Some thoughts on AI’s impact on digital design and development

I drive my kiddo to school in the morning, and my 12-minute return trip home is a fruitful time to think and prepare for my day. I had the realization that my little drive home is a nice window of time to share my thoughts, so I’m exploring what that might look like! This feels like a fun and natural thing to try out, and I’d welcome any thoughts and feedback on this new experiment.

I woke up this morning with AI on my mind, and a quick glance at my Twitter feed made it clear that it’s on a lot of other people’s minds as well. So I used my little commute to explore the topic of AI’s potential impact on the world of digital design and development. Here we go!

Here’s the (cleaned up) transcript:

Hey there, this is Brad Frost, and this is a brand new thing called “I’m a dad that just dropped his daughter off at school, and I have about a 10-12 minute commute back to my home and about to plug in in my office”. But I was like, “Man, 12 minutes is kind of a good amount of time to be able to dig into something. I’ve been desperately wanting to share more. And thanks to the Ol’ Musky Twitter situation, I’m finding myself going, “Well, I’d like to operate in maybe some other spaces, but also my brain moves a lot faster than my fingers. So I’d love to kind of have this time to be able to share things of interest to me: professionally, personally, talking about web design and development, talking about creativity, talking about art, design, music, mental health, wellbeing, spirituality, and other consequential matters.

I wake up in the morning with a bunch of things on my mind and, and I’m like, “You know what? Maybe I should do this.” I promise I’m being safe. I am not looking at the camera. I’m keeping my eyes on the road. It’s a pretty straight shot from my daughter’s school home, so hopefully I could walk and chew gum at the same time. So anyways, that’s what I’m gonna do.

To kick things off, I woke up this morning with AI on my mind. I actually was interviewed for a podcast yesterday where we got into that a fair amount. And so I want to pick up on that. This podcast is gonna be published I think next week, which I’ll definitely make sure to share whenever it comes out.

A lot of [the podcast] was talking about how AI is going to impact the world of web design and development, design and development in general, and some of the really big cultural and societal ramifications of this technology. And I woke up this morning and naturally checked my feeds, and I feel like every other tweet was about GPT4 and all of these really fast-paced, breakneck-speed improvements to these technologies.

I’ll admit, I feel like I’m pretty measured when it comes to technology. You know, I’ve seen tons of trends come and go, things like blockchain and web3 stuff. It’s a “there it is and there it goes” kind of thing. I’ve always kind of been more foundational in my thinking as far as what I feel are important technologies. I still think that betting on the web is a good bet, and there’s a lot of other technologies and things that come and go that just kind of haven’t stood the test of time. This, to me, AI really does feel like the first thing that has really made me go, “Oh yeah, this is it. This, this absolutely is going to change the very nature of the work we do.” And I am simultaneously excited and concerned about it. And I think that it’s important to be able to hold both of those thoughts in your head at the same time, because like most things it requires a lot of nuance.

We need to be talking about the good and the bad in an open and honest way. And so that’s what I’d like to do here is just pick at that a little bit. So I wrote a post about how design systems might be impacted by the world of AI and the most commonsensical and easy use case is that you are able to take a whiteboard sketch of an interface and kind of teach AI to learn your component library and basically take that sketch and turn it into a real thing. And sure enough, just this morning I saw somebody do exactly that, right? They took a little picture of a sketch of the webpage and it splat out some markup and and styles and whatever.

And, and that’s that. I think that that’s amazing. I think that’s genuinely amazing and it gets really to the heart of what it is we do and how we do it. And I think that it’s inevitable. I think that it’s a matter of time. I don’t think it’s going to be mature and integrated and implemented in large organizations around the world like next month. But next year? Certainly in five years. So when we talk about what we’re doing with all of this, I think it’s increasingly important to decouple the what and the how of how we go about creating digital products from the why and making sure that it is sound.

I think that those are, those are little kind of different equations. I think that the what and the how are going to continue to be just eaten away at by AI but also just efficiencies in general. I’ve spent a lot of time in the world of design systems, and that’s kind of an express goal, right? It’s like, “Hey, you don’t have to go and create that card for the 17th time. You don’t have to build that button from the ground up. You don’t need to hand craft every piece of markup or hand draw three versions of the same design comp in Figma.” That’s waste. That’s not a respectful use of people’s time. So design systems are about efficiency, and we’re all trying to become more efficient,

I know there’s a lot of like, Taylorism, industrialism, blah blah blah; I’ve seen a lot of commentary and I’ve commented on that myself, like where you could kind of get into the capitalist wheels of it. But let’s get at it from like a really bedrock level. We want to make great digital things and we would like to do that as quickly as possible. I would say to the people who are pushing back against that, let’s maybe challenge that a little bit because he thing that’s important when it comes to increased efficiency is that there’s there’s the risk of reduced quality.

I think that this is going to be kind of a big shift in our field, a big shift in our industry. If the means of production for creating digital products are increasingly assisted or led or maybe even taken over completely by artificial intelligence, I think that the role of the designer and developer kind of becomes one of quality control.

I see design and development stratifying into the beginning of the process, meaning you’re on the creation end of these processes. You are the ones crafting the design systems and making sure that the raw materials of these tools that AI wield. I think that there’s still going to need to be a fair amount of human talent involved in and kind of making sure that the tools that we’re using to build these things out of are solid, are accessible, are performant, are resilient, are ethical, and all of that stuff.

And once those toolkits exist, then slapping them together and composing them I think are increasingly going to be handled by non-design, non-technical folks. And it already happens in the world of like CMSs and stuff. You have a bunch of marketers who wanna just splat out a hundred thousand different landing pages for things, and they don’t necessarily know how to design and develop, but they’re able to log into a CMS and spin up a new page and publish that.

And I think that trend is going to continue where you’re going to have a bunch of non-technical folks wielding AI saying, “You know, as a business stakeholder, I need this new website that has this hero unit that says this and calls to action and drives to this page, and then has a grid of three cards that have these three things that are featured there and go grab some of her product imagery from a database, and of course a header and a footer. Please generate that for me and wired up to the appropriate systems.”

It’s not that far of a cry to really envision that thatcan happen, and I think can happen quite easily. It’s really a matter of time, right? So I think that’s it. So back to the stratification. There’s the people who will focus on the creation of the raw materials that then get assembled in myriad ways. And then I think that there’s the other end of it, which is, “Okay, some marketer plus AI just came up with a brand new thing, and now we really need to validate it and make sure that it is not evil to make sure it’s ethical first and foremost” That it’s not garbage, right?

So how do we make sure that these things are built properly and soundly, are accessible, are performant, are doing all the things so effectively, many hands I think will need to become more brains that are checking against what this AI-assisted or AI-led work is doing. So there are these bookends of people creating things that will get used in a bunch of ways, and then people on the other end that really serve as a safety & sanity check and as a QA check for the stuff that gets produced with it. And I think that’s not that far away. I think that all the building blocks are here at present to make this happen.

Especially in the word of like digital interfaces. We’re not talking about anything particularly exciting: it’s forms, it’s marketing sites, it’s dashboards. A lot of your standard fare interfaces are not the most wowee-zowee, doing triple back flips. They’re just kind of helping people along the flow or whatever. So I see that as an inevitability.

Now, I’ve talked about the pros, which are, “Okay, this is efficient.” And you can go from like a concept to (with that human guidance) something that is a sound digital product coming out the other end. And rather than taking months and months, it might take weeks or days or hours or or seconds, who the hell knows? And again, that’s inevitable. I see that as something that we should be actively encouraging at the expense of other things, namely quality and just basic ethical judgment. But I think that that’s the trajectory it’s on.

The downsides of all of this are obviously massive. I can’t help but think about the entire economies that are built on basically farming out a lot of rote development tasks or design tasks, just like feature factory kind of churn. We work with a whole slew of organizations that employ many different-shaped sizes and flavors of offshore development. But also a lot of not even offshore, like a lot of teams are effectively doing this kind of composition work today, right?

They are the ones saying, “Okay, well, I need to put this homepage together. I’m taking these pieces, I’m stitching them together, I’m wiring them up to some backend system.” And I think that means of production becoming more efficient and more led, or at least guided by AI is a real genuine existential threat to large swaths of people producing digital work. And that’s terrifying. That’s utterly and completely terrifying.

I think that what is terrifying about it is that we don’t have the social structures in place, certainly not in America, where human value is so coupled to the amount of work that we produce. That work is being encroached upon by AI much in the same way that self-checkouts at grocery stores and driverless cars [are encroaching on jobs]. (Actually, forget about driverless cars; that’s still a little bit of like a future thing.) But self checkouts at grocery stores is a great example of “here is this thing that used to require a bunch of humans to make this thing go.” And now machines take care of that and it seems to be working. A lot of people prefer it, and that means that there are fewer jobs for cashiers. And we used to think as knowledge workers that we were exempt from all of that, but here comes these things that are just totally going to eat our lunch.

Going all the way back to the Industrial Revolution and what that was supposed to unlock for us, which was we just would have more time. We’d have leisure, we have the means of economic production, the machines and now the computers churning along and producing value, and we just get to kick our feet up and live happier, more fulfilling lives, not focused on drawing rectangles or sitting behind a till at a grocery store dealing with irate people all day. And that’s great and that is the ideal, but we are not set up as a society for that. Politically we are like worlds away from that, again speaking from an American perspective here.

And I think that’s the biggest risk because at the end of the day, if we were just able to snap our fingers and create a bunch of sites and apps and have those things work and make information more easily accessible and build new things for us, that can be great. And again we get to kick our feet up with a light beer and just spend time with our families. That’s great, but so long as our ability to feed our families and pay our rent and pay for healthcare is tied to what we do for work, [AI encroaching on work] is just a gigantic and huge and urgent threat to a lot of people’s wellbeing.

And of course, I’m coming at this from the lens of design and development, but we’re talking about entire industries just being totally upended seemingly overnight, certainly faster than the wheels of society move.

So anyways, pros: we get more efficient ways of doing things. We get to take some of the rote and repetitious and — I don’t want to say thoughtless — but less intellectually-challenging work off of our hands, where we are able to just stitch these things together and spin them up faster and do all of that. Cons: I think that there’s a lot of risk that a lot of people’s jobs are going to be impacted by this.

And of course, then there’s just also the more existential stuff around AI in general, around what value system is AI operating by? What are its blind spots anddo we even like know where these things are are going or how they even work? I think that [those questions] are really important, and there are far smarter people than me that are dealing with and talking about those things. But I think that the cons are many and the cons are massive.

I also think that in our discourse today, these things are being either vilified or deified, and we ought to be having real open and honest and nuanced conversations about all of this. These things aren’t going away; you can’t put these genies back in the bottle. So it really becomes a question of how we approach this in a very eyes wide open way to mitigate the risks and promote healthy use of this stuff.

How do we do this in a way that balances the efficiency and the superpowers that the stuff can give us with a real humanity? How do we do this in a way that preserves the integrity of humanity and how do we not just use the efficiency to bulldoze and steamroll a lot of people’s livelihoods? And that’s especially for the people who have the authority and the ability to green light something like this.

Um, yeah. So these are things that are on my mind after dropping my daughter off at school and heading home to get to work. These are things that I just kick around in my brain all the time. Not all of them are this big and consequential; I’d like to talk about more fun stuff too.

So this is an experiment. We’ll see if I keep at it. I would really welcome feedback and ideas around it. I would really love to have more of these types of conversations, and I’m exploring a bunch of different ways of making that happen, and I’m really excited about some things that are on the horizon. So here we go! I’d welcome your feedback on this. What should I be doing more of or less of? Or should you say, “Brad, stick to your lane” hahaha. All right. Hey, take care. I hope you have a fantastic day.

I’m hoping to continue publishing stuff like this, so if you think this is interesting feel free to subscribe to my YouTube channel and my RSS feed.