Font AwesomeStrange things are afoot at the Circle K.
There’s a version of this post where we could tell you AI changed everything overnight and that we’re never looking back. The thing is, that’s not what happened.
What actually happened is slower, and we think, more interesting. At the moment, we think AI use is more about intent, and what it means to build something for other people, not just crank out high volumes of slop.
We Didn’t Come In As True Believers
When AI tools started becoming more common for devs and designers, it’s not like we were in a huge rush to embrace them. But we were curious, aaaand also a bit skeptical. Travis put it plainly: “It’s not necessarily the world I would have envisioned or I would have picked.”
That’s an honest thing to say out loud, and a sentiment worth considering. This is a little Captian Obvious to say, but there’s a real disruption happening. The things a lot of us have gotten really good at (like the craft of writing a function, solving a gnarly UI problem, debugging until 1am because the thing just wouldn’t work) are changing. Some of them are getting eaten up by AI.
But rather than become doomsday prophets about how we’re all getting rolled over by the AI Overlords, we’ve decided to stay “curious and reasonable”. It turns out, curious and reasonable is one of the three core behaviors we try to practice at Font Awesome, and that means we don’t jump to a dogmatic conclusion before we’ve actually spent time with something. But it also means we reserve the right to change our opinions when the evidence changes. So, we approach AI tools with some healthy skepticism, but if the tools get better, and it opens more opportunities, we’ll acknowledge that as we continue learning.
So we ran an experiment. We called it Build Week.
Build Week: What We Actually Found Out
At our most recent all company Snuggle, everyone on the team came in and spent a week building — and not necessarily on Font Awesome projects. Nobody was under any particular pressure, the goals was just to build using AI to see what happened.
The results surprised us. Not because the AI did everything for us, but because of where it helped most. Most folks gravitated toward their areas of deepest expertise and used the tools to reach further around the edges. Security folks got more thorough security. Designers got more precise design. Developers finally got to clearing out the decade-long backlog items that never made it into a sprint.
Dave Gandy, who spent the week building Color Awesome (a tool he’d wanted to make for years) put it this way: AI helps you be more of who you are, not less.
That was the unexpected part. With AI, there’s often an assumption that it would flatten everyone into generic output. What we found was almost the opposite. When you know what you want and you understand the domain, the tools can amplify your judgment. The stuff that fell out of the week felt specific, personal, and fully ours. Dave’s color palette visualizer ended up looking like stained glass — because he’d just been to the Sagrada Família and that’s where his head was. The tool wasn’t a replacement for true taste. It gave the taste somewhere to go, but faster.
The Question That Actually Matters
Here’s a question we kept coming back to: What are we doing this for, and why? That may sound simple, but it isn’t.
For example, if the goal is to learn to code, leaning on AI to write all your code is a bad idea. You’ll ship stuff you don’t understand and can’t maintain. Kind of like how a student might use AI to write every paper isn’t learning to write, rather, they’re just checking an assignment off a list.
But if the goal is to build great software for real people, that’s different. The end user doesn’t care how long it took you to write the function, they care about whether it works, whether it’s fast, and whether it gets out of their way. And if these tools let a small team ship five or ten times more of the right things for the people they’re serving, that’s not cheating, it’s focus.
That framing shifted something for us, because the best engineers aren’t the ones who write code in the way that’s most fun for them. They write code in the way that’s most useful for someone else. AI, used right, is just another way to stay focused on that goal.
More Power, Harder Choices
Here’s the part nobody warns you about: when you can build almost anything, the hardest skill becomes knowing what not to build.
Travis framed it as a quality problem. When you can produce so much more output, the bottleneck of workflows shifts to review, to make sure what ships are actually good, reliable, and useful. The speed is real, but so is the responsibility to stay user-centric, and useful.
Dave made the same point from a product angle: more features aren’t better. Anyone who’s ever moved to a new house knows what he means. Every box you carry down the stairs is a feature you have to maintain forever. It’s the right stuff that you’re looking for, not more stuff.
So we’re thinking harder about discernment. Who do we talk to? What do our customers actually need? What does simplicity look like when you suddenly have the ability to add … anything? The ability to say no confidently, quickly, and on purpose has never been more important.
The Concerns We Take Seriously
We’re also not going to skip past the hard parts of the conversation. AI raises real questions. Energy and compute costs are real. The ethics of training data — what was scraped, from whom, and under what permissions — are real. The effect on jobs, on what it means to be a developer or designer, on the skills people spend years building, that’s real, too.
Travis put it honestly: you can’t put the genie back in the bottle on some of this. Maybe we get better moving forward, but we also need better governance. And we definitely need to keep having the hard conversations rather than pretending they’re settled.
Dave’s take is that this is the same story technology has always told. You can do remarkable things with powerful tools, and you can do terrible things with them too. The question has always been intent and character, not capability alone. All capability and no character is, as Dave put it, just a supervillain. Zod doesn’t have a product strategy, he just has power and no guardrails.
Character has to come first. That’s true for people, and it has to be true for how we adopt these tools at Font Awesome.
Why We’re Still Here, Still Curious
Here’s what we keep coming back to: the people who get hit hardest by major technological transitions are almost always the ones who refuse to engage with them at all.
That’s not a reason to abandon your judgment. It’s a reason to use it. Stay curious, and ask the hard questions. Don’t just accept the hype, but don’t perform skepticism as a substitute for actually figuring out the right thing to do.
We ran Build Week because we wanted to start poking around and start formulating opinions, and find a way forward. What we found was that these tools, used with clear intent and real taste, actually help people become more of who they are. That was the surprise, and that’s why we’re staying curious.
Obviously, this conversation will look different a year from now. And it will probably be different a month from now. But the framework we’re working from, to stay curious and reasonable, humble and helpful, adventurous and dependable — that’s all the same.
Now, go forth and go make something awesome!
This post is based on conversations from Episodes 1 and 2 of Podcast Awesome with Dave Gandy and Travis Chase. Listen wherever you get podcasts or at podcastawesome.com.
