know, have certain behaviors you, you need to sort of consider, even for that sort of small of a thing, there's a lot going on in, in just the lowly checkbox, right? is it unchecked? Is it indeterminate? Like all the, all of that stuff. But then they're, but they're not terribly practical on their own, right? So it's like putting those , atomic bits those atoms together into these relatively simple [00:10:00] components that I call molecules. Now all of a sudden you have like atex field that's comprised of, you know, a label, an input field, maybe some helper text or some sort of like required. Icon or something like that, but now all of that takes on its own unique properties, the same way that in chemistry, two hydrogen atoms and an oxygen atom come together and take on their unique kind of water molecule properties. And then kind of from there, then it grows in complexity. Those molecules get combined together further into say, like a. Whatever, a contact form or something like that. First name, last name, email, text area, whatever. Here's now this discrete chunk of interface that could be dropped in wherever that's needed. And then from there, those more complex organisms, what I call them get combined into a page layout at a template level, where you're kind of like having all these things hang together, and [00:11:00] then the page layout level, which is the kind of final form of all of this stuff is where we kind of pour real content, real scenarios, real state into a given sort of template, right? So here's what the homepage template looks like. Here's what that hero looks like with this specific photo applied to it. Here's what it looks like with this specific tagline. Here's what it looks like with this specific CTA. Or here's what it looks like if you're familiar. A user versus an admin, right? Here's like all those, like all that sort of different variants that sort of comes into play when you're making digital products that, you know, whether it's user generated content or whether it's just all the complexities that kind of come with the software landscape, your design system that sort of powers that whole experience needs to be able to account for. All of that complexity, right? Like it's, you can't just be like, Oh here's [00:12:00] this accordion, but it doesn't take into account the need for, Oh, well, we need to slot in a badge for this specific thing. So it's it. What atomic design does is it really kind of paints that picture of like how that lowly checkbox or radio button Actually finds its way into like real products that real human beings use and that's where like I'd say that You know this sort of spoiler of my talk is that all these years later It continues to be I think a really relevant and important mental model because a lot of times What we tend to encounter our design system teams increasingly growing up there. Many organizations are establishing their own centralized design system team that sort of manages this. And a lot of times you can kind of see this drift or this disconnect between the people working on the system. And then that sort of product teams [00:13:00] who are like trying to move fast and ship things and all of this stuff. And there's this kind of like. Weird dichotomy sometimes, you know, whatever, The design system team has become the pattern police and they're trying to slap the wrists of everybody for using things incorrectly, or they're just kind of off to the side doing these other things and these These you know, components are these kind of platonic ideals, but aren't actually like relevant to the product teams using them. So like what atomic design as a mental model does is like really underscores the fact that there's a design system, there's products that use that design system. The design system informs and influences the products that are built with it. And in turn, the products that are built with it inform and influence the design system. So it's really that kind of virtuous cycle that's so incredibly important to get right. And that in our work, as we duck our heads into all of these [00:14:00] different you know, jumbo size companies, especially, it's that's, Often absent or at least it's just hard. It's hard to get that right Yeah. I think that, that makes a lot of sense and I'm sure a lot of devs are still in, in places where they are, you know, they haven't really been able to go through this journey for one reason or another. You know, there's just not support for it in the company. Don't have time, whatever it may be. So I think the question that may be on their minds is like, why what is the threat? Two atomic design. Why would it be dead? Is this just because, you know, people, is there anything recently that's been kind of, you know, a pressure in the opposite direction? I don't think so not it, you know, like it's ultimately like a buzzword You know what? And for good and for ill, it's been an interesting journey in that department, but being part of a thing and I think that like That the only real sort of like threat is just like something that's a [00:15:00] buzzword or new and shiny is no longer right think of like web 2. 0 or, you know, just even responsive design, which is still like a very sound concept that is very much a part of our lives isn't like the new hotness anymore. And so I think that's really the only like threat is that it's just it's no longer a new and shiny kind of thing. But I think that sort of, that doesn't necessarily impact whether or not it's sound or not. Yeah. Yeah. I think kind of after that, you know, the, I guess in general with trends in tech specifically, there's always the fervor around the thing and then the debate is is this actually a thing that's going to be a long term you know, like milestone or something that impacts the industry and how we develop software versus are we just excited about this at the moment? And boy, do we have very short. It's Yeah. Attention spans. Yeah. Yeah. Yeah. just it's did you, do you literally not remember going through this a half a year ago? Oh, for sure. . Is there, I guess [00:16:00] in, in the talk, you talk about like a global design system. What is that? Is that what we've been talking about here? Like a system that's used just kind of, you know, ubiquitously through an organization or is there something more to it than that? Yeah, so what I was saying earlier is that a design system is meant to kind of capture the boring stuff right for creating user interfaces at a company, you know, you go into one of these big company, I use big companies, because we work with a lot of big companies, but it's, it helps highlight the point it's the same holds true, even if you just have Two websites. It's like it's hard to juggle and manage things in between but it especially matters whenever you're talking about You know dozens of teams managing dozens or hundreds of applications and whenever you see all of these different app teams, you know, each creating their own button, each creating their own accordion, each working on form controls just [00:17:00] because of how they're wiring them up to the backend or whatever. You see a lot of waste, right? Like it's this wow, like they shouldn't be doing that because that's not a good use of their time, right? It's fraught with errors, accessibility errors, like all this stuff. A lot of times, just like good front end. Chops are hard to come by and organizations, especially whenever you're just like hurry rush, trying to get a feature ticket done or whatever. So what we see at these organizations are like, is all of this waste and duplicative work. So let's create a design system to You know, solve that problem, solve those boring bits those downstream teams can then just consume it better quality, faster production work, because they're not having to reinvent the wheel every time, whatever. So the irony is that if you zoom all the way out and what my vantage point is, as a consultant. that sort of works with all of these companies. And I've been for the last [00:18:00] decade, helping many organizations establish and evolve their design systems is that we're seeing every single organization creating their design system button, their design system accordion, their design system tabs, their design system, text fields, and so on. And so this idea of a global design system is really Hey, in the same way that we're trying to take. This pain and reinventing the wheel out of individual app developer teams. What if we just take that pain away for literally everyone? Like, why don't we just have one accordion component that like the world uses? Sure. And why don't we just have this sort of again, not every bit of UI, but the sort of common fair, the stuff that if you were to duck your head into many of the popular design systems out there, what you would see is. You know, the same stuff, [00:19:00] right? You would see buttons in almost every design system. You would see badges. You would see, you know, cards, you would see these things But yeah, so, so you're going to see the same stuff the world over, right? And whenever you, especially look adjacent to this world a little bit, you look at the world of accessibility. There's this thing called the WebAIM million, it's a company WebAIM, they go across, they crawl the internet every year and they give a report on like how the top million websites are doing in the accessibility department. And spoiler alert. Word awful. It's something like 95 percent of websites of these like top million websites have like many accessibility errors. And when you look at like the nature of those errors it's really common stuff. You know, it's, it's labels not having Like linters should be catching this kind of stuff. totally. Right. And so it's like, well. Clearly we're not, [00:20:00] clearly we're not doing this right and so what would it look like to have sort of a centralized thing, whether that's my, my best sort of proxy, and again, like I haven't workshopped this, this is an idea at this moment, here's this W3C sanctioned it, What component library if there are accessibility issues with the accordion? Well, we do a PR against that fix that accessibility issue and literally the world benefits Right from that fix So it's no longer just this like, you know We're all sort of building the same stuff and I think it's important to differentiate that it's like I'm not necessarily like Advocating for oh, we need like a bunch more HTML elements I don't actually think that's the case, although potentially these things could graduate into, honest to goodness, HTML elements. I think what we need is a layer that sort of sits [00:21:00] on top of it. The HTML elements are kind of like the low level, if you think of like, IKEA furniture, they're the dowel rods, the screws, the real sort of raw materials to make this stuff. And what I think that I'm advocating for is we need more prefab. Stuff that's you know, halfway there. All you need to do is either like skin it or compose it. And you're going to get way further down the line than you would on your Yeah. Do you think that like, I'm gonna put on my skeptics hat, devil's advocate, for a minute here. Do you think that the frameworks are the reason that we don't have this kind of like a standards body of some kind giving us a set of, you know, slightly higher order primitives already? Or do you think there just hasn't been enough kind of, pull Or, you know, desire for this to really make this happen. Yeah I frankly just chalk it up to, this is an immature medium, the [00:22:00] web, you know, this whole thing has only been around for 30 odd years or whatever, but even kind of coming back to ES modules as a thing, it's that's only, what, 2007 ish or something, but web components as a technology. Has only, I think really matured enough in the last like couple of years. So it's like up until now, it's and you've seen this, we've seen people go, why don't they just build React into the browser? Let's like pick this stuff. That's a little harder to come by because that's like a proprietary technology Now what we have is like the like foundation level technologies to actually accomplish this which Wouldn't have been true three or four years ago even so so I think that it's like there's a number of things It's like these concepts take time to mature and like for all these teams are like As I see in my daily work, all these teams are [00:23:00] like getting it, they're seeing the value for it, but they're still like at those early days of kind of like building this stuff out, or if they have it, they've only had it for a couple years tops. So it's this is all just still kind of like new science, which is kind of a good thing. And that's kind of, I think why this idea. is interesting because it's like, okay, we have the tech in place now to actually accomplish this. Everyone's primed for it mentally for it. And also things haven't gone too far into like kind of crazy framework land or anything like that, where it's we're able to, you know, I think still ride the wave of, oh yeah, we should make components is like still. Not well, it's not like technically like the newest concept right the second It's still pretty young and it's still like it's still I think be Yeah. I think organizations haven't made that leap yet, like overwhelmingly. [00:24:00] So, so I feel Yeah, that's right. It's like Definitely we see some good success stories. It's proven to be like a sound concept I think now it's like okay now, how do we do it? Especially like kind of just removing the kind of like capitalistic parts of it and it's more just like thinking of it as just general shared infrastructure for the web. It's that same sort of type of thing as like a standard or whatever. I love the W3C. I love, you know, all of the great stuff that they've basically, you know, provided in order for us to always build upon it. I think that there's like now this opportunity to kind of, they're at the basement level, like really making sure that these like ones and zeros are like doing what they need to do. And that's awesome. And that's like very necessary. But I think that there needs to be something in between that layer. And then the sort of like end user developers who are like making individual things, it's like, we need this [00:25:00] like, kind of what's this layer in between that sort of takes those ones and zeros or those sort of basement level elements gives us all like a solid base to start off of That provides enough flexibility, but also enough, just like kind of You know, feature set to, to focus on other things, like just thinking of a date picker. Right? So let's just talk about that for a second. How many millions. Millions upon millions of human beings, like life energy, is going into creating, maintaining, debugging a date picker. And it's, that's a tragedy. That shouldn't happen. That shouldn't happen. I was like let's actually fix that. Let's do something about that. Yeah, for sure. Yeah. Again, I think there's like nuance debate one can get into there. It's like, Oh, well we kind of have like the date form input field and that's supposed to be doing what a date picker is, but we've decided that UI abstraction is gross. So I don't know. It's [00:26:00] yeah, whatever. But I agree. Do you think that there is I guess maybe, you know, this ecosystem better than I do. Are there any just kind of prominent Just like web component libraries that are like kind of coming out as front runners or trying to push into this space at all. Cause it, it feels like that's the most analogous thing we've got right now is even just you know, libraries and in frameworks like react and view libraries and stuff that are kind of trying to do this. Hey, we're not very styled. You can use this as a baseline, do whatever you want. But we give you all the primitives. So is there anything kind of that you're, you have your eye on? Yeah there's a library called Shoelace that, that's really good, and it's, here's Web Components, but it does provide it. I think sort of some styling defaults, obviously, you can rip those things out. Like for a day picker, you're gonna need some styling. Yeah. to have it be well, yeah, but there's, yeah, there's styling and then there's like the design language, more of the like brand or like theme or kind of the more obvious, these [00:27:00] buttons are purple or these buttons are gold, right? So it's like, I think, There's plenty of stuff to mine from this, whether that, whether it's a library like shoelace, which is web component base, lots of other like web component base, kind of like libraries, but also any proprietary frame, you know, the component library, something that is like react base or angular base it's, here's a component, here's the API that exposes, here's like the shape of this thing, and the technology doesn't really matter. Yeah. At least from like an inspiration perspective. But what's really interesting with respect to design systems is. What we found is like the actual sort of architecture and the build of a component isn't the thing that makes or breaks a design system, a lot of that technical architecture really needs to be informed and influenced by the organization's culture and just kind of whole vibe, like what [00:28:00] they kind of believe in. And like, How much do you trust downstream developers versus not? There's a lot of like value system kind of stuff baked into that, that influence a technical architecture for a design system. So it's not just like a matter of going. Oh, let's just use material and call it a day because that's built through the lens of a specific organization. So I think this sort of exercise would be how do we take the best qualities and the shared qualities of all of these, but really kind of come at this from a, this is for the world. This is like shared infrastructure and , given that brief, how would we go about kind of architecting this? So, so there's plenty of good stuff to mine from and steal from and all that stuff, but ultimately the exercise would need to be one of this is a, you know, we're bringing the world together with UI components kind of thing, which is really, I think it's fun. I think it's a really exciting concept, like [00:29:00] whether or not, this idea has legs or not. I think that like the problem as I see it or the opportunity for something like this to exist is needed. Like it's there whether or not it actually comes to life or not. Yeah. I, yeah. I have a suspicion that a lot of people see the utility, and they feel the need for this. Yeah. Yeah. And that's what it's been fun to start airing this out a little bit and introducing it because. Everyone so far has just gone, yeah, it makes a ton of sense. We had that. I don't think that there's anybody that's no, like I need to really own my accordion component. Cause here's the other thing is like, just being really pragmatic here. It's talk about an 80, 20 kind of rule for something like this. If you need to do some crazy date picker stuff, something that's like totally, you know, wild and out there. You still have all the, again, those like low level IKEA parts. You could still do that. You can still do your own thing. [00:30:00] That's what we're doing right now. So you can continue to do that. But for all intents and purposes, like you're talking about again, 80 percent of use cases for a lot of these components let me guess a text field needs like some sort of like error treatment. You know, maybe a valid state. Oh, you need to be able to put like an icon in there, slot in like a button for a, you know, show, hide, toggle kind of thing. It's there's really common architectural stuff that you could solve. Many of the world's use cases with, just with a solid foundation. yeah, for sure. I feel like you could go out to like NPM and look at a bunch of the most downloaded packages and be like, these are the things that should be like a primitive, right? Like we could iterate all this pretty quickly there. You also, in your talk, you briefly touch on AI and how it's going to change this relationship or might, what is your hypothesis there? Yeah. Obviously this is like brave new world for all of us here. [00:31:00] But as it pertains to kind of the world of, you know, designers and developers and stuff like that, I think that there's some really. No brainer use cases so for one, just Hey AI, make me a component called, you know, a button that has a prop called Variant with primary, secondary, and tertiary, and then, you know, disabled prop, and whatever, and then it would just splat out the source code, and my colleague at Big Medium, Kevin Coyle, has been doing exactly that, where we're able to feed in our design system code base and the code guidelines that go with it, and it stays in its box, meaning that it's not going out and crawling all of the internet like, kind of, ChatGPT does. It's a very localized thing, which ends up being this really... And that's really needing to be tuned for the organization. So we're [00:32:00] able to like kind of suck up a lot of the design documentation, developer documentation, source code, stuff like that, and create like a real sort of shortcut for getting stuff done using the grain of the design system. So that's really cool. And I think that's really helpful. And. So much of it really boils down to, it's well, how much do we want, you know, what does that take away from the developers? And if you're talking about like basic boilerplate kind of stuff, like it's hopefully just taking away some drudgery from their lives. And then they're able to worry about more pressing issues. So in those, in that respect, I think it's good. I think there are many more problematic and terrifying sort of prospects for AI, but setting aside you know, world The social problems. yeah, like a lot of that stuff. I think that one of the most [00:33:00] obvious ones that I do think is important to just. Kinda keep front of mind is just like really understanding these things are out there. These things are going to get used and abused by a lot of people, right? People going, oh, great, I don't even need to hire a web developer designer anymore. I'm just like a person. And I could say make me a marketing website that, you know, does this. And then it just spits that stuff out. And I think that's the kind of stuff that without a real sort of thoughtful. analysis or like kind of going through some sort of like quality control human, human sort of side. It's it's one thing to use AI to help generate some stuff, but there needs to be this kind of like human Analysis and curation and judgment, ultimately of is this good? Is this accessible? Is this sound? Is this ethical? Is this all of this stuff? And I, my worry is [00:34:00] that these tools are just kind of getting just lobbed out there into the abyss with. real incentives to just be like, Oh yeah, I'm just want to get this up and get it done now. And that we're going to basically enter a world where there's just like a lot of garbage out there, but also I think a lot of stuff that could do a lot of harm. So like my like main sense of urgency is that we. It's not no, don't use these tools at all they're bad and evil and whatever, but it really is that we should be really... You know, judging and analyzing and curating the output of these things before we release anything into the world. For sure. So yeah it's crazy times. It's going to be really interesting to see how this stuff plays out, but unlike a lot of other, we were talking about trends earlier and stuff like that. And unlike a lot of, you know, whether it's like blockchain or [00:35:00] whatever, where you kind of like squint and you're like kind of looking for the use case, like with AI, you could like really see some real genuine. Oh, that's helpful. Yeah. For sure. Taking, taking 50 percent of my drudgery production work out of the equation, like that's welcome. Yeah. Or I think like the opposite in there. I think it's very easy to do things like, you know, have the accessibility scan happen. Like running all the time, you can do slightly more powerful things there without running it on a code locally. Yeah. And I think there's, I think there's benefits on both sides. My, My, my colleague, he's been doing a lot of really amazing stuff. And like some of the stuff that it does is like on PR. It will like create an AI like summary of what's going on just by looking at the changes So it's like obviously you can see the diffs and code and whatever but it's to have this thing going on there But also yes running like accessibility audits and like linting and [00:36:00] stuff But like kind of playing it back and just having this It's kind of very human feeling, or compatible, it's less about Line 81, three characters in, so much as it is Oh, this is like really bad or yeah, there's some bad stuff here, but it's not that bad. Exactly. Yeah. Yeah. Yeah. What does it agree? Yeah. Yeah. Very cool. Yeah. I feel like I don't worry too much about Oh, we're all going to be in this hellscape of auto generated websites. Cause I feel like we're like. If people didn't want to do any work, they could already go like use a website generator and spit out a website that I don't feel would look vastly different than anything that, you know, if it's programmatically generated, it's going to be either way. So I don't know, to circle things out on a We'll see. note, we will We'll see. I think that I agree with you on that part. It's Oh, the tools are available. I think that there is something where just how accessible it is. Meaning like the affordances of this thing are just, it's Oh, if you could [00:37:00] send a text message. You can like harness the power of this stuff and that's both very potent, but also very problematic people. Make me a copy of this exact banking website. Log in. But yeah, do this. Yeah. Yeah. Mm Yeah, for sure. For sure. Cool. Nice. Is there anything you wanted to call out from the talk or just in general before we wrap up? Brad? So I'd say that sort of building on the talk, one of the things that my colleagues and I have been doing is I just published this kind of I think it was like 5, 500 words this article that kind of goes deep into the different layers of kind of a design system ecosystem. So that, you know, we've been like kind of talking about the core design system, but then there's like all of these like kind of optional, but often helpful layers that go into a design system. So we've. We were talking about like web components and how they're great, but then still most [00:38:00] applications are building with some sort of framework or c m s or whatever. And so it's like how do you get these like kind of crude, low level web components into actual software? And what we found is that there's this kind of like layer cake. Ecosystem that kind of can power in organizations UI landscape. And so, so I think that's at least worth kind of directing people to just because there's like a lot of nuance in it because it's like, Oh yeah, here's web components. There's a little bit of like glue code that needs to work in order to get react to work, or a team might find it necessary or more ergonomic to have downstream React developers. writing, pulling in React. Formatted components, even if they are under the hood web components. So there's just, there's just like a lot of stuff like that gets, that is what we see in day in and day out. It's not as crude or [00:39:00] Shallow of a depiction of here's a web, here's a design system, here's products and they inform and influence each other. That's true. But then what we've found over the years, especially working with these like really large organizations is there's stuff like, okay, if material design is your base design system, well. What in YouTube and Google Maps and Gmail are all three products that are, you know, using material design. Ostensibly there's this kind of like notion of well, there could be like a layer sitting on top of the material that is contains what we call like the recipes, right? Like the YouTube specific recipes that creates the toolbar for that, or the Gmail sort of slat you know, inbox slat and stuff like that. So, so there's a lot of nuance that goes into design systems. And I think that a lot of the language. Around how we talk about design systems are still very much here's design [00:40:00] systems and here's products in the same way that like whenever we created an atomic design It was like here's components and then here's a product and you can make things with that and it's that's true but I think we need to kind of put a finer point on that and that's kind of what we Is attempting to do with this latest article is let's actually unpack All of the kind of potential layers. Let's talk about native and how you get design systems like into native iOS and Android and other non web kind of platforms. Let's talk about how to take a dumb. Text field component and wrap it in something that is ready to be plugged into, you know, react hooks form or whatever library you're using to manage your state or collect payments or Sort of data grid and stuff like that. So there's again, there's many different layers and tiers to this whole world that and I say in the article it's like this is all pretty complex [00:41:00] and But at the same time, this is kind of the nature of the beast for the modern landscape. We're not, we're no longer just like managing a website. Yeah. We're like managing like a lot of things and so there needs to be Unfortunately, well, unfortunately, it's just the reality of the situation is sometimes if you're a big jumbo organization that needs to deploy many websites and apps, we are going to need to be kind of sophisticated and use some advanced techniques to manage it all. Yeah. And It's tough. Yeah. Yeah. It's a lot. It's a lot to take in, especially, I think yeah, if you're a dev focused on, you know, one little slice of the... Of the pie right now to use your analogy. Cool. We'll get a link to that in the, to that post. He says like a blog post or kind of an article. Yeah. Yeah. I'll link it up. Yeah. We'll get a link in the show notes and we'll get a link to the talk we have been talking about as well yeah, it looks great. I appreciate Of course. Of course. Well, thank you so [00:42:00] much for coming on and chatting with me, Brad. It's been a pleasure. Yeah. Yeah. Likewise. So thank you for having me. This has been Of course. Of course. It was awesome.