#141 Designing Perplexity - The Builder's Guide to AI Design w/Henry Modisett
40:18 Share

Designers are used to sort of like put ornaments on the tree or visualize all the possible ideas that like leadership has. And for me, like a design is a problem solving exercise. It's a practice of like making the business work better or just like generally kind of making the world around us better through.

both like form and function. You know, when I, especially young people, if I could give them any advice, it's like, yeah, the best way to actually be good at that and have experience is like make something and try to get other people to use it because you're going to learn a lot when someone says no or if you get them to download your app or whatever it is and they like don't reach for it every day, that's like,

incredibly humiliating, but also one thing to learn. And you got to learn that. You got to put all your ideas out there and find out how wrong you are until you can kind of find the truths eventually. And you just don't get that experience when you're in a big company because all you're going to get is like feedback from other people in the company and maybe some data from A-B tests or whatever. And that can be obviously very valuable, but it just depends on how it's set up. I mean, most of the time it's so easy to learn the wrong thing.

And I think for many of us, it's a time to unlearn some of the things we've been taught. That's really interesting. Adaptable now means forgetting some of the things you were kind of fed.

Hello designers and welcome to a new episode of Honest UX Talks. This is a very special episode, one I'm incredibly excited to share with the world because it features one of the most inspiring conversations I've ever had with a designer. And it just so happens to be with someone I've admired for a long time, someone I truly consider an icon in the world of AI and design.

We're diving deep into AI design and product thinking with someone who's been building right at the frontier, someone who has helped disambiguate how AI should surface in products in clear, meaningful ways. Henry Modissette, the founding designer of Perplexity. If you've worked in product design over the past couple of years, you probably already know why I'm so excited.

Perplexity has quietly become the gold standard for what good AI design looks like. Back when I was at Miro, our team constantly referenced perplexity for interaction patterns, tone, flow, branding, everything. So, naturally, I had a million questions to ask Henry. How did I figure it all out?

What were the messy parts behind the scenes? How do you even design for something that doesn't behave the same way twice? Henry shares all of that and more, from his early scrappy projects to designing Gmail at massive scale and now leading a fast-moving creative team of what he lovingly calls misfit designers at Perplexity. We talk about designing for unpredictability, the power of prototyping in code, why he feels zero anxiety about the future of design, and

what it means to design AI-powered products with care, clarity, and confidence. It's a smart, deep, and fun conversation that genuinely inspired me and helped me grow as a designer. I'm so grateful we had this chat and I hope you'll find it just as energizing. Henry is one of those rare minds who can go from technical to philosophical in a single breath. It was such a pleasure to unpack ideas with him."

And speaking of building fast and staying in flow, let's talk about tools for a second. Weak Studio is pushing the boundaries of what modern design and development can look like. Their AI code assistant built directly in the IDE help you generate clean code, automate layout tweaks, and get smart context-aware suggestions so you can focus on what really matters, solving problems and shipping ideas.

It's the same principle that Henry lives by at Perplexity. Fewer static mocks, more real prototypes, tighter feedback loops. It's exciting to see a platform with such skill offering tools that genuinely empower creators rather than getting in their way. So big thanks to Wix Studio for supporting design conversations and creating an awesome product. Now, let's get into it. Here's my conversation with Henry Modisette. And yes, I absolutely loved it.

So without further ado, I'm very happy to welcome Henry on Honest UX Talks. It's a very exciting moment in my career to talk to someone who is so well-versed in building AI products and has been doing a lot of pioneering when it comes to design decisions and AI. So I'm very happy to introduce Henry. Welcome to our show. Thank you for accepting my invitation. It felt super natural and super friendly and

And yeah, you were very approachable. So I really appreciate the attitude. And I want to invite you to tell us a little bit about yourself, your background and how your journey with perplexity started. And yeah, just welcome.

Yeah, thank you for having me. My background, depending on how far you want to go back, I would say professionally started, studied computer science in university. So that certainly shaped how I think about software and making software coming from a perspective of one of someone who makes software itself, thinking of it as a mutable surface. And I think that's important because a lot of designers that I know that come from more of an art school background, they really think of themselves as making artifacts. And I think that

really can hold you back, especially for, you know, the type of software that we're working on, it being non-deterministic and interactive, and we're always changing it every week and on and on. So that's sort of like very formative. I did study computer science, was very kind of into just making little stupid startups every opportunity I could. I was more interested in that than doing well in school.

Which was great because it gave me a lot of practice, a lot of reps of just like having an idea and trying to get it working and then see if other people would use it. To me, that's like the most important thing I could have ever gotten practice with. I also just found that I liked working on the user facing part of it of whatever I was working on. So yeah.

the design, I guess. Then that led me to start a freelancing, but I didn't tell anyone I was in college. I just said, I'm a designer if you want. I'm a pro, pay me money. Again, like that just sort of got me a lot of experience. I was designing stuff that I really had no business designing, but I think that's the one most wonderful thing about being in this industry is, you know, if you have a nice portfolio, people will take you seriously.

Nothing else really matters. So that was sort of like where I figured out who I was. And then I got my first like normal job working at Google, working on Gmail. And that really sort of opened me up to thinking about consumer design at scale. I mean, hundreds of millions of users when I joined.

I think that just completely changed how I had to think about software mostly because before I was just thinking, what do I want? And maybe what do five other people want? And then with Gmail, it was more like, how do we make this thing work for hundreds of millions of people and in every language in the world and every type of computer and screen and device that exists? It's just a completely different...

problem space. So that was like a very sort of additional formative experience. And also first time being exposed to this problem of every user is going to have a different experience, which is very challenging, especially if you do come from graphic design background, right? Like if you make the thing with the tool and what you see is what other people see, if you're used to that, it's going to be very challenging for you to accept that like

every user is going to get a completely different outcome. Obviously with email, everyone has a different tapestry of email, right? Some people are using it for work. Some people use it to like get coupons to buy clothing, right? Like it's just truly all over the place. It's a universal system. And so it's pretty hard to like figure out, okay, what should it look like?

And how do we apply hierarchy and create sort of predictable systems that anyone can figure out? And the way that we did that was we would build prototypes and we really wouldn't rely on mocks too much. And that really helped us think about this and prototypes where we were plugging in our own email. So, you know, you didn't really know that the design felt right until you got an important email and maybe you missed it. And you're like, oh shit, that probably should have been easier to see that that was an unread email, things like that. So that was a

important experience for me. And then I would say my next kind of biggest chapter was working at Quora.

Quora was important because not only was it a free consumer social product and all the challenges that come with building a social network, but also it was at the time we were very early in applied machine learning and the product experiences. So we worked on personalized ranked systems like feeds and notification systems and email scheduling systems. We were doing this, I would say, pretty earlier than most companies. And therefore, we had this incredible talent density of machine learning engineers. So there's

Me working on this, most designers in the world weren't at the time. I think it's sort of now how most software is made. And then I was with all these machine learning engineers who were getting to try out their thesis from their PhD into an actual applied thing. So...

that was again sort of almost the same problem like okay everyone's going to get a different experience how do you design a feed for everybody i don't even know what the content is going to be i don't know what it's gonna how good it's gonna be and how do we even measure the efficacy of both the ui design but the ranking so that was like a wonderfully challenging problem and the kind of fast forward to here

With Perplexity as well, two of the co-founders of Perplexity worked with me on that feed team. For me, I'm the only designer they'd ever met, so they hit me up when they wanted to start Perplexity. But I also think I was particularly well-suited to this because of all those experiences, working on consumer, working with very technically-backed product experiences, and particularly ones that were non-terministic.

I found that it's like something that I'm very comfortable with. And actually, most designers aren't. So that's my off the cuff bio. This is one of the most entertaining things I've heard in a long time. So I really have to say that. I think I also reached out to you because you seem like a very fun and articulate and an interesting person to work with. And my thesis right now, after 15 years in tech, is that

people with personality infuse that in the products and especially for products that are ambiguous or just they're in a way shape-shifting very often and they don't have a clear structure you really need that kind of person that has opinions and express stories very nicely and it's just not an

early in my career, just talking about it felt very stupid. But now I realize that so much in the success of a product is how well the people behind it think and construct narratives. I was reading through your story about how everything connects, right? There's that narrative arc, like at Google, we were designing for, um,

predictability when it comes to how an inbox looks like and having to accommodate all sorts of things that might happen in a person's inbox from coupons to important mails. And I was listening to this kind of everything led up to being able to operate with the complexity, ambiguity and non deterministic behavior that AI is presenting us with. So it was a pleasure to listen to you.

I really loved your story. It's inspiring to see how things connect. Let's talk about your time at Perplexity and coming into a product that's essentially inventing something that was there in a different form. We've been searching for things and your work at Quora and even at Google kind of informs this new structure that you're building. But at the same time, it's

essentially novel and it's infant technology. We don't have the principles, we don't have the frameworks, we don't have the patterns established. So you're inventing stuff and making decisions. And I loved how you talked about this idea that you had to prototype in order to make a decision and understand how it feels and works. I think this is one of the key things in working with AI, but I would love to learn more from you on how it feels to just create something that wasn't there before.

And how did you make design decisions? Was it just vibes or was it a lot of prototyping, testing? How did your work at Perplexity look like until now? Totally. You know, with my early kind of beginning of my career, that was the only way to make stuff, right? You had to just sort of think like, what, how should this work at all?

you know, there were no patterns or frameworks, especially on like early mobile design. We were all just making it up. And there was a lot of obviously bad design and stupid stuff and confusing things. And I think like where we ended up is fine, you know, in terms of the kind of most common interaction patterns. But I think it's an important thing to be able to do as a designer to say like, okay, forget everything, you know, and remember how do people actually interact with anything, right? Whether it's like a

confusing showerhead or TikTok, you know, they look for affordances. They look for, you know, the brightest thing on the screen. They look for contrast and they want to, they're going to push the thing that looks interactive and on and on. Right. So I think it's like super important to just like be able to fall back on that and then figure out how to apply that to any type of surface. Like maybe you are making an app or maybe you're making a voice only experience or maybe you're making, you

door handle. It's, it's kind of like all like the ergonomics of humans are not changing in any of those contexts. It's just sort of so that, you know, everyone's got, most people have eyes and ears and whatever. Right. So that is going to be the thing that you have to be able to fall back on. And I think most designers are kind of hungry to get to use that part of their brain. And really the only thing since mobile that has been available is VR.

as like an AR, I guess, like as like a way to sort of like, OK, let's invent something. And the problem with those is that they're not comfortable, right? They're not comfortable to work on. They're not comfortable to use. Like it makes me dizzy, whatever. And it's going to be hard to like work those things into people's lives. So the fun thing about AI is it's sort of got that property, but like we're

putting it on your phone, we're putting in all the technology that people already use. And actually we're putting it in places like it can be even simpler than that, right? Like we, you know, a lot of the best product experiences with AI are just completely voice only, you know, there's like a true reduction of the interface. So I think like that is popped up available. Like if you can see it for what it is, you'll probably fall in love with the problem space and not be afraid of it. That's like a very kind of interaction design, kind of HCI kind

framing of it. Like another framing for me is always just what matters with consumer products. People want the thing that is fast and useful or entertaining. There's really not much thought in it. If you make something that does whatever they're used to doing and it does it faster, people are going to use it every day. It's sort of as simple as that, but incredibly hard to actually accomplish that. I mean, I have like some benefits from thinking about

knowledge within the context of Quora. How do people obtain and create knowledge? And how is it organized and presented on the internet? How is it discovered, both from humans, but also technology, whether it's through SEO or whatever. And the major limitation of Quora was always it was a two-sided marketplace. You needed people to ask questions and other people to answer them.

The positive of that is it was mostly incrementally adding new information onto the internet. But the negative side is like if you wanted to know something that like was already out there, it wasn't faster. It was actually way slower because you had to wait for somebody to answer your question. So that when we were prototyping ideas with Reflexity, we got this like really simple thing working.

And the wonderful property of LLMs is they always, it always says something, even if it's completely wrong. And so, and you know, like that was ChatGPT when it first came out, it was like, oh, cool. It's responding. If you want something that's actually important, like, you know, if you want the information to be like important or robust, be careful. And that was our opportunity. It's like, okay, well, what if we just made it actually accurate?

And that was our original innovation is sort of hyper-focusing on that and building all the technology to make that happen, which was a search engine under the hood of an LM. So there's like that, we got that working and then, and then, you know, just basic consumer thinking like, oh, what are the things that people actually need to know every day in simple things? Like what's the checkout time in my hotel? Okay. I can find that out on my phone. It takes me like maybe four minutes from Google to finding the website to scrolling and whatever. And then it

And with perplexity, it was like 15 seconds. That's not that much time saved, but it's actually like a lot of time saved on your phone and when you're traveling. And if you multiply that to millions of people, that's like actually a huge difference. There were these little things like that where it sort of becomes very obvious. Like, okay, like there's a thing that there are all these things that I do every day. And we just made all of them a little bit faster. In some cases, a lot faster, right? Especially with any kind of like more major research project.

That is, you don't really go into any of this thinking like, okay, I want to be innovative. You go into it thinking, what are the fundamental properties of like our existence and like how people decide what to do with their time? And, you know, people need to know stuff and they probably will, if there's a faster way to know stuff, they'll do that.

And it's sort of as simple as that. It's just hard to get there. So that mindset kind of gets you, if you kind of know those things about the world or the universe or whatever, and then you find a technology that kind of like opens a door that wasn't there before. I think it's pretty easy to disconnect the dots. I feel we're in the realm of poetry.

Listening to it feels really poetic and philosophical. And I just wrote down fundamental properties of the world because I want to reflect on it a little bit deeper after our conversation. So thank you for this journey into how to deeply, but in a way, straightforward, right? Simple thinking about building products without overcomplicating with, I don't know what design framework or process or like these rules that we make up so we...

Totally. This is how our world should look like. And I love it. It's really straightforward and powerful. And I resonate with everything you shared. I was the first AI product designer hired at Miro. And...

they're in the capacity of this uh ai leader that should kind of infuse with these new things new systems of building products and so on at the end of the day at the end of every day in my work at Miro I realized we're just building for these what you called fundamental properties like people are asking things through writing it's just they're doing things that they were doing with their friends but now they're doing it with an ai entity in front of them so it's

But fundamentally, it's the same kind of components or tokens, small pieces that come together to assemble an experience. It's true that some things are very different, like this non-deterministic nature and designing, accommodating for the unpredictability of everything, which is also kind of abstract. Like, how do you really design for unpredictability? But anyhow, I want to just say that even back then when I was starting out at Miro and

Miro is essentially an infinite canvas where people put information. So knowledge retrieval and operating with the knowledge that's in there was something that we were also deeply thinking about that something that AI could contribute to. And we were always referencing perplexity.

So somehow in your simplifying the thinking and building a straightforward product that also looked good and felt good, like it was as opposed to what GPT was early on. And it's still not a very pleasant experience. It's very basic. It's in a way boring.

Sterile, I don't know how to say it, but you get my point. It's stripped of any emotion in its design. But with Perplexity, we kind of had that feeling that, oh my God, this is a nice product that we want to use. So we were always referencing it. How does Perplexity do that? You know, in conventional design, we used to ask, how does Apple do it?

But for us in AI, it was like, how does Perplexity solve this? You really managed to infuse this kind of clarity that you just shared with us. It shows in the product you've built. And I think that's an incredible win. My follow-up question would be, who was by your side on this journey? Meaning, how does the design team look like at Perplexity? I think everyone is curious. Is it two people? Is it a big team?

How does it really look like in reality? I'm also personally curious if you've hired designers that have AI literacy or AI fluency, or is it just a team of talented people that came together? Totally. We've been around since 2022. It's been a rush for me. Like it feels like yesterday, but it's been a couple of years. And I would say for the, maybe the first year, it was just myself and

and a few engineers. So, you know, I just joke like, oh yeah, we were five guys on Slack and we're remote. Like we didn't have an office. That's like the wonderful thing about this, you know, building software is you can kind of just sort of, there's not like a requirement on any of that stuff on how to do it. So I was designing, but honestly, I was coding every day. I mean, it's like, this is like how I like to work. I just...

need to make stuff and it's hard for me to think statically like I don't like to use design tools that often maybe to maybe help plan out some stuff or figure out colors or whatever but mostly it's like if I'm building something interactive I need to start building something interactive and then I'll be more confident and it could be anything from the you know the flow of the software but but also just like the details I mean I love the precision of code if I'm trying to get

the right shadow or the hover interaction. Like why would I spend time doing that in a tool that's not the thing that users actually use? Just kind of helps me with my time and it's just where I'm comfortable. So for the first year, I was not only the only designer, but also the only front-end engineer. Nobody in the company knew any CSS at all. I was doing SwiftUI too on their mobile app. And so that was like the early part of the company was just, you know, I would just say we were all just engineers just making stuff. And I was the one doing like the

the user-facing part of it. And once we had traction and raised some money, then I was like, okay, I need to hire someone who's very different than me. Because obviously I think in a very certain way. And I hired a

a brand designer that I'd worked with before, my friend Fi. I love Fi. I hired him mostly because we just get along creatively. In a startup where you have to be really decisive and just trust each other, that was my main goal. It's just someone I trust that I can communicate with and I don't have to worry about it too much. Well, it also turns out that Fi is incredibly talented and

took all of my ideas for the brand and just like took them to another level. One that I couldn't have imagined. Most of my role for the brand is just sort of creating a strategy and then trusting the team to bring it to life. And so my kind of team construction process

I think for the first maybe like six people was just building out skills that none of us had. Me being like an engineering designer, then I hired a brand designer, then we hired someone who was like a really great kind of UI, you know, design systems person.

And then I hired a clone of myself, my friend Pete. We think the same way. We work together. He's a better engineer than I am. I like kind of assembling. I would say it's like a group of misfits, like people who are generalists in some form, whether they're designers who are also great engineers or designers who could do brand design if they wanted to. We have a...

single design team. There are brand designers and product designers, but we're all just one group. And so fast forward to now, there's the design team's 22 people and everyone's a weird shape of skills and background and that's all very intentional. But I would say the archetype is, I really want

people who are flexible in their skills, kind of used to either they were like a founding designer or a solo freelancer, or maybe they worked at like a small agency, people that like have to are used to going really fast, being decisive, having conviction with their ideas, having the skills and the generalist, uh,

ability to just sort of figure it out. I really like that type of designer rather than someone who's sort of used to being in a big company where there's an existing design system, there's a product strategy that gets sort of trickled down to you. And it's something against those people, but the reality is we have to make a lot of decisions every day. And the best way to

to be able to do that is to have like the experience and intuition. And that comes from the experience of shipping lots of stuff. If you're working by yourself or you're in a founding designer role, you have to make all those decisions. And, you know, over the course of six months, you're going to have made a thousand hard decisions. Whereas a lot of designers that are in bigger companies, just the way that the organization operates, they maybe had to make five. And so that has led to like a very specific kind of hiring strategy for me is finding people because I don't want

The most important thing for us is that we launch 100 things every week. And the only way to do that is for me to not really be that involved in the work. So we don't have a weekly crit. There's no approval process. It's just divide and conquer. And so I need people who can live up to that and just generally do the best work that's possible. Because it's this balance of like,

what we were talking about earlier, one of us know what we're doing. We're making all this stuff up as we go, right? And the only way to figure it out is to just try stuff and think of all of it as very mutable. You know, okay, like, do we think this is the best interaction for the way that, you know, this experience? I don't know. I mean, I can give you feedback, but like, what do I know? Let's just try it and we'll use it together and then we'll react or maybe we'll launch it and look at the data, whatever. So there's, I think, a humility built into our process. Yeah.

This sort of combination of trust and humility to just figure it out and embracing the agility that's required to stay alive in this market. I'm going to go back to this point of agility required in this market. I just want to respond to some of the things you've shared here.

I can almost feel the surprise in the audience listening to this because we've grown up with this as designers, like my generation of designers, let's say 10 years into the industry. We grew up with this idea that once you get a big name, big brand name on your CV, then you'll forever find a job and everybody will want to hire you because you have Apple in there or you have Google.

or whatever. So this is like the nirvana dream. We were all communicated so far and now you come in and you crash that dream and you say, you know what we want doers, people who have worked in small, agile, fast, quick, challenging environments where you couldn't stay for one year to make a decision. You couldn't afford that. And that's really interesting. And I think it opens the door to a lot of talented people in the market who are overlooked because they, uh,

like I feel like it's the time of builders to shine like there's this new category of designers that's now succeeding in the age of AI it's not new it's a category of designers that are the new rock stars I really hate this word but I couldn't find a better one but this idea of the designer who can make create generate think dream imagine and then build that thing and so that's

And you are that archetype, right? And so it's really inspiring to see that there's finally the time has come for these people to shine who really deserve to shine. Because I think it's not that hard to push a pixel for six months and have 100 meetings a day and just put up with corporate structure. Yeah.

And then, yeah, call yourself a senior designer. I will say getting a job at Google when I was 20, 20 years old, that changed my life. You know, and I think anyone that can take that should probably take that experience. I think and it's really not a knock on. I have no knock on any designers. It's really just the way that these organizations operate.

right? And the role of design in the organization. It's quite different in every company, to be honest, but usually it's one of like, I would say disempowerment. Designers are used to sort of like put ornaments on the tree or

or you know visualize all the possible ideas that like leadership has and for me like a design is a problem solving exercise you know it's it's a practice of like making the business work better or making things more intuitive or or just like generally kind of making the world around us better through you know both like form and function you know when especially young people they're

if I could give them any advice, it's like, yeah, the best way to actually be good at that and have, and have experiences like make something and then try to get other people to use it. Cause you're going to learn a lot. Um, you,

You're going to learn a lot when someone says no, or if you get them to download your app or whatever it is, and they don't reach for it every day, that's incredibly humiliating, but also one thing to learn. And you got to learn that. You got to put all your ideas out there and find out how wrong you are until you can kind of find the truths eventually. And you just don't get that experience when you're in a big company.

Because all you're going to get is like feedback from other people in the company and maybe some data from A-B tests or whatever. And that can be, you know, obviously very valuable, but it just depends on how it's set up. I mean, it's so easy to learn the wrong thing. Most of the time, it's so easy to learn the wrong thing. Yeah.

And I think for many of us, it's a time to unlearn some of the things we've been taught. Yeah. That's really interesting. Adaptable now means forgetting some of the things you were kind of fed. And just a quick note on big tech to not be mean with it. I was interviewing with Apple. They reached out to recruit me after I had a South by Southwest talk and I seemed eloquent. Yeah.

And so they reached out and we were having this conversation about how their design team works. And I decided I can't move to San Francisco because I'm in Romania. So it's the other part of the world and I have a small kid and so on. But we managed to have a couple of conversations where they were talking about how they hire people. And I think it speaks to your philosophy a little bit as well. So they were looking mostly for interesting people.

that can come in and opinions at some point I remember having to ask that how do you actually make decisions and they were like oh no because these people are very senior and opinionated and they just run with it they just do it once the decision clicks for me that was a very different discourse from everything I've

learned again, like we have this process, these rituals, we have ceremonies, you have to go through this lengthy process before anything is shipped. And yeah, fluff in a way. And so you're closer to there is a big tech company that I think is doing things right, even though everybody's hating them these days.

I personally like Liquid Glass, so I think they're still doing great. So I want to make the most out of this conversation. I have two other questions I want to ask you before we wrap up. One of them is, and I think you touched briefly on it, if people are now moving from their big corporate role, let's say someone who's on a similar path as you are, how

How are the challenges different from working in this big company and working at the speed? So I think you mentioned that you really have to make decisions quickly. But what have you found changed fundamentally in the way you work today?

leaving big companies like Quora and Google and going in a small company when it comes to learning, especially? How do you now learn? You don't have all those resources, the research department. How do you make sense of everything when you're in a completely different environment?

Yeah, I think for the most part, it's ringing for the right person. I mean, I think some people feel overwhelmed by the demand of it, the pace. I mean, you know, we're also like a very prototypical San Francisco community.

startup where, you know, we probably work more than we should. And, you know, we're all tired all the time and, but also like excited and ambitious. And there's these kind of like human sides of the shock that I think people have to get used to or decide that it's not for them, which I think is totally reasonable. I think then there's another part of it. I mean, it's all the soft things, right? There's the other part of it. It's just every company operates differently and every company has a bit of a different culture and, you know,

how decisions are made, how leadership is, how intense people are. And I think startups are more unpredictable in that regard, like because it's really all of that comes from maybe the first 10 people. And so it's like whatever that composite is, you're going to get a different outcome for real. Whereas, you know, like the difference between working at most bigger tech companies, they don't feel that different in the end.

That's a big part of it. I think that kind of on the design side, I think it's mostly, it's like kind of all very freeing, but you're in an environment, it very much depends on how important design is to the business and how self-aware everyone is about that. Because a lot of businesses, it's like, hey, just design is just table stakes. Just make it look parity with the market so that we can find another angle to win.

And I don't mind, as long as there's an honesty about that, I don't mind that being true for things that I've worked on. I'd like to think I can push it a little bit beyond that. But then there are products and companies where design is going to make you win. And perplexity happens to be one of those, I think, because a lot of the opportunity for us has always been, how do we take this technology and make it intuitive and present it in a more colloquial manner so that people

people can fit it in their lives without feeling uncomfortable, both from a product perspective and a brand perspective. I do think that is like the design challenge of AI is like applying it in a way that works for real. And so that's true of our company. The challenge of that and doing that as a startup, especially for me as the leader of the team and the company is getting everyone to buy into that, right?

buy into how design should work and how decisions should get made and when to trust us and like just let us figure it out whether it's product decisions or brand strategy or whatever it's the fun part about having to make all these hard decisions every day is like a lot of times you're wrong and just sort of having that humility and eating it or having to fight for it even if you're really not sure but we need to find out anyway so there's a lot of like ownership is a double-edged sword

like real ownership and real agency. And that can be really exhausting, but I think it's worth it. It's, you know, it's sort of closer to a creative enterprise. You know, if you're writing a song or something like there's infinite ways to do it and you never really know when it's done until you think it's done and someone may listen to it and love it or they may hate it. And, you know, there's so much confidence you have to have and, and, and,

And it's just sort of like, you have to be fearless. So there's a lot more of that, right? You're really unsafe all the time, both creatively and, but I like it because when it works, you're like, yeah, I had a hundred ideas and we've shipped them all and they're working and I'm a genius, you know? So that's the duality of it, which I prefer because I hate feeling like a cog in a big machine. I hate.

like having a bunch of ideas and not getting to try them. I don't like working on other people's ideas if I don't think they're right. You know, this is very much like my personality. So I think joining a startup really needs to be, and certainly founding a company, needs to be like almost like you have to be spiritually tuned for it. Because a lot of times it sucks.

And most of the time, you know, they don't work. Most of the companies I've ever worked on don't. Yeah. They're just total failures. I was a founding designer myself and we raised, preceded one million something, one and a half million and we spent it. Yeah. Yeah.

So I can speak to that. Not all startups work. Anyhow, Henry, the last question is a very controversial one. The pop magazine question is design dying. Will design exist in five years from now? What are your thoughts around how this world is changing? Will we all become creators of stuff?

we're just builders and the design as we see today is not going to be a thing are you thinking about this or is it just not interesting for you as a topic like what are your thoughts about the future of design and how we can prepare if we want to be designers and in five years from now well i do think there's this there's a funny thing of everyone wants to ask people that work in ai what the future holds and what i find annoying is that most ai people answer confidently i'm like

How do you know about the future? I mean, the world is this like tapestry, culture, technology, politics, whatever. Like, why does like, whatever, why do AI people have the hubris to think they know about the future? But usually they're talking about like bigger stuff than just like design. I can comment on design. I mean, for me, I feel absolutely zero anxiety about any of that. You know, this job, we made up this job like 20 years ago, you know,

Why wouldn't it change? It's like all new anyways. And, you know, the titles change every five years as well. And it's all semantics to me because in the end people...

people are going to pay for software. Markets need like different choices. And when you're making software, you're making any business, you have, somebody has to make a lot of decisions about what it is, how it's going to work, how it feels to use it, whatever. And the reason we make those decisions is because those things actually affect the market outcomes, right? Like when people buy a car, they're thinking about

How does it fit into their lives? Who do they become when they buy that car? Is it my BMW guy or a Mercedes guy? You know, that's this like total composite of brand and product and engineering. So all of that's going to always be true. You know, maybe the thing that's powerful about designers is especially, I mean, I really don't like to...

force people into archetypes too much, but often people who come from a creative background, they're more comfortable in a situation where there's infinite possibilities and you just have to decide something because you think it's right. Maybe you can call that taste. Designers are very comfortable with that total ambiguity. And there's a lot of that in making a business and making software. And there's going to be more of that because what's happening is it's getting easier to build

anything and particularly dumb ideas, right? Like there have always been, you know, for every one good idea, there's 99 bad ones. And if it's easier to build them all, you need a lot of editing. You know, usually it happened before the building started. Now it happens afterwards. And I think that's like, you know, who better to kind of

take that on than design and and i actually think that ends up doing the best thing for the user because i think a lot of things are probably ruled out before they're trying to get built in the first place and so i think it'll all kind of average out in a way that's great to be honest obviously our tools will change but like yeah somebody needs to make all there's a lot of decisions to make every day in a business and a startup and the creation of software whatever someone's got to make those decisions maybe companies will get smaller but i think there will just be more companies

I think there should be. Just like take us, for example. Why is Google the only search engine? Like there's no other market in the world where there's one choice. So what the hell? But you could take that line of thinking to a lot of companies that sort of trick the world into thinking like you can't compete with them. So that would be my call to action for all the designers that maybe are thinking about starting a business.

This was extremely interesting and nourishing for my brain. I feel my brain feels good now after this conversation. And I really appreciate your positive and optimistic take without being naive or markety or salesy. Because I hear these two versions, right? The hype, AI is going to be great. It's going to create jobs. It's blah, blah, whatever. And I'm tired of it. And then there's also the...

catastrophic take. It's going to kill us all and whatnot as an extreme, right? Harari vibes. But I really, really appreciate this. I wouldn't even call it middle ground. I feel it's positive, but without being over promising or marketing, it felt very authentic and I resonated.

with it. I feel the same things. I couldn't express them as nicely as you. And I kept thinking throughout our conversation, oh my God, this is something I had on my mind in some format that I couldn't surface it until now. And yes, yes, exactly. Like I nodded all the time.

people can't see the video, but I nodded like all the time. Henry, thank you so much for making the time to join us. It was really, really compelling and interesting for me and like a bootcamp in how to think about building products in the age of AI.

Just to add that thing. Thank you for having me. Yeah. Okay. And to everyone who's listening, thank you for making it till the end and submit your topics and guest suggestions for future episodes. And yeah, just embrace the new age we all live in. And it's fun. It looks very fun now. Thank you, Henry. Bye.