Why Figma CEO Dylan Field is optimistic about AI and the future of design

We’ve got a fun Decoder today. I talked to Figma CEO Dylan Field in front of a live audience at South by Southwest in Austin, Texas, and we got into it: we talked about everything from design, to software distribution, to the future of the web, and of course, AI. 

Figma is a fascinating company; the Figma design tool is used by designers at basically every company you can think of. And importantly, it runs on the web — it’s really easy to share Figma designs with other people and work on them together. That sounds obvious now, but it was a revolution in design when Figma launched in 2016, and it quickly grew into a serious player in the creative software world.

So Dylan and I talked a lot about where Figma is now as an independent company. About 4 percent of Figma’s employees took a buyout offer after the deal fell through, so of course, we talked about how Figma is structured, where it’s going, and how Dylan’s decision-making has changed since the last time he was on the show in 2022.

Beyond Adobe, the two big topics I wanted to talk to Dylan about were AI and the web. Like every big software company, Figma is investing heavily in AI. But it’s also in the unique position of serving the creative community — a community that has strong opinions on whether new generative software tools are taking jobs while also creating a deluge of low-quality work. 

You’ll hear Dylan make the bull case for generative AI. He argues that more empowering software should increase the demand for human creativity and let more people in, and let those people work better and faster. He’s optimistic that even in a world of C-plus AI-generated content, both creators and consumers will push great stuff to the top. 

This topic dovetailed nicely with another we dug into: the state of the web. Even though the web is under a lot of pressure right now due to AI and the dominance of big platforms, Dylan is still a believer in its potential. After all, Figma is a web app, and Dylan and I discussed how he’s looking at the landscape of software distribution in 2024, especially as European regulation like the Digital Markets Act has taken aim at app stores and mobile browser restrictions to try to create a more level playing field for developers. 

We also talked about design trends, which is always a good time — get ready for something called solarpunk to take hold.

Okay, Figma CEO Dylan Field. Here we go.

This transcript has been lightly edited for length and clarity. 

Dylan Field, you are the co-founder and CEO of Figma. Welcome to Decoder. Welcome to South by Southwest.

Thank you.

I am very excited to talk to you. Figma is at the center of many, many, many big stories in tech right now, from design to building on the web to antitrust law. You name it, that’s you. So I just want to start with a very big question: where’s your focus these days?

I think that the way a lot of people think about Figma is as a design tool. And obviously that is where we started and where we continue to put a lot of work into making that better. But how we think about Figma has really evolved from just how do we make the best tool for software design to really supporting the entire process of how you build software. Everything from starting with an idea or brainstorming or figuring out where your team is at, getting all the way to code and production.

And we have three products that we offer: FigJam for whiteboarding and brainstorming, Figma Design for software design, and now we just launched in [general availability] about a month ago, Dev Mode, which helps developers translate those designs into code. And we, I think, have so much more opportunity to figure out how do we actually make it so that you have the most efficient but also best way to explore and bring more people into the process, going all the way from idea all the way to app or website or reality.

So that last turn is really interesting, that we’re going to look at a design we made collaboratively and actually turn that into production code. The last time you were on Decoder, you were at the beginning stages of thinking about that, now you’ve actually shipped it. In the meantime, there have been what I might call some distractions. How have you managed to stay focused on shipping ideas that big?

Yeah, so I think what you’re referencing as a distraction is — which I think actually in some ways created focus, but we can talk about that — was we tried to merge with Adobe, and various world governments didn’t see the same way we did. And so we’re back to being an independent company again. I guess we never stopped being independent. 

While it was certainly distracting to have our exec team spending a lot of time going back and forth to Europe and London and talk with regulators, I think we also tried to make sure that the team was as isolated from that as possible. And we knew what our road map was, we knew what our goals were, and we tried to just have absolute clarity about that. And what ended up happening was, as a result, I think we actually picked up our pace during this deal review period. Now we’re picking it up even more, and I think we’ve been accelerating through all that, but it’s been interesting and definitely we’ve had to push hard to make that happen.

Let’s talk about Adobe just for a few minutes. It’s been several months. The deal was called off in December. We’re obviously sitting here in March. Time has passed. I had the general counsel of Adobe on Decoder a few months ago, it was like the week the deal was called off. He wanted to talk about AI and copyright law, and I was like, “No, let’s avoid that bomb and talk about this deal being called off.” And he said, “Look, we understood that the regulators are not going to let us go through and we made a decision.” That’s his side of it. Talk about that moment for you when you knew that it wasn’t going to go forward.

There’s kind of like the Twitter meme of, “We’re so back, it’s so over,” and you kind of had those cycles through the deal process. And maybe the general trend was toward, as you saw, more thinking from the regulators over time, it was more clear that this had a less probability. But even toward the final months, there were these moments of, “Oh, this is going to go through,” and moments of, “Fuck, what are we doing?” And obviously at the end, there’s a mutual understanding of, “This decision has been made for us and let’s call it.”

But look, you’re only allowed to have so much coordination / collaboration and learning from the other sides in these deal processes legally, and you know all about that. But for me, I actually really enjoyed my time that I got with the Adobe team, especially [Adobe CEO] Shantanu Narayen. He’s, I think, pretty legendary and that was probably one of the things I was most bummed about is not having him as a mentor, but now it’s full steam ahead. Let’s go figure this out and be the best independent company we possibly can.

So the theory of the case from the Europeans was that Figma should be a meaningful competitor to Adobe. It is not yet, but it will grow into that. [Adobe executive VP] Dana Rao, I think he was skeptical of this claim. He’s like, “I have 800 people, we’re Adobe.” You have to make Photoshop. That’ll be very hard for you. Do you think Figma can now become a meaningful competitor to Adobe?

I mean that’s just not where we’re headed right now. Like I said, we’re really focused on the software process. How do we make it so that people go from idea all the way to app or website? There’s a certain amount of building, some amount of buying, some amount of partnership that’s going to go into that. We’ve got to figure out exactly how that all will work. 

But yeah, it’s a very different market and very different process than Adobe’s in. I think it’s interesting if you think about the future that way of what’s possible. Regulators were kind of telling us you can do anything, anything’s possible. And it’s like you kind of want to believe that because I really think our team is great, but at the same time it’s like, “Well, here’s our priorities, here’s our road map, here’s our focus. It’s all documented.” And it’s a weird conversation to try to disprove that you can do anything. But ultimately, that’s sort of the way that regulators think, and it’s flattering, I guess, in some ways.

I think telling a charming European bureaucrat that I can’t actually do anything must be very difficult.

Well, the reality is it’s all just tradeoffs, right?

How much time? What do you think the opportunity is? We have these very adjacent opportunities that we’re going after because a third of our users, for example, are developers and we know from our research that, pre-Dev Mode, we were not doing the best we could for that audience. And even now, we know that there’s so much opportunity ahead to make the experience of being a developer in Figma better. That is a near-term opportunity that there’s so much we can do quickly on. We try to go build Photoshop, that’s going to take a very, very long time. What exactly is the income or revenue? You have to work through that.

Let’s talk about that opportunity. I want to talk about what you see ahead, but let me ask what I think of as the Decoder questions so we can... This is the brand. All of you’re supposed to cheer when I say the Decoder questions. [Cheers]. There we go. I feel very proud that I branded some LinkedIn bait. Thank you for indulging me. About 4 percent of your folks took a buyout after the deal fell through, right? They were kind of waiting on this to be the big moment. It didn’t happen. How many people is Figma now?

We’re about 1,400, which I think is actually the same size as you guys. What you’re referring to I think from the buyout term, we had this program we ran, we called it Detach, which is a Figma pun. We said, “Okay, look, if you’ve been at Figma for a long time, you’re tired. Or maybe you joined and you thought you were joining Adobe, but surprise, you’re at a hard-charging startup. If you want to get off the train, that’s okay. Here’s three months of pay and you’re still in good standing with the company. We’re still friends. If you want to reapply in six months, that’s fine.” And about, like you said, 4 percent or so of people took it, and I think it’s really positive. Now everyone that’s here is committed and we have a lot to do, so that’s sort of the momentum I wanted.

How are you structured now? How’s Figma structured?

We have engineering led by our CTO Kris [Rasmussen]. It’s the largest part of our organization. Product led by our chief product officer, Yuhki [Yamashita], product design research. We have go-to-market, our CMO Sheila [Joglekar Vashee] who recently joined, is leading, which includes not just marketing but also all things support. And we try to really connect the customer experience there. Sales, which has been ramping up quite a bit. Our CRO, Shaunt [Voskanian], is running that org. We also have biz ops, which includes finance and accounting as well as various business functions, led by our CFO Praveer [Melwani]. And we also have people, which is led by our chief people officer Nadia [Singer].

Now, this all reports to you?

This all reports to me, and it’s very functional.

And as people have come and go, people have detached, have you had to change that structure at all?

No, there’s no changes.

And then as you think about growing and expanding, I know you took a billion-dollar breakup fee from Adobe, which is the greatest breakup of all time. More people should exchange cash when relationships end. I would’ve actually lost money on that deal, I think, over time. But you have a billion dollars from Adobe. I think you told my colleague Alex Heath you might acquire some things with that money, do you think you’d have to restructure as you acquire or you think you could plug right in?

I think it depends on what you acquire and how that goes. But yeah, after obviously fees, taxes, we definitely are excited to pursue strategic acquisitions. I think that there’s a lot we could do there, but also just a lot we can do in partnerships, too, and a lot we can build. And so I’m excited to keep Figma as an efficient business. We’re very efficient right now and I’m proud of that, but also make it so that as we continue to grow, we’re able to bring in new perspectives to the company.

You could acquire things — there are indie versions of Photoshop in the world. There are other creative tools or markets where Adobe is winning or losing. Would you acquire stuff that puts you in direct competition with Adobe?

I mean, it’s not what we’re looking at right now. I can say that.

I know a lot of designers who want the indie Death Star. I don’t want to name the apps. So it’s like Pixelmator, all these indie apps that people use that aren’t Creative Cloud, you would never roll those up? 

I mean, look, “never” is a long time, but I can say that those are not conversations we’re having right now.

Makes sense. Last Decoder question, how do you make decisions? You’ve come a long way. When you started as a CEO, you were pretty open that you were learning on the job how to do this. So you’ve come a long way, you’ve fended off or not fended off some charming European bureaucrats. How do you make decisions now? How have you evolved?

I think that every decision is different, and I like to think of myself not as someone who just sticks to one framework, but rather that I have a kind of collection of frameworks. And I think that, even for a single decision, you don’t just want to make it with one framework in mind, but you want to have multiple frameworks and almost do a meta framework approach and then think about it from these different points of view. 

For example, the way that the company makes decisions maybe is more interesting to talk about. I think we are a good mix of top-down and bottoms-up. What I mean by that is there’s a lot that the team just drives and I don’t see, and that just keeps going. There’s also a lot that I push on, and my leadership team and I debate vigorously, and then we have an opinion and we say, “Okay, we’ve got to march in this direction.”

Multi-edit rules, by the way.

Thank you.

It’s very cool. And there are very few companies, I think, that are like, “What is a selection box?” And it’s like this very basic thing. I think it’s very cool. Talk about that.

Thank you. Yeah, so that’s one where, honestly, it took years to develop. And it’s the sort of thing where it exists now and you’re like, “Yeah, of course it works that way.” One weekend into launching, our users have already forgotten, I think, the old way it was because the reality is it almost feels like a discovery. It’s, I think, just clearly the way that it should have always worked. And the reality is it takes so long to figure out these very simple discoveries. And what was cool about the process for Multi-edit was it started with individuals in the company that showed this off and said, “Hey, I think this could be really powerful.” And then over time, we got to maybe 80 percent of what it should be, but there was maybe 20 percent that was really hard to figure out still. And we shelved it for a while because we were like, “We have other things we’ve got to work on. This seems really tough. It’s kind of a research-y problem.”

And then we came back to it over time, shelved it again, came back to it again, and over three years, various people worked on this project until it eventually got to completion. And I was seeing what the team was coming up with and saying, “Okay, this is high enough craft, this is high enough quality or not,” but I was less pushy on that one. It was more really coming from the team, and I’m really proud of the team for delivering something that I think is going to be seen as one of our most fundamental releases looking back 10 years from now in terms of how we actually fundamentally moved the paradigm of UI design forward, even though it’s a simple thing. 

Maybe a different example of top-down would be Dev Mode. This is something where we said, “Okay, we’ve got to figure this out. A third of our users are developers. We’re doing the research, we’re looking at NPS. They are not as happy in Figma as designers are. How do we make this experience better for them?” So the first thing we did is we acknowledged that we didn’t necessarily have the right expertise in the team, and we actually went out and acquired a company called Visly. And then we went and really iterated with them, built out the team more, brought more people in. And over the course of, again, years — I mean both of these projects took a lot of grit — but we were doing weekly or bi-weekly syncs, then sometimes many times a week syncs as we got to the product that is now Dev Mode. And that was a very top-down driven initiative of we have this direction, we need to go in this direction, we need to figure this out. And I think as a result, probably got a lot more consistent attention, but it was sort of a different process.

Let me ask you about a distinction between those two. I think it’s interesting and challenging for a lot of people who make things. Multi-edit is very cool. Everyone should go look at just the GIF of how Multi-edit works. It’s very cool. That is a bunch of people thinking about, “Man, I wish my computer worked better,” which is in a very fundamental way. “I wish I could pick four things on a screen and apply the same edit to all of them, and right now I can’t do that. And I just want my computer to work better.” Dev Mode is, “Boy, we’ve got a whole chunk of our users, we could get more users and make more money if we can build this tool and...”

And, most importantly, add more value.

Sure. So that they’ll pay you more money. But it’s fine. I’m a capitalist. I run a business show, don’t worry, it’s all fine. But they’re different ways of thinking. From the way you’re describing it, they require the same amount of tenacity, they require the same amount of focus. How do you hold onto the focus of the thing that’s just like, “We want to make this core feature more interesting” versus “We’re a startup, we’ve got to grow, we want to grow our market over here,” which everyone seems to understand? 

I think that it’s all about empowering individuals and making a culture where people that are really innovative thrive and they have a path to pushing ideas forward. One thing we do is we have Maker Weeks, where we basically have the entire company for a week. We give them permission to say you can do whatever you want as long as it’s benefiting Figma in some way. And some people are like, “Cool, I can work through my inbox.” Other people go and — this a majority of them — fundamentally reinvent process or come up with ideas like Multi-edit. Then it’s about, “Okay, how do you recognize which of those ideas have longevity and we should keep putting more resources into?” And also, how do you make sure that the people that are those starters that have the tenacity, the grit, but also the creativity to come up with new ideas like this, that they’re still excited to just stay as the company grows and that you are creating room for them to be creative?

One of our values is play, and I think that it’s really important to lean on that. There’s periods where you don’t, where it’s very gritty and we have to really be just hunkered down. It’s like we’ve got to road map, let’s execute, and there’s periods of more exploration. AI is an interesting example where we had an AI hackathon, and it was just two days. We said, “Okay, let’s see what comes out of this,” and there were a bunch of ideas that people presented. A few of those were ones we doubled down on. So one of those was something called Jambot, which is, in FigJam, you can have this widget that lets you explore and do basically GPT queries in a nonlinear way. So we all have experience, we know of ChatGPT, where you basically are having this very linear conversation. Well, what if you could make that actually branching and try to basically explore that in a way that’s nonlinear? And people really enjoyed that and continue to enjoy it.

We also have what became basically this prompt to a FigJam template that came out of that initial AI hackathon. Basically, we see a lot of people that use FigJam not just for diagramming but also for whiteboarding, but in the context of a meeting and making a meeting just fundamentally run better. And if you look at the way that Figma uses FigJam, if you tour our office, for example, you’ll see FigJam up on all these screens. Sometimes these people are all in the same room, sometimes they’re hybrid or all remote. It literally works across all those modalities. But it’s really neat how, if you set up FigJam the correct way, it can really help get more people to contribute their ideas and also it can help you converge as well. So it really maps design thinking in that way. But there’s that requirement of you’ve got to set it up right, and that takes a while. So how do we help people do that? And that’s what prompt-to-template created.

And we’ve seen people that use prompt-to-template — this AI feature for FigJam — they use FigJam a lot more and they’re more successful in their meetings. And I think that’s something that also wouldn’t have come up as quickly if it wasn’t for this more grounds-up innovation culture we have.

You’ve predicted my cards very well. Well done. My next questions are all about AI.

Glad to help.

Thank you. I need all the help I can get. I understand how AI helps enterprise tools like FigJam, right? I understand how it helps engineers write code better. I understand how “I don’t want to write emails, I’m just going to let the robot write my email.” I get it; do it. Figma is a tool for creatives, and creatives, I think, have a much more charged relationship with the concept of AI. How do you see it? How do you feel about that right now? You’ve got a big audience, a market of creatives who are, I would say, most charitably of two minds about AI, especially generative AI. How do you see it?

Well, the way we’ve talked about it internally is that AI has this opportunity of lowering the floor but raising the ceiling. So it could enable us to get more people into the design process, but also it should make designers way more efficient. You should come to Config this year. It’s in June, all of you should. I think when we are at Config, you’ll see a bunch of stuff that kind of points to both those directions: ways to get people more involved, even if they’re people that are not comfortable typically with putting components on a screen or drawing in Figma Design, and you’ll also see ways that designers are able to get a lot more time back, be more efficient. And to your question about how will designers navigate this and how will they react, my point of view is that AI will help people navigate abstraction boundaries.

So, for example, we have brainstorming with whiteboarding in FigJam, we have Figma Design, we have Dev Mode. How do you cross those abstraction boundaries with AI? I think that will become something that’s really interesting. I also think that if you think about the design process and just building software in general, we’re all navigating this idea maze, and the idea maze has all these branching possibilities. AI, if you are extremely generous, let’s say that we get to the world — in three to five years or whatever — where we have AGI. Let’s just take that sort of bull case for AI.

Three to five years?

No, I’m not saying that’s my prediction.

Okay.

I’m just saying let’s say that you have some AGI sort of thing in three to five years [or] 2030, seven years. Let’s say it makes decisions right 95 percent of the time. It sounds kind of scary. But then if you actually think about it, if you’re going across that branching path of decisions and you have all these different steps along the way where you have to make a call, if you take the compounding effect of 95 percent right, you get to 1 percent chance over a long series of decisions that you get a good result. 

And so my point of view is that even in this kind of world, you have so much context that a designer has about all the past screens and affordances they’ve ever seen, the emotional context the user’s in, the cultural context of what’s going on, the business strategy, the business goals, and more. Even just the temporal aspect of how do you have people navigating this application and software?

And maybe you could write that all into some 30-page brief, but at that point, even if you do, it still might make the wrong calls if you were just to say, “Here’s my brief. AI, go create designs and software.” I think there’s got to be a human in the loop. And so I think what ends up happening and what we have a chance to facilitate is making it so that you get to that first draft of something faster and you can explore the possibility space of what can happen faster. And that should be a way for designers to have even more superpowers and actually give them even more power in the organization.

Alright, let me offer you the exact flip side of what you said.

Okay.

Which is, right now, every open input box in the field is being flooded by AI garbage. We just see it everywhere. My favorite example of a choke point on this is Amazon reduced the number of books you can self-publish in a day from 5,000 to 200. And it’s like, I don’t think that was going to fix it, guys.

So why did they do that?

Because they saw this flood of AI-generated books that are being put onto their platform. And that’s just one example. Wix, who has been on Decoder, they just released a full generative AI website builder. You type a prompt, this thing ships a website for you. On the one hand it’s very cool, there’s something very cool about that. On the other hand, it’s terrifying because the web is going to get even more polluted and that’s what you’re competing with. Your vision of the future is you will not want to make all these compounding errors, you’re going to want to make a great product, and the competing vision of the future is, what if we make even more garbage and make it harder to find the good stuff? How do you navigate that?

Well, first of all, I think the people that use Wix are not the people that use Figma.

Shots fired.

I’m just being honest, right? I don’t see them as a competitor actually.

Okay.

Anyway, we can talk about that another time, but the...

No, keep talking about that.

So if I take a step back for a second, the fundamental bet that I now understand we were making when we started Figma was that more software is going to be created in the world over time. And that’s definitely been true. We started Figma [in] August 2012. It’s just wild to see even, over the past decade, how that curve has gone in terms of software in the world. And I think that the interesting thing is that we almost all just kind of forget that that’s happening. It’s just kind of taking it for granted: “Oh yeah, of course software is important.” It’s in the air that we breathe now. But from my vantage point, what I’m seeing with Figma, we’re still very much in the exponential curve of more software.

And I think that with AI, whether it’s Wix making a website that’s generated that way or it’s us helping people create more designs or many other effects that will happen, I think there’s going to be even more software that will be created. And I believe that design will be the way that you differentiate that software and make it something that people want to use and your business wins. Or if you’re not strong enough in design, that’s why you might lose. And I think the digital experience is hugely important today, and I think it’s everything tomorrow.

I think you’re getting right at the tension. A bunch of creatives in this audience listening to Decoder, whenever we write about an AI tool, we get a flood of angry people who are like, “That’s not good enough and it’s actually built on our work, it’s trained on our work without us capturing any value back.” And the overall desire of the audience to have good stuff will go down because they will live in a world of C-plus work. And I’m just curious, people use Figma — Dev Mode for example — you’re inching closer to type in a prompt and Figma and get all the way to a shipping app. You can see how the next four turns of Dev Mode get you there. Are you worried that you’ll contribute to the sludge problem?

I think that even if there is more content and more design and more software, that just creates more demand for more great software. And so, I actually think that in a world where you’re able to navigate this idea maze faster, you’re able to point in different directions. First of all, you still need to refine. You still need a human loop the entire way you go through that for the best software. But also I think that there’ll be more opportunity and more demand for design in order to make it so that you’re able to navigate how do we create the best experience? Because if you’re right and there’s way more C-plus content, of course you’re going to want to rise above that if you’re trying to make the best thing and attract the most people to your digital experience, whatever that might be.

Yeah. One of the things I think...

Does that make sense?

It does. I think you are more optimistic about it because you sell software and I write about software, and I think that’s a natural tension that we can be okay with. The thing I think...

I think what we’re both saying is there’ll be more stuff put out, and that, to me, implies that there’ll actually be more need for more people to work through that. And I think maybe where we disagree is you’re saying that just like there’s going to be people that say, “Okay, write this book in a thousand ways. And AI, go.” And I’m saying, actually, what will be needed for a business to succeed in this environment will be more people working in collaboration with AI and AI agents over time and making sure that you create the highest-quality results.

I don’t disagree with that, broadly. I think what I am worried about and consistently worried about is that having the best thing win implies that people can find the best thing. And when you are in the sea of garbage, and our internet right now is like a sea of garbage...

With a lot of gems.

But it’s getting increasingly harder to find them, right? And the ways for me, an app maker, [or] you, a software maker or someone else, an author, to go find a real audience that’s going to value them more than 50 AI-generated ripoffs of that that are free on some platform that is designed to make you download Temu or whatever — it’s the internet we’re in. That’s going to be too hard. And that’s the thing I worry about. I want great design to succeed, I want great apps to succeed. Every day, I see that it’s getting harder and harder to build that business than the “Here’s some garbage” business.

It’s interesting because you look back at the history of the web, and I feel like people have predicted the web’s demise so many times. When I was a kid, I grew up using AOL and even then, I was reading people saying in the tech section of the newspaper… Because in the 2000s era, there was a tech section; that was really nice. It went away one year.

You’re saying you don’t like The New York Times tech section?

No, I meant there’s a full section in the newspaper every day. Anyway, people were saying AOL is going to kill the internet. And in reality, I think it opened up a lot of people to want to explore the internet. And then I was a teenager and people were like, “Oh, Facebook, it’s going to kill the internet.” And I think the same thing happened. A lot of people onboarded the internet through Facebook. 

And I think that there’s a lot of ways you could say that the web is going to die today. People might say, “Oh, well, you’re going to just interact through a prompt or you’re going to be lost in the sea of garbage.” There’s always a distribution challenge, and the platforms are all trying to figure out how to navigate that. But ultimately, there’s always going to be ways to find the gems to go explore, and that’s the beauty of the web. And that’s why I’m optimistic about the web and excited about the future of the web.

Are you sure you didn’t see these beforehand? Because you’re right on the next one, which is the web.

I asked you if I could see them before, and you said no.

I refused. I gave him one question to think about, which we’ll get to later, which is a fun one. Figma is a web app. That’s a remarkable thing to say in 2024. We run a website. I keep joking that we run the last website on Earth. I think there is a view of the web, particularly in mobile, the web will be a document viewer and the application model will be inside the App Store. And we’ll have the full horsepower of the phone, we’ll be able to manage battery life. You bet on the web, you bet on some key enabling web technologies from the beginning like WebGL, WebAssembly, all these things. Do you still see the web as the best application platform?

Oh, that’s a good question. I think for desktop computers, for sure. I think on mobile, you’re more limited. It’s very good in some ways and in other ways, we need improvements still. I think that in a world where Google and Apple really invest in the web via mobile, that’ll be a very good world, and I’m excited for them to just engage more with the developer community over time.

Are you feeling like a real love / hate with Europe right now? Because the DMA in Europe is going to make Apple have to support alternative web browsers on the phone. They’re going to put a horrible choice screen in front of people, but they’re supposed to support alternative web browsers on the phone. The idea that the mobile phones will have richer browser innovation, I wouldn’t say it’s real, it’s potentially real. Do you see that as something that’s exciting?

Well, I think we’d be a beneficiary of that probably.

Yeah? What would you need from that process to be a beneficiary of it?

There’s fundamental limits that we can’t necessarily influence as much, and if you have alternatives, then I think that there’s probably more you can push on for other parties in order to get...

This is very careful. Be 10 percent less careful in this answer. What do you mean by “there are limits and certain changes”? 

Like for example, I’ll be very specific, Figma Mirror is a way to mirror designs on mobile. And we still rely a lot on the web for that process, and there are fundamental memory limits. Sometimes that app crashes for us, and that creates a lot of work for us to make it really efficient and make sure that we’re not filing those limits. But there’s also things that are sort of like spooky action and distance problems where you really need collaboration with browser vendors and operating systems in order to fix them and figure out what’s going on. And you can poke around and figure it out over time, but more open source, more transparency will help us achieve better outcomes there probably.

You said just a second ago, the web is the best way to ship applications on desktop. I think we see this everywhere. Every new exciting app ships as a mobile app and then as a web app, and there’s very little native Mac or PC software, in the productivity space at least. Why is that? Is that, you don’t want to fight their app stores, you don’t want to deal with their weird binary ideas? Is it just the web is easier and you can just get everywhere?

I think that the power of a link is so understated and misunderstood still. It’s amazing to have an atomic unit like that that you can share, and I think that the web default enables collaboration through the cloud and that’s the other shift that we’re seeing across basically all software.

And so, you think that some of these changes that are coming to mobile might bring that to the phone... Because you just don’t see that on the phone. The idea of shipping a great phone web app, I think, has hit a stasis point. Like, this is as good as it gets and everyone kind of pushes you to mobile. Could you see some changes that bring out great phone web apps?

I think that we’re confounding a few things in this question. You don’t have to predict that people navigate to applications through a web browser versus through apps to believe that apps will get better because of just more technology options available on different platforms. Does that make sense?

It does, but I think that somewhere in there are two giant companies who would prefer that their phone platforms get monetized through in-app purchases and mobile apps and limit the power of the web. And so I’m wondering if you see anything that overcomes that beyond the DMA or whatever other regulation, if you see user behavior or technology shifts that might actually disrupt that.

Yeah, I see what you mean. I think that it’s interesting. Like right now, if I go buy a book on my iPhone for Kindle, I have to go through this weird browser experience. And yeah, it’d be nice to just buy it and I don’t have to log into Amazon through the browser again. But that’s pretty light friction right now. And I think that as you open up a platform, still a lot of people will choose those default payment options because they’re a great user experience, because they’re just easy to implement. And yeah, you’ll have other people that choose other options and they might not be as good of a user experience. But, at the end of the day, I think it’s more both than one. Where I think about it more from, again, is how do you make the best user experience and what technology can enable that?

Let me ask you in a much simpler way. When I was growing up, all designers used Macs. It was like the computer of choice and that company was aligned with Macs. Now all the designers I know use Figma, and a lot of them use Figma on Chromebooks because you have abstracted the actual thing away and the product lives in the cloud. That right now, structurally, cannot happen to the phone. You have to be bought into one application model on the phone. Figma, I think, is a shining example of that abstraction on the desktop. Is there something happening you see right now that can make that happen to the phone where Figma could ship a bigger and better web app on the phone?

I see your point in terms of where it could go, and I also think it’s exciting that we’re starting to see more hardware platforms, software platforms come out for mobile.

Are you going to get a Rabbit R1 with Figma on it that’s like, “Make a website”?

What I will say is I’m pretty excited about just the fact that there’s people building that kind of stuff. I don’t necessarily think that that’s the vision of the future that I think is going to happen, but it’s just cool to see more innovation, more people trying. That’s unique, and we haven’t seen that for a while. But yeah, I think there’s something to that exploration. I’m the guy in the MMO that goes and tries to explore the entire open world, and I sometimes ignore the quests. Modality should be in sometimes where you’re trying to achieve and you’re trying to get an answer really quickly, or order on your DoorDash. I think the demo is like he orders a pizza, it knows exactly what he wants and does it for him. And there’s a mode where you’re like, “Okay, I’m in Austin for South by, let me see all the different options,” or “Actually, I want to go read all the different analyses on this topic.” I think that both exist.

Alright. I’m going to ask you a couple design questions and we’re going to open up for questions. You’re a design nerd. I think coming into South by, the biggest design story I’ve seen floating around is the Rivian R3. There was a Cybertruck parked in front of this hotel yesterday that drew a crowd. Those represent, I think, different poles of design, charitably different poles. One, for example, is a triangle and the other one is not. So it goes. But they have a different feel. And we were talking about this backstage.

Yeah, exactly.

That you think there’s a shift in design coming and the R3 kind of represents that.

I think that a lot of things in trends in design are sometimes reactions to what was happening before. I think that we were really futurist, really cyberpunk for a while. We were talking about it, like a lot of neon lines, a lot of hard edges, a lot of poly or low-poly sort of metaphors. And it feels like we’re going more humanist. It feels like we’re going maybe more solarpunk.

Solarpunk?

Yeah.

Alright.

I think solarpunk’s pretty cool. It’s more of this optimistic vision of the future, and I like the idea of that, whether that’s what is happening or not. Maybe I’m a little bit projecting my own wishes and desires for what I want to see, but yeah, I’m hopeful.

Where are some of the places you see the solarpunk trend?

It’s a good question.

And solarpunk is in opposition to cyberpunk?

Yeah, I think they’re two poles of the same thing, but cyberpunk is dystopia, everything’s metal and concrete and low poly. It’s kind of depressing, but it’s also got that edge to it. Whereas solarpunk is more natural, the architecture is blending with the environment. I think it’s still a vision of the future, it’s still futuristic, but also it’s more human and it’s got a lot of curves, it’s got a lot of... It’s sort of modern, but it’s modern in a sense of it’s blended into the environment more rather than trying to stand out as a piece of technology or a piece of architecture. And yeah, I think we’re going more in that direction.

Do you think Figma, just as a tool, as a design tool… When you see your customers use it, do you think, “Okay, we’ve got to lean into letting people make that more or less,” or is it just a blank canvas?

Well, it’s interesting because at the start of Figma, the metaphor we had before we even shipped was we are the concrete floor and the user’s content is like the nice rug. And we went super brutalist in our design direction. In the earliest days of Figma, I remember showing designers the tool, and the reaction that we got almost universally was, “Really cool idea, I just am not sure I trust it.” And I was really worried because I thought maybe that this was some reaction to the web and we had made a fundamentally bad platform decision, getting back to our earlier conversation. And I think what the reality was is that we did a redesign basically, made it a little bit warmer, a little prettier, and suddenly everyone got over that concern. The reality was just designers didn’t like the look and feel, and designers are very tuned into look and feel.

But I still think that there’s something to that idea of how do we get Figma out of your way? How do we make it there when it should be there, but how do we keep the focus on your content? And there’s so much we could explore in that direction still. I see 10 years of work to figure that out. There are ideas that I was pushing for in the first two years of Figma, that we had to shelve, that have recently started to come back up again. Not from me, but from our team. And I’m like, “Yeah, I’ve looked at that. That’s a hard one. Let’s do that after we finish this project.”

Give me an example.

One thing that I just get so excited by, but it’s a very hard problem, is just more contextual UI. And you see this in some tools, especially in the 3D space, but I think that there’s a lot of challenges with how do you both maintain the predictability of an interface while also making it so that you don’t have to move your mouse too much? And I also think that the more deep that you want an interface to go, the more power you need to be able to get to, the more sub menus and whatnot that some people have to navigate. So it’s a very cool vision to go, how do you bring up exactly the right things at exactly the right time? And I think there’s something there that Figma needs to explore at some point, but I don’t think it’s an easy problem. And I’m definitely looking forward, though, to just going for it at some point and trying to figure it out.

Is it the curse of every software CEO that you try to invent Clippy at the end?

Try to what?

Try to invent Clippy at the end.

I’m actually not talking about Clippy here.

Really?

No.

You think bigger than that?

Bigger than Clippy? No, I think Clippy is... Maybe I just didn’t describe…

You don’t want like a Figma avatar popping up and be like, “Yo, this design’s kind of played”?

So it’s funny you mentioned that. Our internal mascot that just won’t go away, we call it Flippy. It’s like a Figma logo with this weird hat on. So it’s an internal joke and then it just keeps popping up in random places. It’s like a meme that just won’t stop. And then Flippy ends up getting killed in various ways, internal decks and stuff like that. But no, I must have not described this correctly before and maybe that’s because it’s still vague and still out there, but it’s a fundamentally different metaphor than an assistant.

It understands what you’re doing and provides you the tools.

No, I am not talking about that at all. I’m talking about, how do you instead provide the right actions and interface at the right times.

Do you think that is an AI-related problem?

I actually don’t.

Really?

Maybe it is. Maybe that’s some gap we need to explore more, but I think that you still need the predictability if it’s a professional tool. You have to think a lot about how do you always have the thing available that people expect, and how do you have a consistent way to get there, and that’s a hard thing to solve.

One of the things you and I talked about the last time that this brings to mind is there are things on the web you cannot do, and the one that you brought up was pen input. WebAssembly and WebGL are just not fast enough or performant enough to do pen input. Maybe that’s changed in the past year since we talked, but it feels like the pen, that modality is where you want more of that. My pen can just do whatever I need it to do right now. Is that a better framework of thinking about it?

I think you could do a lot of stuff with pen. So I have to look back at our last interview and conversation to figure out what I said there. Maybe one thought experiment is what would it take to make it so that Figma is as least intimidating as possible, as approachable as it could possibly be? Could you get away with somehow removing the properties panel on the right-hand side, the layers panel on the left-hand side? What would be your path to get there? And this is sort of an extreme thought experiment, but I think if you follow that, you still need to have the right properties at the right time, and how could you surface those contextually in the UI but in a consistent way? I think that would be a really interesting way to both blend into the environment of the user’s content as well as creating flow for the user. And to me, somehow that rhymes with solarpunk in my head. Getting back to the original conversation.

Outside of solarpunk, are there other design trends you see on the web that maybe are not as aesthetic design trends but more functional design trends?

I think it’s a trend in conversation, but I think that there’s two trends that are happening simultaneously right now in the culture. We’ve maybe touched on both of them in this conversation so far, which is, there’s existentialism and people being like, “What is the future of humanity in a world of AGI?” And there’s optimism. There’s the point of view that it’s going to be okay. Look, when we did the deal with Adobe in the first place, my head space in 2022 was, “Oh my god, AI is coming. This is clearly exponential as a technology. I don’t know what this does to us. Is this one-tenth our market, is it 10x our market? What does it mean for creatives and designers?” And I was like, it’s better to team up in this world with Adobe and to navigate this together and to figure this out together than it is to go it alone.

Through the deal period, through the last few years, I have grown increasingly optimistic about… I don’t know what’s going to happen with super intelligence. If we get there, who knows? But in a world of just AGI even, I think that there’s a human in the loop. I have a lot of optimism about that future and how things will work out, not just for design, but for knowledge work in general. I hope that more people start to see that and that trend of that optimism continues.

That’s great. Alright, that’s it. Thank you Dylan so much.

Thank you for having me.