60: Design Engineering, Interviews & Job Search | Career Growth with Adam Argyle

Brittany Ellich (00:00.757) Welcome to the Overcommitted Podcast, your weekly dose of real engineering conversations. I’m your host Brittany and I am joined by…

Bethany (00:09.496) Hey, I’m Bethany

Erika (00:11.054) and I’m Erika

Brittany Ellich (00:12.971) The three of us met while working on a team at GitHub and realized we were all obsessed with getting better at what we do. So we decided to start this podcast to share what we’ve learned. We’ll be talking about everything from leveling up your technical skills to navigating your professional development, all with the goal of creating a community where engineers can learn and connect. Today on Overcommitted, we are joined by Adam Argyle, and I’m very excited about it. He spent seven years as a Chrome CSS developer advocate at Google.

Co-hosted the CSS podcast with Una Kravitz, currently is a co-host of the Whiskey Web and Whatnot podcast, and has spent a lot of his tenure trying to convince the web development world successfully that CSS is worth taking seriously as a craft. His role was eliminated in 2025 at Google, unfortunately, and he navigated a months long job search before landing now at Shopify as a staff design engineer, which he has been very candid and public about.

He’s also been the kind of engineer who builds a 30 demo scroll driven animations notebook at bedtime just for fun. Welcome Adam.

Adam Argyle (01:20.851) Thank you so much, great intro, excited to be here, and nice to meet y’all.

Brittany Ellich (01:25.685) Yeah, great to see you again. To kick us off, what is one thing you’re currently building or obsessed with learning right now?

Adam Argyle (01:34.033) Yes, I am currently building a, well, I have a color picker. It’s a web component that allows you to pick, you won’t be surprised, any CSS color. So all of the CSS color spaces and it kind of tries to make this humongous concept of HDR and HD and all these color spaces easy-ish. So it bundles it all into this very familiar feeling color picker that gently helps you pick HDR colors and it outputs

I think the most sane amount of syntax. So it optimizes for like percentages. So instead of you learning that OKLCH Chroma goes from 0.0 to 0.35 and maybe to 0.5, which designers are already like, why are you making me do decimals? It uses percentages and goes from 0 to 100%. So you can be like, I do, I want 100 % Chroma. Just give me all the Chroma. And so I’m like, I will make the design tool, I’ll make this tool for you so that it’s easy to use. And I just announced v0.4.

The other day I did an RFC and people just loved it and didn’t have any complaints and so I released it and now I’m currently trying to get it into Figma which they can render web components in their little iframes and so soon y’all I even have a working version I was working on it right before this a Figma plugin so that you can choose HDR colors that are CSS compatible in Figma.

Brittany Ellich (03:00.427) That’s incredibly cool. Very neat. So it seems like you’re like kind of into CSS a bit, right? That’s like.

Adam Argyle (03:10.129) Little bit, little bit. Yeah, it was funny. Everyone at work all the time was just like, it’s easy. Just give it to the juniors. And I was always like, uh-uh, I’m gonna sit here and I’m gonna make sure it’s done right. here, yeah, you’ll keep carrying on. Like it’s easy and it’s child’s play. Just go ahead, you know, whatever. I’ll be over here in the front end just making sure it all looks beautiful and owning it. And I just, and I loved having something that was visual to work with. So I just haven’t let it go. And it’s only gotten more complex.

and more respected, which is nice. It used to be just poo-pooed on all the time and I was just, didn’t care. I was like, whatever, I know I’m right. You’re wrong.

Brittany Ellich (03:45.27) Yeah, no, that’s totally valid. is, it’s really somewhat, I guess, easy to do. I don’t think you can even say that anymore, but like, it’s really hard to do well. And there’s so much to know, very easy to get it wrong. So.

Erika (03:46.476) Yeah.

Erika (04:01.78) Yeah. I mean, yeah, I feel like it’s an extension of design and people will say that design is easy until they try to do it. And it’s the same with CSS. You think it’s easy until you really try to do something well and like maybe one or two things. Sure. But then when you get to like actually putting together a whole design and putting, you know, implementing something at scale, it does take

It takes architectural skills, takes deep technical knowledge, it takes an eye for when something looks right or when it’s broken. Yeah, so definitely have experienced the underestimation of what it takes in myself and also from other engineers as well.

Adam Argyle (04:53.52) Yeah, I say it’s the easiest to get started, the hardest to master. It’s subjective, which means LLMs suck at it. They’re like, I did it. And you’re like, did you? You don’t even know what you did. All you did was make a sentence. And then you’re like, I did the CSS thing. And you’re like, no, you didn’t. And it gets even harder. Like, you know Scott Pilgrim and you’ve got to fight all the evil exes. It’s kind of like with CSS where you’re like, all right, my layout works.

Brittany Ellich (05:04.129) Mm-hmm.

Adam Argyle (05:20.184) I beat the boss and then the, you know, the next boss shows up. yeah, just, which is a work in Safari. You’re like, there’s another X. Damn. Okay. And then it’s like, there’s another X. And then someone shows up with screen readers and they’re like, guess what? There’s actually seven screen readers that you need to support. got mobile ones and desktop ones. So you might’ve beaten the four bosses, you know, Chrome, Safari, Firefox, and somebody else, but it doesn’t matter. There’s seven more. And you’re just like getting, and you have language translations.

You have different viewports. All of this stuff just adds to the complexity. And that’s what a lot of people get wrong when they do CSS is they get started with color red. They’re like, yeah, I control the universe of this painting space. This canvas is mine. And then later they’re like, wait, users have input on what happens in here? Their font size preference needs to do. I need to respect their motion. Like, my goodness, I thought I could just define the whole world. And then you just realize you’re not in control, which developers hate.

Developers love their control, they want their rust. Give me rust. And it’s like, CSS is the opposite. You pretty much have no control. You have two choices, which is try to grab all the control and fail miserably, or relinquish control and write CSS defensively and with a system and with a way that you of corral the sheep. Never assume you know how many sheep are in the pen, never assume how big the sheep are, and assume there might be a sheep that’s quite rowdy. And now you’re ready to write CSS.

Brittany Ellich (06:44.673) That is an incredible framing. I love that so much. What was the initial thing that got you into going so deep into the world of learning and mastering CSS? Was there anything in particular that occurred?

Erika (06:45.43) Okay.

Adam Argyle (07:01.764) Yeah, I was in a computer science college. was taking SQL and PHP, Java. So I was writing these Java applets. And then eventually we got to Flash. And even in PHP, I noticed something really quick was pretty much everybody in all my classes were really happy if the console printed the result. know? Hey, what’s seven plus seven? 14. It did it, everybody!

my program works and they were just so happy it worked. And then I was like, yeah, but it looks really dumb. You know, it looks bad. And I was like, and I cared. And so mine would look better than other people’s and then be like, why’d you do the extra part that did the sizing? And I’d be like, cause it looks better. Why are you being so judgmental about my version you turd? Anyway, so then we got to Flash and we got to Flash and I’m a video game player.

Brittany Ellich (07:37.697) You

Adam Argyle (07:54.995) And you know, and I love animations and I remember specifically in Rock Band, you hover over one of the buttons and like 50 little rock hands would pop up from the bottom and just kind of like spring in and float. And they all had like a different scale and a different, I was like, that’s cool, staggered animation, rock fingers. And so I did that in Flash in one of my classes. My button looked rad and I was the only one. I looked around the entire room. I was the only one that added animation. did something like meaningfully visual.

visual and cool with my code. And I went, I’m in the wrong place. I am in the wrong place. I need to go somewhere else. And so I finished my two year computer science degree and went to Art Institute of Seattle and studied design. And I skipped out of all the coding classes. I was too advanced for that. And started taking design courses, motion graphics, 3D rendering, lighting, concept art, typography. I traced, I traced letter forms on really big pieces of paper and I loved it.

All the while while I was there, I got a job at an agency building apps for early iPads, iPhones, web apps and stuff like that. And so I just immediately was like, I like both. I really liked to code. I really liked design. I really love learning from the designers on the job because they create things that I couldn’t imagine, but then I finished them, you know, and then I learned intimately how to do a good design because I wasn’t just moving pixels on a canvas, changing border radiuses and not using tokens. I needed to make a system.

and I needed to make a system that had consistency and I would feel inconsistencies where Figma never really made you feel them or it wasn’t Figma back then. But anyway, and so that was kind of how I got into this space and people started to rely on me to be the one that actually did the good front end work because so many other people made franken turds. You you’d have the beautiful design and you’d give it to the dev team and they’d be like, we built it. And you’re like, that is not the same. There is a lot of things different.

from that one to this one. so mine always looked good and I took pride in it. In fact, I would add things because back then there’s static images. I would add animations. I’d always add hover effects, focus effects. I cared. So that’s the origin story.

Brittany Ellich (10:06.901) I love that. And you’re currently a design engineer, correct? And you were a developer advocate at Google. Is that right? Or?

Adam Argyle (10:15.57) Yeah, at Google my first year was on the design system team of Google Cloud. I was a UX engineer there for a year and I was building a design tool to empower the design teams, managing the design system, building components, et cetera. And then all before that, I’ve been building apps since like 2000 something. So I’m very savvy and just like building. And I’ve been a creative director. I’ve had many, roles, but generally I preferred coding because

People were always like, you can code? I bow to code. And then if you were a designer, you’d be like, I can make designs. And they were like, I don’t like that blue. And I’d be like, OK, I don’t like this role. I like the role where people think everything I output is great and not the role where everybody has an opinion and it’s rude. So I pursued the coding one. And so yeah, now I’m at Shopify working on Admin, which is very similar to Google Cloud. A lot of the same problems.

you know, there’s just tons of people writing code all day. You’ve got a design system. You need them to use it. You need a reliable design system. anyway, all those sorts of things. So that’s what I’m doing at Shopify now. Yeah.

Erika (11:23.298) I’m curious, like you kind of mentioned a couple times, like the, well, you mentioned the handoff between the design team and the engineers and that maybe not always going well. And I guess with your experience in developer advocacy and in all these various roles, how do you think about educating engineers in these design systems and like getting them up to speed on

what skills are required to implement these designs well.

Adam Argyle (11:57.776) Yes, yeah, reversing it where it’s, yeah, need developers to care about the quality that comes out. And they don’t have very many good testing tools. So like one of the first things I noticed resonated with developers specifically was the do’s and don’ts. You know, they’d pull up the design system, look at a component, I need to use this accordion. And then they’d be like, I can’t use it that way because that’s almost always what they’re gonna do, right? They’re like, I’m gonna use an accordion in this part of the UI. And then they’d.

go check the don’ts it’s like don’t use an accordion in this particular case and they’d be like why you know and you’d be like well that’s because it would be a bad decision and we want you to not do that so the do’s and don’ts are really good it’s very binary for them because it is quite subjective and i think that’s what a lot of people don’t like thinking about is well users they they like servers with the game compilers because they’re binary they’re yes no and so they know when they’re doing good and they know when they’re doing bad

And that’s why I think a do and a don’t speaks well to them, just like TDD. They know when their component is done because it’s all green check marks. There’s no red. So as long as you can. So that’s a good step forward. We used to run like, I did these all the time at Google because the developers there weren’t, there’s no magical human out there. There’s like world class at everything. And so we would sit, I would sit down with teams and either intake from them and hear their complaints about what features weren’t available on the platform. And that was a lot of the Chrome work.

But I’d also educate all the time. Like, hey, I saw that you wrote height 333 pixels. That’s a red flag. And they’re like, why? What’s wrong with height 33? And I’m like, well, here we go. First off, you should probably, you might want min height because whenever you set a height explicitly.

you’re gonna run into clipping issues, which is one of the things that people always tease CSS about. CSS is busting out the box. And I’m like, that’s because the turd told the box to be 50 pixels wide and put a humongous letter of text in there. You know, it’s not CSS, it’s fault. It’s the developers. And so then I’d be like, also consider alternative units. Also, 333 pixels, that’s what we call a magic number. And those are not good in your code base. You’re not gonna be able to explain why. You’re like, why’d you pick it?

Adam Argyle (14:11.555) It looked good. That’s not a good enough answer. You need to have a good answer about your units and what you’re choosing. So there’d be a lot of educational pieces here about how to build reliable systems. You had to kind of use the verbiage that they want, which is you can’t break it, right? So it’s resilient. It’s adaptive and flexible. Just kind of like speak into the things that…

orient around their code that they like to write, but make it also about the CSS in the front end. Like these things need to exhibit the same behaviors that have your backend, like all your TypeScript. You like your tight TypeScript. You should like tight CSS. It has the same fulfilling properties to you as a developer. You just have a mental block about why one is better than the other or whatever, and you need to get over it. So I usually was a little blunt with people about it because, well, they came to me and they’re like, I heard you’re good at this and I’d like some help. I’m like, cool, I’m going to tell you what it takes.

Maybe you won’t like it, but it’s okay. You’ll live through it just like you lived through learning Rust. That was way harder or maybe not. Anyway, I’m going to tease Rust a lot, guess, today, Rust is great.

Brittany Ellich (15:18.817) Rust can take it, I think. There’s enough people out there supporting Rust that it can take a little bit of teasing. So you spent quite a bit of time at Google then translating new CSS features for people to use. And now you’re back at Shopify building with those features, I’m assuming, that you had been translating. Has your opinion on building changed at all, having had that experience of building for a browser, and now to?

Adam Argyle (15:22.907) Yeah.

Brittany Ellich (15:47.413) like actually like leveraging those features again.

Adam Argyle (15:51.058) I wouldn’t say change because I came from building intense applications like really big, you know, corporate applications before going to Chrome. Chrome and being on the web standards teams, you definitely start to get some tunnel vision at the, you don’t live in the same reality as enterprise software. Enterprise software has, you know, like Shopify, for example, has, we have people with a POS device that’s an old iPad and that old iPad, they’re not going to update it.

So the app needs to work on a pretty old version of Safari on an old iPad. That drastically limits all of the cool stuff you can use. And progressive enhancement tends to scare a lot of developers and project managers because no one wants to hear that you’re going to do a code fork and that you’re going to manage two versions of the component now. We’re going to have one that does the traditional way and one that uses the new CSS way. And they go, no, that is combatorial explosion.

which we do not want. We already have combinatorial explosion, Adam. We don’t want more of it. So progressive enhancement or graceful degreed, these things don’t tend to exist in enterprise software because it’s hard enough getting new features out the door, let alone building it for two different strategies. So a little bit of like what I’ve, I guess what’s changed and shifted is so yeah, I went from building things a lot with JavaScript into being it.

Google where I built almost everything with CSS and tried to get as far as I could in a very declarative fashion with very little JavaScript. And I’m back in a very JavaScript heavy scenario again. However, I still find that knowing CSS not only gives you a leg up in terms of like the quality of your output and the resiliency of stuff like that, but you get things done faster and with a lot more confidence and with just a lot more attention to detail and quality. So knowing CSS is…

And even now with LLMs especially, so Shopify is very, very AI centric and LLM driven and LLMs suck at layout. They suck at interactions. They suck at animations. They suck at almost all of CSS. And so it’s really meaningful for me to have that skill right now because I go in there and I correct it all the time. And I can’t just prompt it to correct it because it’ll just spin its wheels. You know, I reach my hands in there and I still code manually a little bit, you know, old school.

Brittany Ellich (18:12.565) Whoa, what a flex these days. How do you recommend then, like if LLMs aren’t necessarily the most reliable for CSS and that’s the new way everybody’s learning everything, like how do you recommend people go learn it as deeply as, know, well, probably not as deeply as you’ve learned it because that would take a very, very long time, but at least getting closer to like a very good understanding of CSS given that.

Adam Argyle (18:16.177) You

Brittany Ellich (18:40.459) you probably don’t necessarily want to rely on what an alum is saying about it.

Adam Argyle (18:45.327) Yeah, that’s gonna play into your learning style. I know my learning style is hands-on. I could read a book all day on how to learn a new thing, and then I go try the new thing, and I’m like, well, all that reading didn’t really do good, because I suck at this thing really bad right now. You know, like, I’ve been playing banjo or a lot of instruments, and you can go watch YouTube videos, and they’re like, here’s the technique. It’s quite simple. You do blah, blah, blah, blah, blah, and you’re like, that looks pretty simple. I’m like, okay, let me go try it. I’m like, clink, clink, clink, clink, clink. I’m like, great. I’m gonna need to put in some time.

So you just gotta put your time in, there’s reps with it. Just like with JavaScript, the first time you wrote the word function and open parenthesis and close parenthesis and then you opened up a curly bracket and then a closed curly bracket, hit enter, you formatted it. Like that was so foreign to you when you first started. CSS feels like that for people too. And I would just say, saturate yourself in it. Have an open mind. A learning mentality is really just gonna be what gets you that really far is be willing to kind of be bad at it, willing to learn and ask LLMs, LLMs can…

give you good tips, they just don’t always implement it great. Ask it for options. Ask for more modern variants of these things. So that’s another thing. LLMs like to give you high probability answers. And you can be like, I want the modern CSS version of this. And it would be like, well here, that sets it down a different answer path, at which point it might respond with the line height unit instead of a pixel value, or at starting style for the way that an animation can introduce itself into the page instead of JavaScript.

And so yeah, it’s a acquire the language. So a lot of with LLMs, you need to have the words. And this was my talk at Cascadia last year was I’m trying to give all these JavaScript and AI developers the vocabulary to ask for the new way of doing things or just a modern way of doing things and a more resilient way of doing things with CSS. But if you don’t know what to ask for, you’ll never ask for it. LLMs are just going to keep giving you tailwind unless you ask for something else. And so.

It’s just deliberate effort. There are people you can follow, so maybe they’re sharing tips. You can take courses. There’s courses by a couple folks that are really great. Whatever your learning style is, I would just say if you want to get good at it, I’m always going to tell you, fingers tippy tap on the keyboard. Find a design that you like and recreate it from scratch. And you’ll learn a whole bunch every time you do it. I still learn that. I just failed at March Mad CSS with a

Adam Argyle (21:12.376) on the Syntax show. There’s like a, I just missed a dom node, you know, and I’m in there, I’m like, crap, I gotta write CSS grid really fast right now. It’s like, okay, good thing I write grid all the time, because otherwise this would be really annoying. Anyway, that was a long answer for, I think just practice and an open mind.

Erika (21:31.799) Yeah, I mean, there’s so many variants, I guess, on top of vanilla CSS that I feel like for me, when I’m doing front end, so much of it is dependent on the context and the libraries being used. Are we using variables? Are we using a component library? All those different things. And you have to understand how all the pieces fit together in order to know where to look.

for the answer you’re looking for. Like if you need to use a view component or some kind of like pre-built library component, then you have to know to look there and you have to know how to test it. yeah, so I mean, to your point, like there’s more than the CSS itself that goes into

the development of front end and yeah, the quick fix of prompting the LLM to give you the answer probably won’t work in most cases. Unless, I mean, maybe if you give it all of that context and like give it the full library and all that, maybe it’s more helpful, but yeah.

Brittany Ellich (22:57.545) Yeah, that’s assuming you’re using the same library that you’ve used the entire time in that code base, which almost never happens because you’re always in some sort of weird mix of like, we used to use this and now we’re using this now. And yeah, that it gets really complicated. And I kind of love the fact that some of these areas of development that traditionally were seen as like, like that’s not as important. That’s just the front end are now actually one of the things that

AI agents are probably going to be last to be able to replace. And that knowledge is actually very helpful because they’re really hard to do well. Like accessibility, it’s so hard to do it well. And all of the LLMs have been trained on such inaccessible code that now actually that accessibility skill is incredibly important. And knowing how to do that is incredibly important. And I love that. It’s like a, I don’t know. I think that’s fun. So do you land then?

Adam Argyle (23:52.78) It’s, it’s for people.

Brittany Ellich (23:54.4) Yeah, it’s for people. It’s for people. The stuff that is for people, you still need a person to do that thing. And I don’t think that that’s going to go away anytime soon. So if somebody is like, hey, I need to AI-proof my career, I feel like these front-end skills like CSS and accessibility and stuff, those are really great to invest in right now, I think. Do you agree with that? is that? Yeah. Yeah.

Adam Argyle (24:16.624) I hope they stick around. I don’t know, I’m scared. Because I have it make really beautiful pages all the time. And to a lot of people, that’s way more than just acceptable. know, it’s like, to them, it’s exemplary. It definitely gets different in a really big application scenario. So it’s like it can make a one-pager really cool from like five paragraphs. So you’re like, here’s five paragraphs. Make me a really flashy website. And it goes, boop, boop, boop. And it’s awesome.

And then you’re like, OK, I have an app and I want to just move this one thing over here. And it goes, all right, I did it. And you’re like, no, you didn’t. Hey, you lied to me. know, like, I didn’t know you could lie. And it’s like, I can’t see. So how could I know? And it’s like, it doesn’t have a browser. It can just assume that that was the result. But I don’t know. Some of these things are getting really good so fast that maybe we won’t. Maybe we will be agent orchestrators and not.

Brittany Ellich (24:52.897) you

Adam Argyle (25:13.296) Brick layers anymore, you know, so I know a lot about how to put bricks down which makes me good at orchestrating agents that place bricks But my job my job feels like it’s changing a lot all the time like every month I feel like it’s shifting less from craft more towards my domain expertise making me empowered in an LLM space and we’ll see how long that lasts I don’t know maybe there’s a

an end to this or it’s just maybe we’re at the beginning. Some people are like, we’re at 0.1 % of AI’s potential future. And I’m like, dang, we got a lot to go. And then other times I’m like, I don’t know, maybe these models are gonna max out. Maybe we’re at 90 % the capability of just the word vomit machine, because that’s all it does. Next word is most likely this. This is the, it’s not intelligent. It’s just an autocomplete. It’s a really, really fancy autocomplete. And so maybe that has a cap.

or maybe all these agent harnesses and the AI things that we put around the word machine continues to make it look like it’s really smart. And those give, I don’t know. Anyway, I’m confused. Hopeful, but also not hopeful. I’m like, I liked my job. I liked being a crafter. And that’s why I still have side projects like the color input, like that uses web components and signals. You you don’t, not going to get that from an LLM. Anyway.

Brittany Ellich (26:34.123) Yeah, yeah, it’s a weird time.

Erika (26:38.094) And like that human element too is never something that can be replaced. Like I have had the experience on the internet of going to a website where it’s clearly like a work of art or something that somebody really cares about where they have crafted an experience for you and then built it versus sort of a website where it’s clear that it’s pretty like cut and dry or you know, it’s there.

and kind of generic. And yeah, I mean, I think those sites will always exist that people who care enough will build them. But yeah, I I don’t even know if it’s changed necessarily from a corporate perspective, how much they care, unless they can monetize that experience in one way versus another. It’s probably a little naive to say that.

big corporations have ever really cared about design from a deep human empathy perspective. yeah, I definitely care. And yeah, I think it matters a lot.

Adam Argyle (27:55.012) Yeah, noticing those details reminds me of factories versus handmade goods. Like when you have a handmade good, the flaws of the human are visible and present, or they’re not. And you’re like, holy, you did an amazing job on this, you know? And then other times you see a factory one and you’re like, I see the mold seams. Wow, you made this nice and quick. That feels like plastic. You know, that’s not natural. But what we have right now with AI is like a race to make the next factory that builds apps.

And you’re still gonna have people that make apps by hand and they’re gonna feel like that and they’re always gonna be more unique, just like homes. There’s a race to make the next template home and you see a thousand of them on a strip of roads and you’re like, that’s not great, but somebody needed money or whatever and like they found an efficient way to make homes, good for them. But I’m always gonna be the one that’s like, where’s the crafts home? know, where’s the home that the most dangerous home on the most dangerous locations? I’m like, yeah, that sounds great.

So yeah, we’re always gonna have the pendulum there to swing. People are always gonna be trying to make the fastest, cheapest version. And I’m stuck in the middle trying to do both, I guess. And that’s why it gets so awkward.

Brittany Ellich (29:06.815) Yeah, try to keep a job basically for until you we need to retire. I feel like, you know, yeah. switching gears a little bit, I, you recently went through a, an interview process and we have heard a lot about the frat processes right now. I’m curious, how, how did it go for you? Like how was looking for a job, especially considering you aren’t quite overcommitted, if you will say, I mean, you have done so much.

Adam Argyle (29:13.72) Yep.

Brittany Ellich (29:36.449) publicly available, like publicly visible within your career, know, podcasts and side projects and everything. Like, do you think that helped you at all? Do you think it made it more difficult in some ways or like how was finding a job for you?

Adam Argyle (29:50.392) Yeah, it was both. And I chatted on a couple podcasts talking about it, like when it was really fresh. And the kind of recap is a lot of times I would apply somewhere and they’d be like, we see your history, CSS. Do you do anything else? And I’m like, you didn’t read very far, did you? Okay, well, yes, I can. I can do a whole bunch of other things. So if you want, I can show you. And I got so tired of like, no one read a resume. They just didn’t.

Even though I put a lot of work into it. I even made a website resume I think you can still go to it, but it had scroll driven animations. You could print it It had print styles, you know, like I even had Web GL shaders on it at one point so I was like people need to be wowed at my resume So they’ll go there and see sparkles, you know It’s got sparkles and it’s wavy or whatever it was I did on there But I eventually took it off because it performed like crap on your phone. You’d be scrolling it on your phone. I’m just like that’s garbage

But eventually what really resonated is I just had to sit down. I’ve done so many interviews myself interviewing other people. And I know that I will, and I’m a weirdo. I always read the resumes, but I know that like a lot of people don’t have much time to review you and they want to understand you in an elevator pitch. And it makes sense. We’re all very busy. So you need to like not just have a resume and not just, think also having a website, my website nerdy.dev speaks a lot for me because that is a integrated social network feeling application.

that immediately proves to someone that, well, A, this is no template, never seen anything like this before. B, there’s server-side rendering and data communication happening with outside services. Like you can just see and feel right off the bat that when there’s cache, I’ve just, I’ve got it all on there. And it’s nice, it feels nice. It’s got shadows and stuff. But I made what was, what I called my sizzle reel.

And this went through a lot of my past projects, a lot of my very visual. And it also, I oriented my sizzle reel towards the job I wanted, which was I wanted to be a design engineer somewhere. wanted to take my ability to code and my ability to execute high fidelity and sell that as my, thing I would bring to your company. And so I made a sizzle reel that did just that. And usually people, and it’s a three minute reel of animations to music and they’re

Adam Argyle (32:02.831) You know, it’s quick cuts and all this stuff. Cause I’m just like, look, people just are not patient. They’re so used to tick tock. don’t even tick tock, but I’m like, I watch enough TV to know that the attention span of humanity is going way down. So I’m going to just oriented towards that. And so it’s punchy. It’s flashy. It was in your face. could barely even tell what’s happening in some of the animations. And you’re just sitting there like, this is all hitting me so fast. that boom instantly I got on top of lists right away because you could send that to a boss.

and bypass the entire situation. It was a URL. Go to this URL. Watch this video. we need that here. We don’t have anyone that does. Bring them in. So that was one of my biggest tips is you do still need a resume. Definitely have a site. Make something that’s very quick and easy to consume that encapsulates, which is so hard to do. Let’s just be honest, too. It’s hard enough to even write a paragraph about yourself, let alone make a minute long video about all your sizzled stuff. But anyway, just

get it done and I found that to be a modern way to kind of bypass a lot of the junk, communicate with people in a way that tailored to their patience levels that they were at. And yeah, yeah, make a sizzle, I called it sizzle, rizzle because I’m a, I just like, I’m a dork. So, and I could, I was like, I don’t work at Google anymore. I can, I could say what I, that’s why I’m at the whiskey web and whatnot podcast. Now I’m like, I can custom my podcast. Ha ha ha ha. And I’ll make a sizzle, rizzle. It’s not even words.

It’s like made up words, you know? Like, do what I want. yeah.

Brittany Ellich (33:35.681) Yeah, that’s incredible. That’s a really great idea too, I think, especially if you do something that’s very visual. But even if you don’t, I mean, even if like your sizzle-rizzle is something that’s like showing, you know, these cool like performance back end, like, you know, everybody loves a graph. Everybody loves seeing a graph, see a graph go up or a graph go down. That’s exciting to see.

Bethany (33:56.228) Yeah, I’m here thinking what it would look like for a backend engineer. like, hey, look at this number. Now look at that number.

Adam Argyle (33:56.399) Absolutely.

Brittany Ellich (34:02.849) Whoa, look at this postman response.

Adam Argyle (34:03.395) Watch this migration script run. Yeah, I mean, it really could. Successful migration one, successful migration two. All right. And watch the pod. Actually, you could do a cool thing with Kubernetes or whatever. Like watch my pods light up. Here comes traffic. Here comes me thrashing the traffic. Watch everything get handled. no, one of them went down. Guess what? I got a backup. Yeah, you could make it fun. You just need a radio host to say what’s going on in a fun way if you’re a nerd. Anyway.

Bethany (34:05.828) Yeah

Brittany Ellich (34:30.463) Yeah, might need some help explaining exactly what you’re seeing. yeah, no, that’s really interesting. It seems like a lot of the process too is very broken right now on like just getting in the door to have the conversation even to about like about the role. Like, I mean, there’s everybody making AI resumes and then there’s everybody using AI to review resumes. And there’s just like, seems like it should be a very easy problem to solve where it’s like we have humans that need jobs and we have people who need to hire humans. So you would think that that would actually be.

easy to solve, but it’s not at all. So that seems like a nice way to…

Adam Argyle (35:03.491) Yeah, that was another part of the problem was even just getting into the interviews and being tested over dumb stuff that didn’t apply. So they saw my sizzle result, right? And they’re like, we want that on our team. And I’m like, cool. They’re like, come in for the interviews. The first couple of interviews, you know, were like, non-technical. Then we get to the technical interview and they’re like, solve this thing. it would be a, maybe it would even be UI related. Like I had a couple that were good in some make a notification system. Cool. You can pick whatever framework you want to make.

The notification system in cool. I’m just gonna do an HTML CSS JavaScript one right here. you didn’t choose react You just told me I could pick what I want turd nugget. Okay. Well, whatever here we go So I’m just I’m already I guess losing points because I didn’t choose react or whatever and then I’d be making it and I’d be like, okay, I’m gonna use the output element so it announces itself to the Screen reader users and they’re like now you don’t have to worry about that. I’m like, yes, I do

because that’s kind of what you’re hiring me for. It’s not just visual animated UI, it’s the UI that is useful and usable by everybody. So I think it is. then even animations, I’d be like, okay, I’m gonna spend the next couple of minutes animating it. And they’re like, no, we just want you to complete the task. I’m like, no, think you want to hire me to make, because you can probably make a crappy version of this that no one likes with React. I’m gonna make one that people like with whatever, I don’t even care what tools, but I’m gonna make a desirable one. And they’d always be like,

You didn’t pass the test. I’m like, I could feel that through the interview. Yeah, I was trying to show you the things that I bring that you don’t, and you clearly made me feel like you didn’t want them. So yeah, I could feel the mismatch. And then yeah, using AI in interviews too. Some of them, like open AI, you’re not allowed to use AI in the interview. And I thought that was quite funny. I was like, I’m sure you’re going to require it on the job, but you’re not going to.

I’m not gonna interview me and see if I even know how to do it. Okay, that’s fine. Meanwhile, the Shopify interview was like, full on AI, whatever you wanna use. Are you gonna one shot the test today? And I’m like, I’m not gonna, it’s someone else. And they’re like, yeah, someone last week one shot tic-tac-toe. I’m like, that doesn’t sound like a good interview. I’m gonna go ahead and just think this out. I’m also gonna make one that’s acceptable. Watch the keyboard. We’re gonna be able to play tic-tac-toe from the keyboard, you know?

Adam Argyle (37:26.125) So yes, it’s broken from the testing. It’s broken from, they filter people in with these, you know, criteria that they want you to know, and then they don’t test you on any of it. It’s super broken in tons of different ways. And so it was not fun. And me being a public figure didn’t really play into that much other than the job I got at Shopify was because I, well, that was a previous coworker was Jason Miller. him and I worked together on Chrome and.

He knew I could build robust UI. He wanted me and him to build projects together. So it’s a little bit different. If anything, being public might hurt you because people create assumptions about you. And sometimes, you know, if you can’t build, become a teacher or whatever the phrase is that people like to use to be rude. I’m like, no, I teach because I’m building all day. The only reason I have anything to tell you that’s a tip is because I was building last night. So turd, more turds. There’s turds everywhere, dang it.

Why are there so many turds?

Brittany Ellich (38:27.027) It seems like that needs to be the theme of this entire episode, really. Interesting. Yeah, I think also I feel like a lot of the way that…

like hiring has gone trending up until, you know, AI was like, we’re just going to make, take these leak code problems and make them more and more difficult and stuff. now they’re like, how, how are we going to read completely redo and mess up the interview process again now that AI is available to solve these leak code problems? Super, super easy. And like, you don’t need to, you don’t need to do, know all of these obscure algorithms anymore.

Erika (39:05.036) Yeah, I guess like, how would you design a front end interview? Like if you were interviewing yourself for a job, like what would you expect on the other end of the table?

Adam Argyle (39:16.117) smiling because Shopify asked me to do that. They were like, hey we want to hire more people that bridge the gap. want X’s not T’s. They used to hire T’s and that was a T-shaped individual. You’re nodding your heads like you know what that is so I don’t need to explain it. Okay, the X-shaped individual though, know, they’re the cross, so that means you can go cross domains versus the T means you could go deep into domain. So they used to hire for

intense technical knowledge and now they’re looking for cross because AI can help fill the gap of a lot of technical stuff. And they want someone that can do AI things. want more design engineers. And the test that I made, because you only get an hour with someone, and a design engineer, tend to think of them as the people that show up and they finish. I like to think of them as the finishers or they are capable of finishing. They can start from scratch and go all the way to like,

polished finished thing. A good example is like a faucet. So you’ve got all these people that help build a bathroom or a room that has a sink and they can put the sink in and they can do it without it being seamless, right? They can have gaps and someone has to eventually go in there and smoothen out the caulking and stuff like that. Another feature of a sink is when the water comes out of the faucet, there’s like this little aerator thing that’s on there that makes it so it comes out and looks like a cylinder, but it has like air in it. It’s like all bubbly.

Anyway, if you don’t have one, you’ll know because you’ll use one. like, it feels like I’m at the beach using just the crappy bathrooms. know, like the sink is all, just the water just looks, anyway, it just feels different too. And so someone needs to show up and give a crap about that. And so like, how do I create a test where it’s like, look, I don’t need you to put the sink in. I don’t need you to run the water to the plumbing. I need you to make this sink nice though. So do some finishing work. And along the way, in terms of like,

front-end web development, let’s do some animations, let’s make sure it’s keyboard. If we get to accessibility, that’d be great. And so what I built was a front-end interview scenario where I’ve already built a lightbox for you with popover, actually using dialogue and some images. So it’s a grid of images. If you click one, it clones the image into the dialogue and shows the dialogue. And then when you hit escape or you click off of it, it closes the dialogue and that’s it. So it’s like,

Adam Argyle (41:39.331) The plumbing is there, the interaction is there, it’s just lame. It just sucks. And there’s so many different directions to go. There’s so many ways to improve it. It comes with some accessibility affordances. It comes with some keyboard affordances and stuff like that. So I get to watch people first figure out what is not working. And then they’re like, I need to go fix that. Like, I expect this to do that. And then once they get the kind of

general stuff working, can go add some polish and some finesse with animations and stuff like that. And it’s on them to figure out how big are the animations that they want to do, all of that stuff. So they get to show up and instead of me watching them plunk out HTML for a grid and then plunk out a dialogue element and then like look up the features and stuff like that, it’s more like, no, come here and make this thing polished and awesome. The fun part, someone else, anyway, so that was the test I made and it’s, you can use AI the whole time.

we have ways that we’re, critiquing your AI usage. we’re, you know, we’re scoring you on a number of things. there’s different phases. They wanted it to be like a video game, which is fun. So there’s basically like levels, that you have to beat. And it’s like, Hey, what level did they get to? they got to level five. Ooh, that’s a pretty advanced level. That means they had done all these other things. and you can continue challenging people be like, Hey, you just passed level four. Here’s level five criteria. Let’s see what you do with these or whatever. so you need to guide them.

towards something else. That’s what we did. And I thought that was a fun and interesting and meaningful way to engage someone’s finisher mentality. Instead of us watching them do the boring part, they get to jump right to the fun part and show off. Literally, I wanted them to spread their wings. It’s so annoying that in one hour, you go to these interviews and all you do, and everyone, can’t see what I’m about to do, but I’m gonna start spreading my wings and then they get stuck.

on the walls because you only had an hour. You’re like, I swear I could have done better. You’re like, didn’t even get one flap in. You’re like, that sucks. I wanted something where it comes in. It’s like, it’s already flapping, but now I want you to like, yeah, make it an elegant flap. Actually, that metaphor got really weird, but you see what I mean? Like, I wanted you to feel free and feel like when you leave the interview, feel like you had an opportunity to actually show something unique about you that showcased your skills, not your ability to.

Adam Argyle (44:03.022) tippy tap, but your ability to execute and deliver something polished and great.

Brittany Ellich (44:08.243) that and that seems like something that would like give a lot of signal too and like shows a lot about the company as well like I think a lot about the way that a company interviews and what that means about like the company’s internal culture as well like if they’re just doing the hardest algorithmic problem as possible like that’s probably not a place that I’m gonna want to work anyway so yeah

Well, we are coming up on time and at the end of every one of our episodes we do what we call a fun segment and since you are our guest today, of course it is CSS themed, and we came up with CSS would you rather? So I am going to go through a list of these would you rather questions. All three of you are welcome to answer because we might have different opinions here and you might get shamed for your opinion or you might not, you know, we’ll see.

We’ll see how goes. I don’t think Adam is judging us, but maybe just a little bit. And that’s also probably, that’s probably fine. So we’re gonna kick it off with this really fun, hopefully not controversial one. Would you rather write vanilla CSS forever or be forced to use Tailwind on every project for the rest of your life?

Adam Argyle (45:22.6) Easy vanilla CSS all day. All day.

Erika (45:25.398) Yeah, I’m right there with you. This is kind of a no-brainer for me.

Bethany (45:31.629) I would do vanilla CSS as well. As someone who does not write CSS well, I still do not love Tailwind at all. Respect to people who do, but not for me.

Brittany Ellich (45:46.178) Yeah, Tailwind came out after I learned CSS and so it’s actually more difficult, I think, for me to learn to like be like, no, this is a Tailwind application. I need to go figure out what this looks like there and lot of looking up. Okay, this one might be.

Adam Argyle (46:01.71) I’m cursed because I know everything it can’t do. So I’m like, oh, I needed to do this thing. Oh, it’s so much easier over here. I’m like, you have to put that class on those 20 items. I’m like, I just have one selector do it over here. Anyway, so yeah, I’m just like cursed. I’m just like the whole time using Tailwind, like this feels so limiting, which is part of its value. It does have some superpowers, so Tailwind is still cool. But if this is for forever, oh, vanilla. Yeah, I want to grow with it, not be stuck.

Brittany Ellich (46:14.23) Mm-hmm.

Brittany Ellich (46:21.567) Yeah, it’s true.

Brittany Ellich (46:30.069) makes sense. Next one, would you rather flex box or grid forever if you only got to pick one? wow, that was fast. Okay.

Adam Argyle (46:36.12) Grid, grid. People flexbox too much. Grid is awesome. It has a big API surface area. It’s hard to memorize. It’s got some weird units. It’s also got a lot of functions. But it’s an awesome, awesome layout engine.

Brittany Ellich (46:57.963) Do you have a?

Bethany (46:58.371) Yeah, I probably would do Flexbox just because that’s what I mainly know and have used, but yeah, I would say Flexbox.

Erika (47:09.708) Yeah, I’m on the grid train. I feel like I’ve gotten to some weird states with Flexbox and I have like boxes on boxes on boxes and I’m like, well now I just made a grid. So why didn’t I just use grid in the first place?

Brittany Ellich (47:25.173) Yeah, I agree. think grid feels a lot more powerful and like the more that I do write CSS, like that’s probably the one I would choose to start with. But I almost always reach for Flexbox first because I’m like, I know exactly how to do this with Flex. Like this is easy until it’s not until you’ve got, you know, 10 layers of nested divs. And yeah, like, wait a minute.

Erika (47:45.739) Exactly.

Brittany Ellich (47:49.574) Would you rather every browser wait to ship with new CSS features simultaneously and perfectly or browsers keep competing and shipping things fastest to see who can get it?

Adam Argyle (48:03.277) Ooh, spicy. I like this one. The interop efforts have been quite nice as things come out altogether, at least in the same year. However, the competition’s mostly in the tooling, which is nice. Not usually in the implementations. Mm-hmm, I am going to go for personal convenience.

and say if they all released at the same time that would be quite dandy.

Bethany (48:36.411) you

Erika (48:37.912) I’m not sure I have a preference on this one. I feel like I’m not deep enough in the browser compatibility. I don’t think I do enough CSS to really feel that feeling much. So I’m going to abstain from an opinion on this.

Bethany (48:57.275) I would say honestly I think I would rather them compete because I still think people are going to be on ancient browsers or ancient versions of it. So you’re always going to have to do some sort of backtracking. So I think like just having people compete and get it out sooner is probably probably more or less going to be going to be the same.

Adam Argyle (49:22.381) Cool answer. I like it.

Brittany Ellich (49:22.465) Yeah, yeah, I agree. That could end up being some sort of like lowest common denominator. Like now we have to wait for everybody to get it out and then it never comes out. Luckily Internet Explorer is out of the picture, but I’m sure there’s, you know, a new one to hate. It’s been a while since I actually had to care about browser compatibility. Would you rather lose key frames, no more CSS animations or

lose CSS variables.

Adam Argyle (49:59.252) Hmm. I’m gonna say keyframes. Keyframes have a very annoying thing that they do to me, which is they’re not interruptible. And that kind of makes them dead to me in a whole bunch of ways. Like, it makes them way less useful. They have a whole bunch of really fantastic use cases. Like, they’re in scroll-driven animations, so you couldn’t do a scroll-driven animation without keyframes. That stinks. But I think they could mostly…

be replaced with something else and I wouldn’t miss them. But variables, variables are a clutch. That’s a really, really powerful way to, well, they just have so many superpowers. I could rant about what you can do with variables. have many blog posts about, I think there’s even a blog post called 10 Powerful Things You Can Do With CSS Variables. And most people only do the first one, which is like they make a variable with like a value in it.

Like we scrubbed the internet and saw everybody’s variable usage and it’s like, how many levels deep do they go? Deep meaning, do you have a variable reference another variable? And that was very rare. It was like 10 or 15 % of websites actually have a variable reference another one. And in that post, I’m like, you can have a variable reference, a variable reference, a variable reference, a variable, know, like look at the cool use case. can do this. So yeah, variables all day. Plus I have open props. I’m like a variable hoarder. That’s, what that whole repo is. It’s like, let’s just.

find all the best variables that you could have in the world and put them in one repo for people to use. And so yeah, I choose VARS.

Brittany Ellich (51:29.441) Any alternative opinions, Erika or Bethany? We’ve been in GitHub UI, anything. Bethany is very much on the back end, I feel like, within your role. I haven’t gotten to do as much UI as I would like to. It’s been a while. But this has been fantastic. Thank you so much for joining us. Where can folks find you on the internet?

Adam Argyle (51:58.281) Good one-stop shop is nerdy.dev. That’s my personal website. I always post there first There’s RSS if you’re into that and then I syndicate out to social networks and you’ll find my posts on Macedon, Blue Sky or X if you want them But yeah, find me at nerdy.dev. That’s the good spot

Brittany Ellich (52:16.897) That’s awesome. Yes, I’m very into RSS. Love that that is making a comeback now. Excellent. Well, thank you so much for tuning in to Overcommitted. If you like what you hear, please do follow, subscribe, or do whatever it is you’re supposed to do on the podcast app of your choice. Check us out on Bluesky and share with your friends. Until next week, goodbye.

Adam Argyle (52:23.874) Me too.