All episodes

61. Josh Miller - A Good Rectangle

·Founder
Read transcript ↓Browse chapters ↓
Listen to the conversation
Watch the video

Related conversations

Description

Josh Miller (X, LinkedIn) is co-founder and CEO of The Browser Company, makers of Dia and Arc. The Browser Company was acquired by Atlassian last year.

Before The Browser Company, Josh was an investor at Thrive Capital, Director of Product in the Obama White House’s Office of Digital Strategy, and co-founder of Branch, which was acquired by Facebook.

He studied sociology and has always been thoughtful about environments: institutions, companies, and digital and physical spaces. We talk about his effort to improve the space many of us spend our days in: the browser. We discuss the tension between Josh’s creative and commercial instincts, optimizing for feelings, and hiring unlikely stars. Josh reflects on much of the journey with Arc and Dia and why he’s still betting on the browser in the AI era, now as part of Atlassian.

I hope you are inspired to believe in people and leave your fingerprints on what you create.

Dialectic is presented by Notion. Notion is a collaborative workspace for your life’s work and a hub for all the context you and your team use to work with each other and with AI. New models are available in Notion the day they launch, and you can work with agents you build in Notion or any other agents and tools you already use. As tools change, Notion is the Switzerland in the middle that keeps your context consistent across all of them. Learn more at notion.com/dialectic.

Timestamps

  • (0:00) Intro: Josh Miller
  • (1:55) Start: Not Commercial Enough & Optimizing for Feelings
  • (9:14) Sociology, the Obama White House, Thrive, and the 5% Favicon
  • (15:38) Bowling Alone, James Turrell, and Roden Crater
  • (22:12) Betting on People, Leaving Thrive, and Where Computing Goes
  • (34:54) “Assume You Don’t Know” and What Was Sacred at The Browser Company
  • (41:58) Intuition, Road Trips, and Leaving Fingerprints on Things
  • (53:57) Building in Public, Pulling Back, and the Road to Atlassian
  • (1:03:23) The Internet Computer in the AI Era: Codex, Cloud Browsers, and Why Everything Is Converging
  • (1:13:30) How Dia Wins
  • (1:17:07) Agents, Muse, and the Red Giraffe
  • (1:22:43) Software Is Becoming Art
  • (1:32:13) Believing in People, Leaving Facebook for the White House, and What’s Next
  • (1:47:17) Closing Questions

Links & References

Show 54 more referencesShow fewer references

Transcript

1:55 Start: Not Commercial Enough & Optimizing for Feelings

Jackson: (1:56) Josh Miller, thank you for being here. This is a long time coming.

Josh: (1:59) It's awesome. I feel like I'm about to have a therapy session on this couch a little bit.

Jackson: (2:03) Only a little bit.

Josh: (2:04) And the guy behind the camera actually helped make the video when we told people that we were transitioning away from Arc. So, really great to see an old friend, but I kind of feel like we're about to get into it full circle.

Jackson: (2:16) Shout out, Mark. A big one to start: Do you think you have been too commercial or not commercial enough over the course of your career?

Josh: (2:29) I've got to lay down on this therapy couch. No, I mean, I think that's an easy one, to be honest. I think not nearly commercial enough.

Jackson: (2:35) Really?

Josh: (2:35) I actually think there's a good chance The Browser Company would have sold for a lot more money if I wasn't the CEO.

And I'm not trying to diminish how proud I am of the work that we did and I did, but we were a web browser company that pretended it was a consumer company, that made behind-the-scenes YouTube videos and got really excited when we made the Spotify icon.

There was one day in the office where I was kind of dancing to EDM, and then we just decided that we needed to make the little Spotify music player kind of dance when there's music dancing.

I'm not sure how commercial a decision that was, but one of the best compliments I think I ever received about our work was someone was like, "I don't know how you're going to beat Chrome, but you seem like you're having a lot of fun."

I remember just feeling like I've made it in some way. So, that's a long way of saying I am definitely not as commercial as I probably could and should be.

But I think that's also a little bit of the magic of why people love the work that we've done.

Jackson: (3:34) Was that something that you were conscious of throughout? Is that something that you're reflecting on now? Would you have done things differently?

And maybe the biggest question, which is just: what dispositionally about you, Hursh, the team, the setup, produced that?

We should also caveat, you guys had an amazing outcome and you're continuing to build a great business.

Josh: (3:59) I'm very proud of it. I don't mean to be too self-deprecating. I just think, honestly, the gifts I have are not necessarily aligned with the word "commercial."

But in terms of where it came from, if you look at our early team especially, so many of us had worked at Instagram, Snapchat, and Facebook. We were in many ways taught our craft of software at some of those institutions, which were fabulous.

But they all revolved around a very specific kind of dogma at the time, which is you really are optimizing for a metric. I'll never forget at Facebook one day, Chris Cox—unbelievably charismatic, great storyteller...

Jackson: (4:36) He’s the head of all product.

Josh: (4:37) He was chief product officer at the time. He's hopped around a little bit, but he got up on stage and gave this talk about how Facebook's a medium. Just very, very heady stuff.

At the very end—and it was very heartfelt, you felt like you were there for a purpose—either at the end or a follow-up talk, he would basically put up this big pie chart and it said "hours in the day."

It divided it by how long are you on the toilet and how long are you sleeping? And the implied goal was that our goal was to capture as many of those hours as possible.

Jackson: (5:08) Like the Netflix thing: "we're competing with sleep."

Josh: (5:09) Yeah. This kind of metrics-driven culture was extremely effective. For example, I was on teams that were focused on how much people are sharing, how many pieces of content the user is sharing every week.

We all learned a lot. But when you are building software that so ruthlessly optimizes for numbers, a big part of The Browser Company project was to say, "Well, what if you optimize for the polar opposite? What if we just optimized for how people felt?"

What was the experience like using a piece of software in the way that I'm sure a lot of artists or filmmakers say, "Hey, how do I want the person to feel using this?" That instinct and founding motivation compounded because we were The Browser Company of New York, so excited to be in a room together.

And then two weeks later, COVID hit. All of a sudden this "band of brothers and sisters" team was stuck in our own product, which was a web browser.

Jackson: (6:09) Yes.

Josh: (6:10) In this dystopian, weird state of the world, we were going, "How do we literally... we literally wake up, open our product, and stay in it for eight to 10 hours. How do we make this rectangle feel awesome?"

I actually sometimes wonder if we would have been a company that's so obsessed over how people felt if it weren't for COVID, if that makes sense.

Jackson: (6:27) You had to make the digital thing very real. You were so immersed in the digital thing, it was like, "How are we going to make this..."

Josh: (6:35) Yeah. Again, how do we not make it painful? Keep in mind what was going on sociopolitically in the world. There was all this—it's sort of similar to how it feels now. People just weren't feeling good.

So we took it upon ourselves to say, "All right, if we're going to be in this space every day—this office of sorts, your browser became your office—how do we want it to feel?"

Intrinsically, it came from seeing how we all got in the industry. "Twitter's going to bring democracy to the world," etc., etc. We were reckoning at the time, after President Trump was elected, with what world we are creating as software craftspeople.

That definitely was rooted in an exploration of what it looks like if you're not so ruthlessly optimizing for metrics. And keep in mind, at the time, the theory of making a web browser for desktop was that it was actually a consumer product.

Jackson: (7:27) Right.

Josh: (7:27) That was one of the key insights of what we were all there to do: "Wait, we're all consumer people. We've been thinking about messaging and UGC on your phone for our whole careers."

But if you actually look at the landscape, there are a lot of social network messaging apps competing with each other. But if you look at the products that your mom uses, your cousin uses, and you use, the web browser is actually one of the most consumer pieces of software. What if we made it feel like a consumer piece of software?

In retrospect, why did we end up doing that little dancing emoji on Spotify? Was it because of our background? Was it COVID? I think it's probably a mixture of all those things.

Jackson: (8:00) The web browser specifically... maybe, if I'm hearing you right, to put that a different way: the least obvious thing to do something "consumer" in is the desktop computer, at least at this stage.

So maybe it was an inversion of that: yes, mobile's the ideal place to focus, except for this one thing, which is that every consumer does use a Mac or whatever and they have a web browser.

Josh: (8:25) And in fact, the story of The Browser Company was that I was actually an investor at Thrive Capital. Thrive famously incubates companies in addition to investing in them—Oscar and the like.

This was an incubation I was doing as an investor because I would never, ever want to be the CEO or founder of a desktop browser company. Can you think of something more boring? I’m sure your listeners, if they don’t know our work, are thinking that is one of the most boring things you could pick to do.

But the fact that we think of it like that, yet you spend eight hours a day in it... what piece of software do you spend eight hours a day in? What room do you spend eight hours a day in?

When I realized that, we changed our perspective on the problem space and said, "We're just blind to how utilitarian Chrome is. What if we spice it up a little bit? What if we made people feel something?" But again, not very commercial in that regard.

9:14 Sociology, the Obama White House, Thrive, and the 5% Favicon

Jackson: (9:15) So maybe on that note, I think you studied sociology?

Josh: (9:18) Yeah.

Jackson: (9:19) Why do you think you... and you did a startup. I think you dropped out, right?

Josh: (9:24) Yeah.

Jackson: (9:25) To do Branch? White House notwithstanding, the shape of your career has been very commercial. You were a full-time investor for two years.

Josh: (9:30) Yes.

Jackson: (9:32) Was that obvious? Was there a different path for you? Why do you think, despite you—the reason I think it's such an interesting question for you is that you’re actually in this really interesting dance between creativity and commerciality.

Obviously, there are some people who are "max commercial empire builder" optimizers, and there are other people who are the artist who doesn't have a commercial bone in their body.

Josh: (9:55) Yeah. There have always been sort of three parts of me that I've grappled with and will probably always grapple with. There's the part of me that is, let's say, traditionally ambitious, that has looked up to business people and wanted to make a kind of business of some sort.

There's the part of me that has always wished I was an artist or an urban designer or something. And there's a part of me—my mom worked for city government growing up. My godfather was a city manager of Culver City and then became head of urban planning for Los Angeles.

So there's a part of me that's always kind of grown up in local politics, local policy, or community. I'd always tried to combine those into one job or tried my best. How do I find—we've talked about this on walks before—how do you find that perfect job that's perfect for you?

Jackson: (10:35) There was a point in 2023 where I was very searching, and we had a conversation exactly on this note.

Josh: (10:40) And I thought my job at the Obama White House was that. I combined—maybe not the business side—but I sort of found the perfect job. Then President Trump is elected. Everything's thrown off.

Jackson: (10:48) You were there until the election, basically.

Josh: (10:50) Yeah, I was there the day that he was elected. And actually, one story about President Obama that, to me, captures his character. After the election, we were going to do photos with the president in the Oval Office as a thank-you for the time.

The day that he was meeting President Trump was the day these photos were scheduled, weeks or months in advance. He, of course, had no idea that was going to be the day. Parents fly in, wear their best. And he didn't cancel on us.

He met with President Trump for the first time, then met with LeBron James, and then took photos with a hundred White House staffers and their families. He's such a good person.

But to go back to your question, one of the reasons I joined Thrive Capital after the White House is I was like, "I'm over this. I want to do everything the best and separate it." So if I want to do art, just do art, whatever that means for you.

Thrive Capital was a, "Okay, this Josh Kushner guy and their team, it feels like they're creative with capital in a way that I don't see elsewhere, and it seems to be working." They had kind of this style to them that caught my attention and made me curious.

Jackson: (11:55) They didn't...

Josh: (11:55) They had a website with nothing on it. But really, that was an experiment in what if I'm just really commercial, actually? And then I was messing around with other projects on the side.

I almost made this documentary with Josh Lee, who ended up running video at The Browser Company. I made this prototype called Text Journal, which was this kind of artsy UGC thing.

Anyways, at Thrive, it was around the time that Thrive invested in Robinhood. What was so fascinating about that deal, which my partner Kareem led, was what Kareem picked up about Robinhood's strategy. E-Trade is a public company. Charles Schwab—they couldn't really compete with Robinhood because their whole business model was charging for trades.

So Robinhood shows up and goes, "How about trades are free?" And you see this upstart—what are you going to do as a public company? To me, that was actually the first time I was really inspired by commerciality as a lens through creative work.

I need to tell that part of the story because the root of The Browser Company was the heart. It was the hours a day. This is almost like an office, a physical office. But it also was: Wait, why hasn't the browser changed?

You peel back the onion and it is: Google doesn't care about browsers; they care about search, and they really care about search ads. So really they saw Chrome as this big search box at the top, and anything that traded off against those metrics, they knew they'd be optimizing for.

If we can build a browser that is better and it monetizes—or is better at something that is not search and it's decoupled from search as a business model—they're not going to shut it down. Actually, a week ago they announced a vertical sidebar with favorites that looks identical to Arc.

Jackson: (13:39) I was going to bring this up.

Josh: (13:40) This is 2026, and we started the company in 2020. So the thesis was correct, which is it would take Google a long time to do things that trade off against search revenue.

Jackson: (13:49) Why does vertical tabs trade off against search?

Josh: (13:51) Anything that is—just any change. One story. It's a real therapy couch. We ended up working with someone who's actually one of the co-creators of Chrome.

Sundar, who's now CEO, was the PM, and this person who worked for us, that ended up—was the VP of Engineering for Chrome when he left, came to work at The Browser Company, told me the story.

When you hit Command-T without thinking, it's like you're driving. Let's say you want to go to Twitter.

Jackson: (14:18) Yeah, I go Command-T, T-W.

Josh: (14:21) Exactly. And so on the new tab page, there are these suggestions of where you go. "Jackson, you go to Twitter a lot. Maybe you want to just click this big button to go to Twitter."

And they realized that because you're going so fast, if you actually show the favicon or the icon for the website, it's a lot easier for you to pick up what it is and find utility from it.

So they just shipped an update to the new tab page where instead of just saying "Twitter," they'd show the favicon. People loved it; they started using it more. Overnight, worldwide search ad revenue for Google dropped 5%.

Jackson: (14:52) No way.

Josh: (14:53) 5% just because you added a Twitter icon. So you say, "Wow, what does a vertical sidebar do?" Well, the whole theory of Arc and the features we built were: Hey, this is your office. This is your space, Jackson.

It should look exactly like you want. It should have everything you need. You shouldn't have to go anywhere. Actually, we should do everything to keep you right here. That intrinsically trades off with your time and your attention with going to Google Search and finding stuff.

So I guess that's all to say, it's not quite fair to say it was purely commerce. It very much was an experiment in: Wait, could you start a company that was very much driven to do something very different and new, but was doing so under the auspices that there's a massive market opportunity if we are in the right place?

15:38 Bowling Alone, James Turrell, and Roden Crater

Jackson: (15:39) Where did the—just to maybe go back to the business ambition, artistic creative stuff, civic service stuff—it sounds like your parents certainly influenced the latter. Where did the entrepreneurial, "I can do and make new things" come from?

I mean, I think you went to Princeton. You studied sociology. Those paths tend to not lead to "drop out and do something new." They tend to lead to more consensus things.

Josh: (16:07) We're going deeper down the therapy rabbit hole. I was talking to my dad about this the other day. My dad and my parents—I'm so lucky—they are parents that just don't care that I sold my business for a lot of money. They've always been like, "Just be happy."

Neither of them are ambitious in the way that I have been. I used to be a big nurture person. I think so many of the decisions I made in my career and life were—I found tech through a professor named Robert Putnam's book Bowling Alone.

Jackson: (16:42) Really?

Josh: (16:43) I went to see him give a talk as a sociology professor—yeah, about Bowling Alone. And it was the first time I'd heard someone describe the world and this idea of social capital that explained everything I felt like I'd seen up to that point.

I went up to Professor Putnam after the talk and I said, "Hey, what should I do for a living?" And he's like, "I don't know you, but if you liked my book, there is this guy, Scott Heiferman, who started a tech company after he read Bowling Alone. So if you like my book and you're trying to figure out what to do with your life, maybe you should go work for Scott."

I interned at Meetup in New York City in 2011.

Jackson: (17:18) Wow.

Josh: (17:18) As part of interning at Meetup, you have to go to meetups.

Jackson: (17:21) Cause that's the product.

Josh: (17:22) At the time, the New York Tech Meetup was the center of tech in New York. Everyone went to this demo day. At that demo day, I was like, "These guys are my age. These gals are my age. I can make software." And the rest is history.

Anyways, going back to your underlying question, I have become, over time, much more of a nature person now. I have two sons, and you meet your child in the first year and you're just like, they're such a fully formed person. There's a spirit in there.

So that's a long way of saying, I think if I go back with my dad, I've always had this part of me that wants to make and do things. And I think it's been encouraged mostly by my parents in that no matter what you do, if you fail, we love you. Just be happy.

Which, keep in mind, I'm from the west side of Los Angeles—very hippie-dippy parents. But my instinct is it comes more from just how I came out as a kid.

Jackson: (18:19) You said to me on the phone recently, "I'm not particularly an artist." But clearly, this is rolling around. Also, was that just interest in art? Was that making stuff?

Something I want to talk about later is how really talented, creative people—videographers, designers, whatever—are drawn to you, or at least you're good at identifying them. I think your wife, if she is not an artist herself, works with artists. Is there a root for that thread?

Josh: (18:49) There's definitely a—I say a lot that in another life, I'd be an artist. And someone once said, "Well, why don't you become an artist? It's not too late." And so I think deep down, there's some part of me that struggles to even say that word in this podcast.

Jackson: (19:00) Yeah, yeah, yeah.

Josh: (19:01) So we'll leave that aside. I think something more relevant is that we start The Browser Company in 2019 or 2020. And around that time, my wife gets a job running James Turrell's studio.

Jackson: (19:14) This is pre-COVID?

Josh: (19:15) This is pre-COVID. And then at the very beginning of—it literally overlapped with The Browser Company. And so on a dime, because of COVID, instead of Valerie commuting, we moved to Flagstaff, Arizona, in the high desert, up in the mountains.

Jackson: (19:30) It's basically next to the crater.

Josh: (19:33) Yes. For those of you who don't know the artist James Turrell, he's a light artist. People have probably seen his work, but what most people don't know is in the '70s, 50 years ago, he got a grant from the Guggenheim and spent seven months flying over the American West.

He's a pilot, and he spotted an extinct volcano and said, "I'm using my Guggenheim grant to buy that volcano." He landed his plane next to this volcanic crater, and he's spent 50 years and hundreds of millions of dollars, including his own money, on it.

That's what was so fascinating about James. He takes money he makes from celebrities buying his work and just puts it back into his work. And so as we were building The Browser Company in COVID, I was just sitting there with Valerie and James and his wife, Kyung, and going out to the crater.

The ambition of an artist to spend their entire life working on this magnum opus that he wants to be the modern-day Machu Picchu, pouring his own funds into it, obsessing over the type of clay that's going to curve the wall—which to me is like not an important wall.

It was hard to physically be around that in this moment when, in COVID, the world feels like it's falling. And he has just kind of this singular purpose and drive. He builds chambers. He's hollowed out this volcano with hundreds of millions of dollars, built all these chambers and tunnels that align with celestial events.

I remember once he was explaining this long tunnel, the length of a football field. It brings the sun and projects it onto this piece of marble that's a big circle. So it feels like Turrell took the sun out of the sky and put it onto this piece of marble. It's the most glorious thing you've ever seen. I was making awkward small talk with him towards the beginning.

Jackson: (21:11) This isn't open. People don't get to see it.

Josh: (21:13) No. I mean, hopefully soon, but it wasn't at the time. And then he made some comment like, "Yeah, I like that the sun comes here. I built this chamber because once every 18 years, there's this thing called the moon—" and every 18 years the moon's going to be projected over here.

I'm around this guy that's thinking about building experiences that are going to happen once every 18 years with big pieces of marble and football-field-length tunnels. It was hard not to be deeply inspired by that.

Not pretending like we're making things like Roden Crater, but if this man can dedicate his life to making people feel a certain way, and we're building this rectangle that you're in for eight hours a day, how can we bring some of that spirit into our work?

My instinct, in addition to just being a fan of art—I grew up in Los Angeles, I love going to indie films with my parents and museums—I think just being around James and his work for the early duration of the company trickled into the sort of people we took bets on and the sorts of things that we did.

22:12 Betting on People, Leaving Thrive, and Where Computing Goes

Jackson: (22:13) There's some idea that the founding DNA of a company really, really matters. And obviously part of that is maybe the way that early product kind of shapes with business model alignment—the more rational version of that. But I have to imagine that being a part of the DNA of where your headspace was in those early days mattered.

Josh: (22:34) Yeah, crazy. For example, one of the best hires—and it's worth noting for your viewers, I'm a sociology major—I think my main skill is identifying really talented artists in themselves and giving them encouragement and a platform to go do their work.

So many of the things we're known for were co-created by a guy named Nate Parrott. And Nate Parrott was our first design hire.

Jackson: (22:56) Where'd you find him?

Josh: (22:58) So our first designer was a guy named Nate Parrott, who had never been a designer before and was two years out of college. Keep in mind, Hursh and I sold our first company to Facebook. We knew designers. We could have hired someone senior.

This person who I didn't really know was like, "Hey, you've got to meet this Nate guy. He's a kid, but he makes browsers for fun." You kind of look him up on the internet, and Nate's website at the time was a video game arcade where you won coins by doing things like reading his resume.

It was very playful. You can go look it up. And I meet this guy Nate, who makes this really artsy personal website. He was just making browsers for fun on the side, not even designing them. He's just making them full-stack.

We meet him the first time and we're like, "This is our first guy. This is our designer." I think meeting people like Turrell, hearing their life story, seeing how quirky they are, seeing how nonsensical some of the things they say are but how brilliant they are if you look through what you're supposed to say in certain situations, it made us more open-minded to people like Nate.

On paper, he wouldn't pass a screen. But you just feel like there's something in there. Josh Lee, when we decided to bet on video—it's so obvious now, but this was in 2020. Why would you hire a video person? Are you going to make behind-the-scenes YouTube videos?

Part of it is the secular bet on video as a storytelling medium. And part of it was that Josh was just an N-of-one person.

Jackson: (24:24) And you bet on him. Were you looking for someone?

Josh: (24:27) I was looking to hire Josh. I'm someone that when I meet someone that grabs me like that, I'll wait years trying to recruit them.

Josh was someone that had interned for me. He basically came up after we sold to Facebook; he was a fan of our first company, Branch. He was an intern at Facebook and asked me to go on a coffee walk and he really stood out to me.

Then when I was at the White House, he wanted to come intern, so he interned for me at the White House as a designer. Then he decided that his heart was in film and he wanted to go pursue filmmaking. He was pursuing filmmaking at the time, and we did some side projects together.

I just felt like he was such a creative genius, honestly, that I just wanted him at The Browser Company. So the way that I got him to try it was while he was trying to make these artistic film projects—maybe he would not call them artistic, but I thought they were fairly artistic—I was like, "Hey, let us be the for-profit contract."

That was when we were making a bet on YouTube culture and how opaque and untrustworthy tech companies were at the time. So again, I guess since this is a therapy session, there probably is something to my work and our work as marrying both sides of the brain, or at least marrying art and business in some way.

It was led by the artist: Josh is special, get him in the door. But it was also observing that from a business standpoint, we couldn't succeed otherwise. People didn't trust Google. People did not trust Facebook.

Whether you liked it or not, in that cultural environment, how do we as a dinky little startup get you to trust us spending eight hours a day here with what you get paid to do, buying things? Why are you going to trust this little dinky startup?

Our view was that no one's perfect, but if we can invite people in to get to know us as humans and they see we're just a bunch of people in a room trying our best, video is the best medium to capture that spirit. People would feel like, "Okay, when they inevitably get something wrong or there are stumbles, I know these people."

Jackson: (26:22) Yeah.

Josh: (26:23) They're really trying. They're really trying hard. And that turned out to be true. But did Josh Lee come first? Did this observation about betting on video come first? It's sort of been both the whole way.

Jackson: (26:35) It's funny. I worked with all these Twitch streamers and YouTubers, and I always used to joke—I have very wealthy friends and if they bought a Lamborghini and posted a video like, "Check out my Lamborghini, everybody," people would be like, "You're such a douche."

But if Nadeshot or one of these YouTubers did it, it was like, "That's our guy. We did it." And I think you guys have been influential in so many ways. Even the name has become a little bit of a meme now, but I think the work you guys were doing, especially early on, is how the rest of Silicon Valley works now in a way that is not obvious.

I also think it's worth noting that the answer you just gave is the opposite side of the coin of why the reaction with Dia was so strong. It was actually the same thing. It was the same energy that was working for you working against you, 100%.

I want to talk a little about the activation energy of the company. I think maybe an important place to start with that, especially given your previous comment about maybe having your commercial business life as an investor and your creative life compartmentalized.

I talked to Josh Kushner about this once and he explained being really bummed when you left, in part because you felt like you were really starting to hit your stride after taking a little while. Why do this at all?

Investing is a great job, especially if you're really good at it at an amazing place like that. Why abandon it all to try to smush all these things together into one container?

Josh: (28:10) It's funny, going back to the nature versus nurture and the sense of who you are. I'd always been a big planner, or tried to be, about my life and what was going to happen.

At that point, you looked at everywhere I had lived and everything that had happened in my career and none of it was part of the plan. I think I've just learned over the years to just sort of trust my intuition.

I wish I had a very rational reason why I left Thrive Capital to be a CEO and co-founder of The Browser Company, but it really was just an emotional bet on the people. I started the company with Hursh Agrawal, who is my second-time co-founder.

Hursh is like my professional wife or husband; there's something about us that chemistry-wise has always clicked. I'm actually going to blame Josh Kushner. The most I can think of as a reason I left Thrive Capital was when I joined, I thought I was going to start a nonprofit.

Jackson: (29:04) Right, because you were an EIR at first.

Josh: (29:05) I was an EIR, an entrepreneur in residence. Josh was very kind and generous and he said, "Hey, come to Thrive Capital. You start a company, you start a nonprofit, you want to join as an investor—we just want to work with you."

Actually, I think he in some ways inspired the bet on people I've made because he said, "Let's not figure it out. Just come in the building, we'll figure it out." I joined after the White House gig and thought I was going to do something with civic tech.

I brought this idea once to him—I don't remember exactly which it is—but I think it was to make a babysitting co-op company. Josh, very much in my memory, acknowledged how generous the idea was and how much it could help people, but it was very small.

Josh said, "Small ideas are just as hard to make successful as big ideas. You should go for it, Josh." And so the main motivation for The Browser Company was this instinctual desire to work with Hursh again and to build a place where our people could come and we could work in the way that we've decided we really wanted to.

But what really motivated me and put me over the edge was realizing this is a product that billions of people use. They spend eight hours a day in it. They're using it more like an operating system because all the apps that you use are now in browser tabs.

That means it's sort of the new macOS or Windows. We had learned that actually operating systems held a lot of the leverage in the technology industry because Apple just cut off Facebook.

It was sort of like, "Wow, we're building a web browser, but actually, if this works, this is going to be my future children's sense of what a computer is." You go to it and you go, "This is worthy. It's going to be really hard."

Jackson: (30:46) If you're going to do something new, you might as well make it hard. Make it really big.

Josh: (30:52) Yes. And so there is, I think, if I'm being honest—whether I'm proud of it or not in 30 years when I listen to this: If this all works, it might truly change the way we use the internet and our computers.

And that combination of the people and the audacity—again, Roden Crater. Valerie just started working for James Turrell or was thinking about working for him; I think she had started by then. It was just this audacity, like, let's just try really new, exciting things.

Jackson: (31:18) You said something to me in that conversation in 2023 when I was really searching. Speaking of it being big, I think you said it was deep but it was also wide, which is a really good fit for me in the sense that it was this container that would allow you.

I know you're pretty skeptical of crypto, but you even made a comment like, if the crypto thing actually works, that could be part of this. It was able to be an aperture that was wide enough.

Josh: (31:48) And this is one of the things I really learned from Zuckerberg at Facebook. I will never forget this internal talk he gave. I don't remember the size or the purpose, but in it he made this proclamation that he was sure that in five to 10 years, everything on Facebook and Instagram was going to be video.

Jackson: (32:07) Okay.

Josh: (32:08) And I was like, "This disconnected billionaire. What a fucking idiot. How rich are you? Have you tried Buffer?" Does he have some sort of billionaire Wi-Fi? Like, have you tried watching a video on your phone?

Jackson: (32:23) He's got Verizon premium.

Josh: (32:24) And what do you think of us plebeians? You think that we're going to want to watch on our tiny screen? Why would you not watch it—

Jackson: (32:29) What year is this? Like 2013 probably?

Josh: (32:31) 2013, yeah. It must have been when we joined, and I remained skeptical after the talk ended. He didn't convince me. I was just like, "This guy's disconnected." And lo and behold, he predicted TikTok and everything that came out of it.

I was really inspired by the way he made this bet. There were a lot of things he did in response to that thesis that put Facebook right next to the hoop to dunk the ball when they needed to. And that carried through with me. So when I was at Thrive Capital, the adjacent possible...

Jackson: (33:00) You're in the water.

Josh: (33:01) Yeah. And just for me at the time up until that point, it was mobile, mobile, mobile, mobile. My career—the iPhone came out when I started my career. So my whole career in software had been like, everything's gonna be on your phone.

So I go join this sexy VC firm and see all the companies come in where these productivity tools are in your browser. And so it was part of observing, okay, all of the most talented people I know in the industry are all of a sudden building these desktop web apps.

There are all these people doing crypto and saying computing is going to be rewritten by crypto. Now all of a sudden we're using Chrome extensions like MetaMask, and Chrome extensions are cool all of a sudden.

And then there were all these nascent AI ML companies. Obviously didn't know transformers were coming, but that required cloud servers and cloud compute. And so you sort of just looked at...

And then you look right and my wife is working for a 78-year-old artist who is not tech savvy in the middle of the most rural county in Arizona. And she was not leaving her browser strip because some gallery sends a PDF and it links to Dropbox.

And so you just look around you at the micro and you look up at the macro and you go, "I don't see how in five years the web browser is not one of the most strategic pieces of software." It's not going to look like a web browser today, but it feels like there's no chance with all of these trends that the browser won't be more central.

And so it's kind of surreal in that one of the questions I get the most is like, "Hey, everyone is embedding a browser and trying to take over the browser." And people have different theories like, "How are you going to win in this extremely crowded space?"

The jury's still out on our role in the ecosystem and how successful we will be. But in many ways it was the same thing I saw from Zuck saying, "I promise you video is going to be central." And it inspired that kind of long-term thinking, which is not normal for me. I can be reactive and short-term sighted, and I think I observed from him the compounding of patience and conviction.

34:54 “Assume You Don’t Know” and What Was Sacred at The Browser Company

Jackson: (34:55) Yes. One part of it too is like you kind of need to spot something on the horizon. I don't even know how much specific resolution there needs to be, but it's like, "I want to run over there."

Josh: (35:06) Yes. And it was also probably a reaction to the failures of our first company. We stumbled our way into an acqui-hire that made our investors money going to Facebook, but in retrospect I'm embarrassed to call it a company.

One of the challenges with that company is if you're 20 at the time, you try to answer all the questions. "Okay, what's the business?" You check all the boxes and you have this plan. And then whenever the plan doesn't go well, you're sort of like, "Okay, well what's our North Star?"

Our North Star was this Harvard MBA product proposal plan thing that had all the questions answered. And it really changed my worldview. If you go read the essay we wrote in place of a value statement, we wrote this essay about going on road trips.

One of the core values is just assume you have no idea, assume you don't know. And so we needed to go play in an area and direction where as our ideas worked and didn't work, and as things played out in the world that we didn't anticipate, we would be in the right place.

Our conviction would be at the highest level because we knew we were going to get it wrong. And so that was probably honestly learning and evolution from being 20 and trying to control the world.

I'll never forget, for Branch, we moved out to San Francisco. My first company, we started in New York and then this guy named Ev Williams offered to lead our seed round.

Jackson: (36:28) Wow.

Josh: (36:29) And Ev kind of had a similar... in the same way James Turrell has just worked on light his whole career, Ev has worked his entire career on putting thoughts on the internet. He was the one that popularized blogging.

He co-founded Twitter and was CEO of Twitter, and then he went on to Medium after that. And each successive point he just removed friction in publishing. And so we were doing a publishing platform, that was what Branch was.

Ev Williams wanted to be our lead investor. And not only that, he had just left Twitter and he wanted to teach us. He was like, "I'll help you. Come move to San Francisco. We'll work across from each other."

I'll never forget, we move out to San Francisco. I barely know Hursh Agrawal, my co-founder. I don't know Ev at all. We're going out there to the promised land so he can teach us everything.

And in the first meeting with him, I prepared this whole thing on the plane: "Here's our plan for the next five years." Keep in mind, we moved everything to go work with Ev.

In the middle of this meeting, he stops me. "Josh, I have no idea what the answers to these questions are. You are mistaken if you think I have any. I have no idea. And by the way, you just moved out here. Beware of anyone you meet in Silicon Valley that purports to have the answers."

Jackson: (37:36) ...to all the questions.

Josh: (37:37) Because all of us who had made something like Twitter know we were all making it up as we went along, and we all got really lucky. Which doesn't mean we're not talented and doesn't mean we don't have foresight. But just—and so I almost feel like I didn't really take his advice as much as I should have.

But years later, starting The Browser Company, I wanted to sort of—how do you set up for success if your philosophical worldview about how business or industry works is that no one has any idea, even the greats that have done it two or three times?

So I was not wise and mature enough to internalize his feedback then. But in retrospect, I think he preempted everything I've come to believe about what we do for a living.

Jackson: (38:18) So this prompts two interesting things to me. One, our mutual friend Chris Paik, who is also an investor in Browser—I don't know how much you guys have talked about this, but he has this frame of companies: there are bottom-up companies and there are top-down companies.

His frame is basically that bottom-up is most of what we think of in software. It's like you start—and frankly, like Branch—you start with a single feature thing and it just... or Twitter, it just grows forever. And in some sense, it's very much the YC "just iterate your way."

Josh: (38:46) Startup machine.

Jackson: (38:47) Yes. Vision doesn't matter actually at all. Just iterate until you find the thing that catches fire.

Top-down, on the other hand, which is much more commonly associated with a company like Apple or Tesla—it’s more rare in software—is where you're shipping multiple attempts at a vision, basically.

And you are a classic example, at least with The Browser Company, of the relatively rare top-down software company. I'm curious just how you relate to that frame broadly.

And then two, did you waver? Maybe a better way of asking the question was: was there anything about the product of Arc and then later Dia that was sacred, or was it all just in service of this vision of, "We need to build whatever is going to be the web browser in five years"? Does that question make sense?

Josh: (39:25) Yeah. And I think we can be criticized here for not keeping more sacred, but there was nothing other than what we came to describe as the "internet computer" vision.

Jackson: (39:36) Yeah.

Josh: (39:37) So if you go back to our YouTube page and go all the way back to the first video, it was me sitting in a kitchen basically just riffing.

Jackson: (39:44) Recording through Photo Booth, I think. Right?

Josh: (39:45) Yes. And actually, at the time, it was an update to our hundred people that use the product. I showed some graph that was like a hockey stick and—I watched the video this week—the n was like a thousand people. But I'm like, "We made it, man! This is how it feels."

Anyway, the vision, which I've touched on, was that when my son—at that point we had our first child and he was, I think, two, maybe one—it was like, when he's 10, I want his... I got an iMac when I was 10 and it changed my life. I wanted his blue see-through iMac to be the product we were making.

Jackson: (40:21) Wow.

Josh: (40:22) Because our belief was computers would end up spanning all the devices in your life and sort of be this invisible layer that sits on top of it. And that's what we were making.

That was sacred. That is sacred. That's why we pivoted from Arc to—that explains everything we've done. And I think different people at different times that even work for us have believed that or not.

I try to be really honest about who we are. We put our values up there. We write an essay called "We Optimize for Feelings." We say, "Go watch the internet computer video."

So I think you make a product like Arc, which is a cult hit—let's call it indie software. I’m not sure of the label, but people love the thing. They don't care about this random YouTube video.

They actually think you're kind of a blowhard talking about all these proclamations, not knowing all these stories that informed it, whether it's Turrell or Zuckerberg. This long-term bet on where we wanted computers to go was why we were here.

I don't care about some little feature in this product, number one. But actually, you make a whole brand and you make a whole video series and you make a whole storytelling machine around, "Here's who we are, here's what we're working on. We're going to geek out to you about this little craft detail animation and our onboarding flow."

Jackson: (41:24) You get attached to it then.

Josh: (41:25) Yeah. Then you remove it and people go, "You're abandoning your roots."

Jackson: (41:29) Yes.

Josh: (41:30) And I go, no. We're actually staying true to what we came here to do.

Jackson: (41:35) That explains a lot.

Josh: (41:37) I was imperfect in many ways during that period. But to your question of what was sacred: very little, if anything, was sacred at the implementation product layer.

But these foundational roots of the company are extremely sacred. And I think once you view it through that lens, agree or disagree, it explains most, if not everything, about what we've done.

41:58 Intuition, Road Trips, and Leaving Fingerprints on Things

Jackson: (41:59) Absolutely. You mentioned intuition earlier. You seem to be, to me, a very intuitive person.

Josh: (42:07) To the max. For better or worse.

Jackson: (42:09) I'd love to hear about how that maybe squared with this idea of "we don't know." There’s even a core company value of, "We don't know the answer. You can only learn from experimenting." It's clear to me that those two, intuition and that, aren't at odds, but I'm curious how they interplay.

Josh: (42:28) Yeah, 100%. So "we don't know" doesn't mean we don't know anything. "Assume you don't know" just sort of says, "Hey, what's going to happen with AI safety? Is a bioterror attack about to happen?" No one has any idea.

But one of the reasons we were so motivated to start The Browser Company was because it meant we would use our software every day. Turrell lives in the art spaces that he makes.

And so if we woke up and, in order to do our jobs, we spent eight hours in our product, and we thought we were craftspeople that loved what we did and were striving to get better at it so we could make something for ourselves...

A big part of the browser was it meant I could go into a dinner party and know everyone there used a browser. That meant, theoretically, they could all relate to what we were making, and I could explain to them at a very human, personal level why I thought it would make them really happy or why I thought they'd love it relative to Chrome.

So we assumed we didn't know at the macro level or at the strategic level or at the industry level. But when we woke up and were like, "Okay, should we make the dancing Spotify icons dance this way or that way?" we had a lot of conviction in what we wanted to do because we were making it for ourselves.

We knew it might not be right for everyone, but we knew what we wanted to do. And so I think one of the secrets to The Browser Company's success or innovations—whatever you want to call them on the product side—it was probably someone at our company just said, "I really want this for myself."

So it's extreme trust in your intuition and your artistic sensibilities, for lack of a better word, at the day-to-day level and complete acknowledgment that you're not in control and you have no idea what's going to happen a couple layers above. That marrying of a macro bet that it's going to go generally that way—this is the road trip metaphor.

Yeah, but when you're on a road trip and you're on some interstate, you're like, "This doesn't feel that. Let's take that exit." I see you trust your instinct. Take that exit. You're on the interstate, and you know you're gonna get more Dunkin' Donuts if you go off that way—maybe you'll get a good tamale. Go for it. Just trust your instinct to go that way.

I know you travel a lot. I'm sure you've dropped into some random city. How do you know which way to wander for an hour? You don't. So I actually think once you think about it that way—you want to go walk around Amsterdam for three hours, you don't know which way you go, but you've walked around enough cities that you trust your instinct about what street to go down, which coffee shop to go in.

Jackson: (45:03) There's an element, too, I think, of giving way to tolerate other people's intuition more and thus be more experimental on the intuition.

Josh: (45:12) It's so fun because all that means is we need to try a lot of things, go on a lot of adventures. And also when we feel like we're starting to have conviction in something, double and triple down on it. So really, it pushes people to do the work for us. It's just like, "Stop talking."

Jackson: (45:31) Go make it.

Josh: (45:31) If you love it, give it to a few of us. If we love it, then let's all run towards it.

Jackson: (45:37) Yes. How do you train, maintain, stay connected to your intuition personally?

Josh: (45:48) I think just continuing to explore new things in my personal life. I've always kept personal and professional lives very separate. And actually, there's a reason we wanted to be The Browser Company of New York.

If you go and read the "Optimizing for Feelings" essay that we wrote, the main point we're making is for technology—yes, we're making a point about metrics versus feelings, but one of the things we really argue for is, hey, go look at architecture. What can we learn if we're making software?

This browser, I spent eight hours a day there for my job. I work with colleagues in it. Like this Google Meet tab, I do focus time on what I'm paid to do. That sounds a lot like an office. So instead of pretending like we're making this browser thing and the only reference is Google Chrome, and maybe if you go way back, Mosaic, we sort of said, "Hey, what can we learn from office design?" What can we learn from the offices we've been in?

I think we've always drawn inspiration from industries that sort of sit adjacent to software and technology. As a result, in my personal life, I actually stay pretty far away from the tech world physically and spiritually, and try my best to just explore new things. Which sounds very generic and boring to listen to, but it's the truthful answer about just staying curious, I guess.

Jackson: (47:04) What do you think people loved most about Arc and about Dia?

Josh: (47:14) There's this wonderful woman that ran our storytelling team—and that doesn't even do justice to what she did for us—named Nash. Nash is another great example. She's getting her due, and she is brilliant—one of the most brilliant storytellers, hopefully future poets, that I've ever met.

Nash was a venture capitalist when I met her. I remember telling her, "You're in the wrong—what are you doing?" You could just tell. Nash had this line that she started saying in her final year or two at The Browser Company after a bunch of years, which was, "We leave our fingerprints on things."

There's just this idea that if you're having fun and you're making it for yourself and you're expressing your own intuition, it's going to be imperfect, it's going to be funky, it's going to be playful, it's going to be whatever it is you are and what you're trying to do. And I think people can feel that.

I live in a part of Brooklyn where all these condo towers are going up. If you look at a photo, it's like, "Oh, it looks cool," but then you go in the lobby and it just feels void of any soul. So the best I can articulate is that we were genuinely having fun. We've really optimized for that, and I think you can feel that team having fun behind it.

We're a culture that encouraged people to kind of leave their fingerprints on what they did and not try to do it The Browser Company way, but do it the Omar way, or do it the Carla way, or do it the Kristina way. Kristina, who designed the scrim border around Arc—Kristina Saint.

Jackson: (48:49) Yeah. I was just with her in Maine like a month ago.

Josh: (48:52) So Kristina I met at Facebook. She hated it and she decided to quit tech to become a furniture maker. She was doing contract work in order to pay the bills at furniture making. I met her and we just hit it off. We were talking about Christopher Alexander.

Jackson: (49:11) She mentioned Seeing is Forgetting. I think you showed her that.

Josh: (49:16) I reconnected with her after five years as this tech contractor. We went on this rabbit hole and I walked her back to my apartment to give her Seeing is Forgetting about Robert Irwin. I was just like, "Someone that wants to make furniture with wood in Maine is going to make a really great browser." I want my browser to feel that way.

So I think, again, it's so easy to add a narrative in retrospect. Was it that we were having fun? Was it that we hired people like Kristina that, lo and behold, quit two years later because she went to Maine and made furniture and is now quite successful at that?

Is it because we encouraged people to not do it The Browser Company way, but do it the Kristina way? I'm not sure, and I think Kristina would answer better, but I do think part of what people can feel is the craftsperson or artist behind it, doing their thing, not doing it the way they expect. By the way, huge downsides to this, too. It definitely has cost me—I've got a lot of grays now—but it was definitely worth it.

Jackson: (50:14) That's an amazing answer. It wasn't quite what I was trying to ask, which I guess is: are there distinctions between what you think people particularly loved about Arc and what is especially great? Maybe that's a great answer for your overall philosophy on both products, but is there a distinction between what you found people really loved about Arc and love about Dia?

Josh: (50:34) It's hard in retrospect, and especially being so close to it, to analyze what made Arc successful. But in some ways, it felt a little bit like we were building it together. It was in Covid.

We were very public and open about our experiments and the people we hired and our board meetings. We shipped stuff, got feedback on it, and changed it based on feedback, and sort of broke down the wall. I think there was this feeling for people that used Arc that they were part of the story.

We truly were—it did feel like we were building with everyone. That was the intention, and the downside was that, as Valerie said—I remember there was once I started getting a lot of hate online after we abandoned Arc. It sort of crossed into some thresholds where it was making me really sad at a personal level.

I couldn't expose that at work in many ways. Valerie was just like, "If you are someone that leaves YouTube comments—I've never wondered why you were so happy people were leaving positive YouTube comments."

If you are leaving YouTube comments on a video about a tech company's board meeting, you're not representative. She was saying this so I wouldn't let it get me down.

For a bunch of structural reasons about who wanted that sort of content, who wanted an indie software browser, and who wanted to root for this team, we ended up making a very niche but beloved product.

With Dia, what we've successfully done is build a product that can speak to the average person using their laptop for their job—the person who doesn't know which—to go back to what was actually the original vision.

Jackson: Which was actually the original vision.

Josh: (52:08) Yeah. I was thinking about this—Dialectic is one of my favorite podcasts. Oddly, so many of the people I look up to and have looked up to in my career have been guests on this.

I was thinking about how to explain to someone why this person's work—it's an inside world, or a specific subculture of sorts. Dia is really our attempt to be that mass market. How do we take the sensibilities that led to Arc, but put them in a package?

We hire people like Charlie Deets, who invented "swipe to reply" on WhatsApp and was the lead designer of Safari for many years. His designs don't feel like early Browser Company in their color and their flair and their animations.

Some people like that, some people don't, but it's a very strategic and intentional move to say, "How do we get next up the ladder? How do we distill something down?" It's like the iPod click wheel. I don't think we've done it yet.

The people that love Dia, we hear things like our Morning Brief. I believe we were the first product to have something like a daily brief, and we called it the Morning Brief. We leaned into the editorial design and we don't say AI anywhere.

We have these colorful, juicy buttons, and we started hearing from people that it was the first time they understood how AI could help them in their day-to-day.

Tech people and AI insiders generally gave respect for the novelty of the packaging and the interface around the harness and the models, but they looked at it and said, "I can ask ChatGPT to send me a task reminder every morning and it can do that. What's so different about it?"

I think we've kept some of that spirit, but we're really striving to build something more essential and culturally relevant than this YouTube-video-driven mode.

53:57 Building in Public, Pulling Back, and the Road to Atlassian

Jackson: (53:58) Well, also, the person who uses Arc is like, "I can't wait, I want to mod my thing." These things are—yeah. It seems that you've done less of it—I don't think you've done a podcast in a while.

At the very least, the building in public thing broadly—which again, I think is so much more common now. Do you have any reflections on that? Is there a reason?

Obviously, part of that, I'm sure, is you're part of a bigger company now. But maybe one mode was more suited to Arc and another mode's more suited to the current product.

Josh: (54:32) We are redesigning our office for the first time. We've had this dinky little office in Williamsburg since the middle of COVID, and we're renovating right now. We just moved into a coworking space across the street.

We're back in a coworking space after this big, splashy acquisition—we're back to our roots. Like any coworking space, we've got this little see-through door with our Browser Company logo on it, and right next to it is one of those companies that's called "The Blank Company."

It's unreal that we just happened to be put next to some startup called The Blank Company. I'm actually really proud to the extent that it inspired people, but we called it The Browser Company of New York because that in itself was a cultural reaction to the tech startup ecosystem at the time.

It was so not the way—like Ziply or whatever. We were just having fun. Let's call it—we're in New York City with these Gothic buildings—The Browser Company of New York. You see the Paul Rand logo; it was a joke of sorts.

When we started doing behind-the-scenes YouTube videos that were not polished and were raw about what happened at our board meeting, it was us reacting to the culture in technology.

Why is everyone so stiff in these PR blog posts no one likes? What if you share the people behind it? We were trying to zig where they zagged.

Fast forward six years, and "The Browser Company of New York"—we don't even share that name as much anymore because it feels like it's become the culture. I open my X feed and it's just these launch videos and these behind-the-scenes videos. There's nothing wrong with it, but it has now lost that authentic vulnerability.

Jackson: (56:20) I hear what you mean about authenticity. It's funny, I had basically the almost exact same conversation with Gabe Whaley, who runs MSCHF.

He's like, "The drops stopped serving us." Internet virality became something that was just about shots on goal. It was no longer—

Josh: (56:34) It's so funny you mentioned Gabe Whaley because he's actually responsible, indirectly, for us pulling back in some ways. After the deal closed, one of the things I tried to do was catch up with people that I really look up to.

Gabe, to me, is a generational talent. His office is around the corner in Brooklyn. We went on a walk, and I hadn't seen him in many years.

He was telling me that he's basically betting on the inverse of what he bet on before. His whole thing before with MSCHF was tapping into the culture in playful, subversive ways to game the algorithm.

His view was that the world's going the opposite way. It's all going to be about private, closed, exclusive, scarce, human-to-human experiences—often in person.

I walked home from that and there was something about it that was just—okay. We came in a world where there was no transparency with regards to tech companies. People didn't trust tech leaders.

You never saw the people that made the product. This thing we did at The Browser Company—if Nate made a product, you heard about Nate. If Kristina made a product, you saw Kristina. Why should you see me?

Jackson: (57:42) Yeah.

Josh: (57:43) I just think in the same way that Gabe and MSCHF are evolving to the culture around them, so much of what we did before was maybe kind of cool in 2020 or 2021. But if everyone's just doing it as part of the playbook and they’ve got the contractors—what are we doing here? What's the world today as we see it?

Don't get me wrong, we're taking some big swings on storytelling. We actually just did our first national ad campaign: posters in a bunch of major US cities. We commissioned this painter to just paint a bunch of art and we put it up in these cities and said, "We hope you have a great day. Dia."

We can get to the commerciality and why strategically we did that, but you're going to see us experimenting with new mediums and new ways of telling our story that are not going to be direct-to-camera. But I had to talk to you; we've gone back, so it's an honor.

Jackson: (58:37) Yeah, maybe let's talk a little bit about the commerciality. I want to hear about how the last year has been at Atlassian. In some sense there is a great irony—you've talked plenty about how you were so obsessed with not building, even where we started the conversation, the enterprise thing. If I were more commercial, maybe we would have built the enterprise thing sooner. In the end, you work at... well, when you think about SaaS...

Josh: (58:59) Atlassian sort of invented bottoms-up SaaS.

Jackson: (59:01) They know SaaS. You've certainly ended up where maybe other people were telling you you should have been many years ago. How has that been?

Josh: (59:10) It's been fantastic. But one revealing story there, speaking of "assume you don't know," is when we were early in Arc. It wasn't even called Arc yet; I think it was still called Browser by The Browser Company.

There was this guy—and I knew everyone then. I had recorded selfie videos for most people that had onboarded to Arc, just welcoming them. That was the level of new people using our products. Then we opened up a little bit more where I couldn't record the videos and get to know people, but just barely.

There was this guy that kept reporting all these bugs, and his volume of bug reporting was just so off the charts. I was like, "I’ve gotta send this guy a thank you." I copy and paste the email and paste it in, and it's: CEO of Atlassian. I'm like, "What's an Atlassian?"

Then I Google Atlassian and I'm like, "Oh my God, he's like the richest guy in Australia." Mike Cannon-Brookes is the most down-to-earth, wonderful, genuine person. He'd been using Arc, maybe before it was Arc, and not only using Arc but giving us feedback.

Actually, during SVB—speaking of commerciality—I had 100% of our cash in SVB. Overnight, I thought all of our cash was gone. Mike texted me, because he now had my number from all his bug reports, saying, "Hey, do an all-hands. Tell your team their payroll is good for the next two months. I'm going to wire you the money you need from my personal account to The Browser Company."

He said, "I know what it's like to run a company." He wasn't an investor or anything—I think maybe Atlassian had been a minimal strategic investor, but my relationship was with Mike.

The point of all of that was I trusted him. He loved our product, was there since the beginning, helped us build it, and bailed out our team. He cared. And he kept telling me over the years, "What are you doing, man? You're monetizing this thing all wrong. You're going to realize you're an enterprise company."

I respect people that call their shots and stick to it. I found it really annoying at the time—I don't know if the Atlassian people will love that—but it would border on pissing me off. But I respected it many years later because I realized he was right.

He was right about the ideal way to monetize a browser, given everything we cared about in those founding values and hypotheses. Especially because at the time, what we were doing with Dia was saying, "Hey, we think the main interface to AI on your computer, at least for work or school, is going to look like a browser."

How do we make AI cross the chasm? How do we package it up? In retrospect, it's obvious now, but this was when OpenAI was focusing on consumer and everyone was saying it's the next Google. We said, "Hey, this looks like productivity software."

We were playing around with LLMs too, and it seems like they’re really, really good at things related to our job. In addition to just realizing that getting people to pay for it like a professional tool was the business model most aligned with our values, it was also most aligned with the counter-positioning against Google Chrome and every other browser.

What we had done and risked everything with Dia was to say, "Hey, let's go be that iMac or that Macintosh for AI. Let's try. We're probably going to fail; let's go." If everyone's using it for productivity use cases, you're not going to build an ad business off of that. Actually, an ad business will have weird incentives if you do that.

Jackson: (1:02:30) Yes.

Josh: (1:02:31) And so, what do you use your—

Jackson: (1:02:33) —desktop computer for in 2026?

Josh: (1:02:34) Yeah, exactly. One of the things that people are surprised that I admit is when I'm not at work, I hate touching my laptop. If I have a weekend, I'm not touching that thing. It is so strongly a work device.

Jackson: (1:02:48) And it may be in a way that is distinctly more so even than two or three years ago.

Josh: (1:02:52) Yeah, it is. I mean, my favorite test is the Williamsburg coffee shop test. You go into a coffee shop and it's a computer lab. It's legitimately, during the day, a computer lab. I walk around—hopefully not too creepily—and just kind of glance from afar at what they have open.

Everyone has a browser open. And actually, they're using it like an operating system. The thesis was correct: they’re doing a video call, but a video call is a tab in Chrome. So, yeah, it's full circle in a way.

1:03:23 The Internet Computer in the AI Era: Codex, Cloud Browsers, and Why Everything Is Converging

Jackson: (1:03:24) So I think this opens into a question that is worth... I mean, the theme throughout all this has been Internet computer, bet on the cloud. It's not about the hardware, it's not about local. In at least the last couple of years, maybe since Claude Code for a certain kind of person, it's been a total return to everything on device.

And we'll see. Like, maybe Mark Zuckerberg's giving everybody a Muse cloud computer. But it's not quite so obvious as it was maybe two or three years ago that this long-tail bet was right, that the browser could be the home for the computer. Do you have the same conviction you always had?

Josh: (1:04:02) I do. I have it more strongly. I think I read what's happening a little bit different than you. And this is what I love about the bets we made.

We bet on Dia and bet on this new product because we felt three things are going to be really important on desktop with AI. One, we thought your tools — the apps that you use, the ones you have up in the Williamsburg coffee shop — weren't going anywhere anytime soon.

I don't know what's going to happen in 10 years. That's the whole theory of the company. But our belief was that you weren't going to stop using Google Meet or Figma or any of these things. Even if you used AI for some part of it, you're going to pop it into a Notion doc or a Confluence doc.

The second thing we believed was that AI — the most valuable thing as the models commoditized and they all got really smart and cheap — was going to be the kind of personal context or the personal information and your work information that you brought to the model.

You put in the input tokens to make the output tokens better because everyone's going to have the same Nobel Prize winner in their pocket. And so the browser had, in addition to all of your tools — because that's where all your apps are now because of the move to the cloud — it didn't know just that there was a Notion doc or a Confluence doc. It knew that you've opened it seven times in the past 48 hours and you've actually written a bunch in it too.

Then the third thing was that our belief was that we really wanted to try to build this Macintosh. We wanted to be the Macintosh moment or the Macintosh product for the kind of command line interface that we saw. ChatGPT and Claude — we have not done it yet, but that's how we think about it.

What really can cross the chasm? How do you... what's the Susan Kare icon desktop metaphor for this technology? And that without that it'd be hard to cross the chasm.

Okay, so let's replay the last year. You can look at Atlassian's stock price. SaaS is dead. SaaS is dead. It's all going away. I actually — we checked the numbers — we're making all the same things we used to.

And you go talk to your friends from college. I don't think they're going to say their stack of software tools at their job or school is any different than it was a year ago other than the addition of AI. So I don't think these tools are going anywhere, at least not right now.

Second, we've all learned — and I can't even say we appreciate it so strongly — that truly personally understanding your context is the differentiator between these tools being kind of good and extremely good. And we have learned that actually the browser has a lot of that information.

And I would argue that for how powerful Codex and Claude Code and these tools are for people that use them, it's really lopsided right now about who uses these and who don't.

Jackson: (1:06:25) It's niche.

Josh: (1:06:26) It's niche because it's nerdy. I sent my wife a podcast the other day because I wanted to talk about what's going on with RL scale, and I wanted her to know. She just couldn't stop cracking up in the car where the guy kept saying "agent forking."

She's just like, "This is insane. This is insane that you all got to get out of your hole. Agents fork and they share context windows? What are you talking about?" And so I would argue that the average person using AI or ChatGPT is still using it like a Google search replacement.

And so certainly now alongside of it, if you do look at the sort of insiders — let's say the software engineers that are Codex and Claude Code pilled — I actually disagree. I think it's more compute kind of cloud-server based than ever.

Look at Anthropic's revenue. Look at OpenAI's revenue. Why? Because they can charge inference for a frontier model sitting on a bunch of GPUs somewhere that is definitely not local.

Jackson: (1:07:15) The model is cloud-hosted, but the environment the person is running — maybe to put what I'm trying to say a slightly different way — at least the harnesses today are running on your computer. So maybe that's just... and I would argue that it's actually this trifecta of obviously a really good model, all of the context stuff (which I think undoubtedly a browser is an amazing place to access that), and then part three is the harness.

Josh: (1:07:38) Yeah, maybe here's a way I like to think about it. A year ago, OpenAI was ChatGPT, a web app in a tab. Vast majority of their usage for ChatGPT was a web app and tab or mobile.

A year later, what are they talking about? Everything. ChatGPT now is Codex. Codex is a Mac application just like a browser. Why? Because that gives you certain permissions.

And Codex — what's the big center? Why do people keep asking me about Codex? Because they've realized, "Oh wow, you need a browser in order to be valuable for these things."

So what did they do? They took their browser and they stuck it in Codex on a local Mac app and are making this super app, and then they're making these cloud computers so you can access that super app built on top of a browser anywhere on any device. They're just making the same Internet computer thing.

And the difference between OpenAI and us — in addition to insane scale and a bunch of other things — is their belief is that we're going to jump to chat is going to be the center of everything.

Jackson: (1:08:40) Yep, yep.

Josh: (1:08:41) And so their view is that the browser is going to be really important. So important that we have a browser right here and we have a browser in the cloud. Browsers are going to underpin when you render an artifact or you make something—they just launched the slide deck thing—it's going to live on a server somewhere and we're going to render it in the browser.

They are so browser pilled. They're just sticking it under this chat interface and saying chat's going to be the interface. And we're the opposite. We're going: did we learn nothing from every other era of computing?

Whether it's the click wheel or multi-touch on the iPhone, the big default iPhone apps, Macintosh and the desktop metaphor, our view is people are visual creatures. They want big buttons, they want a little bit of color, they want to move things around, they want it to be messy.

When you do work, it's amazing to me. There's something about coding that is really attuned to a linear conversation: "Hey, go fix this bug. Now go do this thing."

Jackson: Okay.

Josh: (1:09:29) Have you ever—when you write, when you prepare for this, I would love to see a video of what that looks like. You're probably bouncing over here, there are notes here, you're looking at this, you're reading. It’s certainly not linear.

Our view is that where we're going to end up is we're going to have a graphical user interface that lets you be messy with buttons and widgets and apps, whatever that means in the AI world. And underneath is going to be a browser.

Again, on the surface, if you listen to our Internet computer video, we said it's not going to look like the browsers that we had. We had a blank screen when we talked about the Internet computer because we said we don't know what it's going to look like. So I actually think everything is converging.

Jackson: (1:10:07) Yeah.

Josh: (1:10:07) And everyone is going to realize that—and has started to realize that—the browser is central. The key disagreements are sort of artistic and cultural about what sort of environment do people want to work in?

If you assume that you're going to have these LLM models and an agent harness and a web browser and SaaS apps...

Jackson: (1:10:28) Yeah, yeah.

Josh: (1:10:29) One is, by the way, that's very scary for us. Just to be really clear. I'm not saying "we've made it." We all of a sudden are competing with all the greatest companies in the world.

So there’s a question of: was the thesis correct? Do you have any reason to believe the thesis? And will we win?

But I actually think there's more consensus than ever that whether it's a local browser or cloud browser, people are going to need browsers. And in fact, what people now believe is it's got to blur into one thing.

That's the big thing in the moment right now. If I have a cloud agent doing something for me or my local agent, I shouldn't need to think about that. What is that describing but the original vision of: it doesn't matter if you're on your phone or your computer.

Jackson: (1:11:06) That's the most annoying part of it right now, by the way.

Josh: (1:11:08) Yeah, that was the whole theory—that if you open a Windows device at work and a Mac device at home and an iPhone in between, you should have your computer span. That was the whole idea of an Internet computer.

Jackson: (1:11:18) Whereas there's two ChatGPTs for me actually, which is the mobile one and then the Codex-like thing that has all my files.

Josh: (1:11:23) Because that wasn't what they believed 12 months ago.

Jackson: (1:11:25) Yes.

Josh: (1:11:26) So actually, the other market participants suggest to me that the thesis, fortunately and unfortunately, was more correct earlier than we expected. We have to run.

Jackson: (1:11:37) Put it another way, it's like—at least from my vantage point—the radical transformation the last few months has been the quality of computer use. I guess what you're saying is that computer use—it's computer use right now locally, but it will be computer use in the browser in the duration of time.

Josh: (1:11:56) Yeah, I think so. And again, when I mean browser, I mean there is a, let's say, Chromium rendering engine technically under the hood that can render web applications like Google Meet and Jira and the tools you use at work.

Jackson: (1:12:08) And run computer use over the top of that, or effectively browser use over the top.

Josh: (1:12:11) Right. So when we say the browser is going to be the future, or the Internet computer is going to be built on top of a browser, we don't mean a horizontal tab strip.

We just mean that if all of our applications are stored on computer servers—which that trend is accelerating—and all of our files are stored on computer servers, and all the frontier AI models are stored on computer servers, then actually, go back all the way to starting the company.

The acceleration toward your computer living off there somewhere, and it's streamed effectively to this little screen—and shouldn't you have the same thing streamed to all the screens?—is just turning out to be more true.

And so, yeah, that's part of the reason we picked Atlassian. We said, "Oh, we gotta go."

And we said, "Atlassian—this guy gets who we are. He's as ambitious as we are and he sees what we see." Let's go run as hard as possible and as inspired as possible. How do we redo what a browser is? Which brings us back to Arc versus Dia.

If that's what we've always been trying to do and we see it coming true after years and years of working on it, we go: we're competitors. Let's go win.

And we just had to have a pragmatic view that we're not going to win as some dinky little startup called The Browser Company making YouTube videos. We need some firepower, we need some distribution, we need a built-in business model and data and security agreements with enterprise companies. Let's go actually try to win this thing.

1:13:30 How Dia Wins

Jackson: (1:13:31) How do you win?

Josh: (1:13:33) As you can probably expect, my real answer is I don't know. We are a company that assumes we don't know. There are three things that I've observed in the past year that make me more bullish that we're going to have a seat at the table and a real shot.

The first is: we still believe things that no one else believes. If one of my Codex-pilled friends was sitting right here, he'd go, "Yeah, sure, maybe it's technically a Chromium rendering engine, but you're not going to see it, and you're going to chat with it."

As much as we think the time is now for this opportunity and everyone has come to the party, we look around and go, "Oh wait, we still believe very, very different things than everyone else." Which means if we're wrong, we're boned.

But if we're right, we have such a different perspective that we're going to have a differentiated product for some period of time, even if everyone ends up copying.

The second thing we've observed is it's really market-expanding. I was an angel investor in Cursor, and one of the things that was fascinating is as Claude Code has this meteoric rise — numbers that are mind-boggling in terms of the speed of their ascent — all of a sudden Codex comes around and they do the same thing.

You go, "Oh no, what about Cursor?" Cursor was like this too. I think a thing people are sleeping on is our computers. What a computer is — which everyone has — is going to completely change.

We're of the belief that there are basically a few categories of software that will eat most of the TAM, most of the opportunity. If you can be player number 3, 4, or 5 in the primary way that people on their laptops get work done market across — that's effectively an operating system that apps and AI sit on top of — we think that's an extremely lucrative opportunity.

It's the same reason that Cursor, Codex, and Claude Code all coexist as legendary, fast-growing companies.

The final thing is why we did this deal in the first place, which is anyone that knows enterprise — which again, I'm new to — you start learning it's not just product that wins. It's trust, it's relationships, it's what they call "enablement" of employees.

Here's a company, Atlassian, that's been around for decades, that was in many ways one of the original SaaS and enterprise companies that both had great products and were embedded in companies.

Our belief is that if we can figure out that product and we can make something people love, it's not going to be what everyone else has built yet. It's going to be one of the greatest TAM opportunities in our industry in decades.

We have a company that knows how to sell it and knows how to distribute it and has the embedded relationships with Mercedes-Benz or Nvidia or — insert your Atlassian customer — that hopefully Dia is the perfect tool for their employees.

Especially for employees that are not software engineers, that are not Codex-pilled, that still find the scene confusing and, quite frankly, don't like it. Have you noticed right now people don't feel good about AI?

I think that's the thing, too: Who do you trust? Whose future world do you want to live in? I'm not saying mine, but I just don't think —

Jackson: (1:16:35) But am I going to be the portal through which you access all this magic? Can I actually deliver it to you in a way that —

Josh: (1:16:40) Yeah, and I think focus, too. We just want to make your day at work better.

Jackson: (1:16:43) Yeah.

Josh: (1:16:44) I still get pushes from ChatGPT to check my credit score. I got that yesterday — it took over the screen. "Check your credit score." Good for them; it's not what we're doing.

As long as you're holding on to all these different goals... Obviously, they're an extraordinary company that's been extraordinarily successful, but I think we look around and go, "We still have something to say that is different than what other people are saying right now."

1:17:07 Agents, Muse, and the Red Giraffe

Jackson: (1:17:06) You had — I don't remember exactly when it was, may have been the beginning of August — observed that the agent thing had not really hit the mainstream in any meaningful way.

The difference between now and then, obviously Instinct was a kind of first, more Silicon Valley-related moment, and then Muse has obviously grown like crazy.

What about the agent side of things? Obviously this ties a little bit into your kind of: Are you going to interact with everything through a text box? I'm interested in this as a sort of gateway into where we are going with what computers are.

Josh: (1:17:40) Once again, I don't know. But since that'll be a boring answer for everyone listening: the tweet in question. I called out that for all the talk of agents, there hadn't been an agent that had broken out in popularity and that was the next big opportunity. Who was going to do it?

Weeks later, maybe someone did. It was not a question of "Would it come?" It was saying that is the big opportunity.

Jackson: (1:18:03) Do you have any observation on what caused the success that Muse has had?

Josh: (1:18:09) Oh, this is tricky. So we're recording this on October 1st.

Jackson: (1:18:12) Yeah.

Josh: (1:18:13) I am skeptical how durable this excitement is and where it is coming from.

What I observe is what I observed with OpenClaw. When I tried OpenClaw and I saw people getting excited, I was skeptical about who was getting excited and why.

I think what it was masking was true architectural and technical innovation. Someone had figured something out about how to take these large language models and hold them or extend them in a way that made something possible that wasn't possible before for people using it every day.

People got really excited about that. But the form factor didn't really change in many ways — it was still chatting. And I think Instinct and Muse are really —

Jackson: (1:18:53) In my opinion, Muse is basically OpenClaw, or not exactly.

Josh: (1:18:56) I think one of the things that really has made these products click in is that now the computer use models are accurate and fast and cheap enough that you can actually really use computer use.

Jackson: Yeah.

Josh: (1:19:05) So we’re putting these labels on Instinct and Muse. It looks pretty like chat, everything I’ve been doing in the industry for a long time, which is chat—just empowered by these really clever and really miraculous new technical advancements.

I look at that and go, I’m not sure that that means those products or that form factor is going to be successful. I’m much more bullish about it on mobile. I think if I’m a person that wants a restaurant reservation on my iPhone, even as a personal consumer, chat’s amazing for that.

But I’m not so convinced, especially on desktop, especially at work, which is messy and requires lots of tools and context switching, that that form factor is going to be the final one.

I have been blown away by Muse in particular. It’s been the best fully formed AI product I’ve used in a long time. But I’d say at least our theories about the interfaces people want—the design metaphors and the buttons that they tap and the spaces they’re in that are on top of these harnesses and models, including computer use—I’m not so sure that we’ve learned anything in the past couple weeks that suggests that might change.

Jackson: (1:20:08) You referenced it early in the conversation in that early video. Your son was one or two, and you talk about this computer that he might use when he was 10.

Josh: (1:20:18) I think he’s five now, turning six in January. Yeah.

Jackson: (1:20:21) What will a computer be to him in the abstract? We could talk about it when he’s 10, or just what he conceives of a computer as when he’s older. Do you have any—and this is obvious, I’m asking less about a strategic to the business—just what do you think to someone of his generation, what is a computer?

Josh: (1:20:44) It's funny bringing this up. I had two interactions with him in the past week that I honestly have not had time to reflect on, so I’m just going to leave them here.

One is, I was demoing something we had prototyped related to some of these themes. I said, "Hey, go tell your phone what you want." It was on my phone. He said, "Picture of a red giraffe." And a red giraffe popped up.

I was like, "Oh my goodness." And he was amused, but for him, it was like, "That's just what a phone does." There was this delta between my own—I’ll never forget this moment—computers are changing, the world’s going to change, oh my goodness! And my son just had this red giraffe expectation. He’s just like, "Yeah, I guess that’s what a phone does."

That led to another thought, which is my son—because probably his memory will start getting really strong in a year or two—will never know a world where humans were smarter than computers.

Jackson: (1:21:40) Yeah, Sam Altman said this too.

Josh: (1:21:42) Maybe. Okay, then I must have heard him say it on a podcast. But it just actually probably connected there. I just can’t stop thinking about that.

My wife’s grandmother passed away a couple years ago in her 90s, and she grew up in—she has memories of the era of World War II. She would tell the stories of living in Paris for a job at one of the first children’s libraries in the world and writing letters home to her family.

Then she remembered when the ATM was invented. She brought this up because at the time before she passed away, ChatGPT was everything in the news, even her news in Switzerland watching on TV.

I just can’t help but think I’m 36. What when I’m 90? So it’s just unfathomable to think about, and exciting and scary. But I think the red giraffe can just appear on the screen that was in Papa’s pocket just by uttering those words out loud—I think that is suggestive directionally of where we may be going.

1:22:43 Software Is Becoming Art

Jackson: (1:22:45) I also think it’s worth calling out that that is something different than computers being smarter than us. Yes, of course computers can be smarter than us, but also, that is a different kind of metaphysical magic.

Josh: (1:22:55) Well, and one of the things that I’ve been thinking about related to that is just this idea that software is going to become art. Software is art now.

And I mean the word art in terms of almost a commercial standpoint. I’m not talking about feelings here. I’m just, very literally, supply and demand. Software for the first... I’m a sociology major. I met Hursh because I needed someone to help make my ideas.

My whole career has been like, "Let’s get money from venture capitalists to go hire software people to help make."

Jackson: (1:23:23) Yeah, you're a middleman now.

Josh: (1:23:25) Don’t get me wrong, the people are the best part of this. I have no interest in starting a one-person billion-dollar company. I think it’s very dystopian.

But if software is now approaching free, that sounds a lot like paint and canvases are approaching free, and a MIDI keyboard and pick your music software and your computer is approaching free.

I actually think the profound shift I don’t hear people talking enough about, that is really exciting to think about, is not, "Oh, if software is free, I’m going to make a better productivity tool just for me." People don’t—as my therapist likes to say—people weren't put on this earth to be productive.

I get really excited that if software is as free as it is to make music and make movies and make visual art, then what are we going to make, and who’s going to make it, and why?

Yes, some of those will be productive tools that make us better at work as utilities that solve this problem. But I think we’re going to see an explosion of software that is everything except something that makes us more productive and useful.

As someone who grew up loving languages, singing, and indie films with my wife or family, that is very exciting. So that’s the other side of this. There’s, "What is a computer going to look like? How are you going to interact with it? How are you going to get your job done?"

And then there’s, "Who’s the Paul Thomas Anderson of software?" What are they going to be able to do, and what are they going to make, and where are they going to come from? And optimizing for feelings—they're going to be like, "Oh, that was really silly." I think that’s going to be a pretty profound culture.

What even creative mediums do we enjoy? I’ve always been impressed and inspired by the way that you’ve kind of danced to your own rhythm. When you joined 100 Thieves, you sort of saw something about video games culturally that most people didn’t see, in the same way that I think YouTube culture...

But you saw this before it was obvious to other people. I think something similar is probably going to happen with software, where everyone’s still thinking about productivity tools, but it’s going to be so much more than that.

Jackson: (1:25:27) The thing that really resonates is not only who's the Paul Thomas Anderson of software — which is to say there's something that comes to mind there which is just, we're in the 20s of film or whatever, we have no idea what creativity is possible.

But specifically, who's the Romero of software and the Christopher Nolan of software? It’s the breadth.

Josh: (1:25:50) That's what I mean.

Jackson: (1:25:51) Yes.

Josh: (1:25:51) And that's why I think software is going to become — "art" is a little too medium-post-headliney. By that word, art is doing a lot of work there.

Jackson: (1:25:58) Right. Yeah.

Josh: (1:25:59) Like, what I mean by that is literally, what do you— red giraffe on a screen? Like, what is—

Jackson: (1:26:05) Software is amazing, but we have a pretty concrete sense of what good software and good software design is. Which, by the way, is why when you see the stuff someone like Nate makes, you're like, "Whoa, I didn't know this was allowed." But that's like the trickle of that.

Josh: (1:26:18) Yeah. And that's what personal computing opened up. People always talk about the 1984 ad and they always talk about the Macintosh and the "hello" world and the icons — and I do a lot — but people forget to contextualize what computing was.

It looked a lot like right now. There were a couple big companies, largely International Business Machines, that made large, extremely expensive computers for corporations to count things and do very, very productivity-centric tasks.

And they try to dress it up and they hire Paul Rand and the Eameses and they try to make it friendly. At the World's Fair in 1964, it was fundamentally these kind of big corporations making expensive stuff for businesses. That's what a computer was.

That's why they did the World's Fair where they hired Charles and Ray Eames, because they're like, "We gotta make this friendly for people, because people are not liking this computer thing." It's almost identical to what's going on with AI right now.

And so to me, the answer is not, "Okay, we're going to make a business machine." It's actually sort of something you can't imagine — I can't imagine — which is: what if you go really wide that way?

Jackson: (1:27:25) About what it will need to come from somewhere else.

Josh: (1:27:28) Yeah. And it's just, I think it's going to be orthogonal to what we think it is and obvious in retrospect. And so to me, that's the art part of it.

There's some little Paul Thomas Anderson or Zadie Smith or insert your favorite kind of creative person somewhere that is going to have Codex or Claude Code and Dia, or whatever it ends up being.

Jackson: (1:27:44) Didn't have this medium before and now has it.

Josh: (1:27:46) And so I have no idea. But I'm— that's the optimistic view.

Jackson: (1:27:50) We have some hurdles from a culture standpoint, I think. Unfortunately, a certain kind of creative person feels pretty disenfranchised from these things right now. So we have some things to...

Josh: (1:28:02) I think it's understating it. There's an artist named David Salle who I am obsessed with. He's a really famous painter in the art world. If you're an art student, you know his work; if not, you may not. I didn't know him before.

Despite being really famous for bringing painting back and painting technique back — he's famous for his ability to wield a brush and sort of moving away from the conceptual art, like Turrell, that had gained popularity a few years ago — he took Stable Diffusion and made his own model for his own work.

What he did is he took the kind of series that he is most known for — sort of like Turrell’s Sky Spaces or something — and he fed all of the artwork he's ever done in that series into a Stable Diffusion forked model. And he made this model that suggests what he should paint next.

Hearing him talk about his relationship with the AI model is fascinating because he says it's like found objects, or it's like a distance. So he doesn't use AI to paint, but it's sort of this new creative tool, a professional tool in his toolkit.

And, you know, he was just struck by how the art world just really is very allergic to AI right now. So that is the thing that's new. I don't know enough about what the cultural climate was around computing in the 60s, but I’ve got to assume it's not as negative as it is right now with AI.

And so I do worry a little bit about that. We're so talking about bioterrorism and all these things which are very scary and deserve the airtime, but unfortunately, I think it's pushing down the creative possibilities that are not being explored right now. Because we're in the International Business Machines phase. Well, not—

Jackson: (1:29:48) I would argue it's both that and all the risk stuff and the fact that, for lack of a better word, AI one-shots the output. Yeah, we're probably going to need a paradigm update on what artistic output means.

I was just talking with Kareem from Clay about Sol LeWitt. I didn't know actually anything about Sol LeWitt and I didn't know it was basically the prompts; it's not the artwork.

One of my favorite, really amazing things that came out of crypto was this thing called Art Blocks, which is very similar. It's not necessarily any individual piece; it's the algorithm that creates this body of work.

And my suspicion is the really interesting AI art will probably be more in that flavor, which is: what can you uniquely do with this thing that isn't like, "Wow, you hand-drew this picture and AI one-shot it"? That's what currently people hate. So it's like we're missing that update.

Josh: (1:30:41) And I think in terms of what software becomes or what do we do with our computers, I think it's going to be the era of what at The Browser Company we called "mutts."

When we were early at The Browser Company, we were trying to put language on why did we hire Nate? Why did we hire Adena? It was because these were people that, like Turrell in very different ways, sort of defied labels in a way.

Nate was an engineer and he was a designer, but he's also clearly a product person. He clearly has this kind of comical, artsy side. He just made things.

I was just struck by Turrell. He'd walk in a room and people would introduce his work. It's like: is he an artist? Is he an architect? Is he an interior designer? Is he a perceptual psychologist? Is he an astrophysicist?

He just says, "Wait, hold on. If we cut a hole in the ceiling, and then you’re going to teach your brain, your eyes are going to get it, and then it’s going to be beautiful at sunset."

Jackson: (1:31:34) What could we do?

Josh: (1:31:35) What could we do? I think we just need new ideas. We're just in a moment in the tech industry and in our political climate where we're so in need of new ideas.

I think we will see most of the most compelling emerge not from these traditional verticals, even, like software and computers. But if the components of a computer are cheap and accessible, and intelligence is free and accessible, and all of it's malleable, and you have these tools—how does that stuff get combined in interesting ways?

It sort of blurs the categories and labels we had put on it before. So, yeah, I think it's going to be the era of the "mutts" in some ways. I hope, and I think it'll be really cool to see what we come up with.

1:32:13 Believing in People, Leaving Facebook for the White House, and What’s Next

Jackson: (1:32:16) We talked a bit earlier. One thing that I think almost anyone who knows you would say is that, probably largely because you're so attuned to it, you are really amazing at attracting great talent—or I should say, spotting it and ultimately attracting it.

Aside from being able to find it, what do you think attracts these people to want to come work under the Josh Miller umbrella?

Josh: (1:32:44) Oh, that is such a hard question to answer about yourself.

Jackson: (1:32:47) Yes, but I'm going to make you try. Or maybe an easier way to answer would be: is there anything you do? Are there any ways that you've obviously gotten better at it?

Josh: (1:33:01) There’s something that someone said to me a couple weeks ago, actually, that really hit home. I should tell them that this meant so much to me. They said, "I think you believe in me more than I do."

Jackson: (1:33:14) Yeah.

Josh: (1:33:14) Because I do believe in them. I hope they don't mind me saying this, but they need to realize how gifted they are. I feel like I got this gift from my mom and dad to a fault; they sort of said, "You can do whatever you want. We believe in you."

So I think probably intrinsic is, yes, spotting talent and aligning it with some strategic priority and articulating our ambitious vision. All that good stuff. But I think I am at my best when I am really speaking from the heart and authentic, and I'm at my worst when I lose touch with that—and I do all the time.

There are a lot of podcasts I've been on where I'm so embarrassed that I showed up with that energy that day. But when it's really coming from the heart, I do believe in certain creative people and their gifts. I see them sometimes earlier or more extensively than they do, and I sort of place them in places at the company that lets that out.

Actually, one of these prototypes I was talking about is from someone that I plucked off of one of our main teams. I said, "Hey, I want to go in a corner with them and just sort of mess around with some prototypes related to graphical user interfaces and duplex. Let's just go mess around."

They obviously didn't arrive because they felt that, but I hope creative people who have worked at The Browser Company, or do work at The Browser Company, say things like, "Yeah, it’s like this campaign we just did."

We were just like, "Let's go put up some art in some cities and make the message be: we hope you have a good day at work. Hey, you want to go paint something? Go paint. You go paint the murals."

I think it's that "you go paint the murals and we'll put them around the city" energy. Those stories getting out probably lead to people believing it's a place they can express their creativity. But again, that's hard to answer. I would be curious what Nash and Josh Lee and all these people would say.

Jackson: (1:35:01) What is so special about Hursh?

Josh: (1:35:08) Very similar to the last question. Hursh is a brilliant software engineer, etcetera, etcetera. But I think he's one of the people that is the most spiritual and in touch with who he is. When I met him, he actually let me come live in his apartment as we were starting the company.

Hursh has meditated every day for 20 minutes—real intense meditation—since I met him. There's just a soulfulness to Hursh and a depth to his breath. Actually, everything I'm saying right now, I realize that came from Hursh.

I remember when we were talking about starting a company or getting the gang back together, it was all about this idea he had about abundance versus scarcity. He felt so many of the places he'd worked—and we had worked—were driven by this idea of scarcity and a scarcity mindset.

He said, "We just need to give. If we can create a working environment where people are their most abundant selves..." And I was like, "Man, what are you smoking these days? This is just so hippie-dippy. What is scarcity versus abundance?" Truly, at the time I was skeptical.

Jackson: (1:36:11) Few more thousand hours of meditation.

Josh: (1:36:13) Yeah. Like, "Man, we gotta get you out of the silent meditation trips and get back in the arena. This is a little too much for me, but I love you."

He was so right in retrospect. I think what is so special about Hursh is, despite being a technical genius that people associate with his engineering skills, he's almost the most "people" of people I know. That has informed so much about the company that we've built.

Jackson: (1:36:35) You guys sold Branch to Facebook. You were at Facebook in a pretty significant time, leaving to go work on politics. Maybe you had the civic background, but why did you leave?

Josh: (1:36:46) Yeah, it was the dumbest financial decision I've ever made in my life because we cared about ethics there. I had to sell all my shares and forfeit all of my Facebook stock from the deal.

Jackson: (1:36:57) Wow.

Josh: (1:36:57) It was a very dumb—the stock price was—anyways, it was a poor financial decision in retrospect, and actually a poor professional decision. It didn't make a lot of sense other than I loved the man. He inspired me.

As someone that grew up around public servants and people that work for the government, I personally view the government as us—we are the government. This is our government, our society.

Jackson: (1:37:24) And that's an important idea.

Josh: (1:37:26) I've always grown up very interested in policy and politics and civic engagement. Obama was the first politician—I grew up too young for Clinton. Most of my adolescence was Bush and honestly hating on Bush and being sad about our president. Then Obama comes around, and he's the first president I ever voted for.

So, my first mentor in this industry got hired by President Obama to report to him and create this new office in the White House that was going to bring technology to the government. He asked if I would come be the lead product underneath him. It was just such an obvious, intuitive decision to say, "Of course." You don't think about that. You just do it. You just jump.

But I remember Zuckerberg, who was so gracious and so generous to me while I was there. I had a call with him about the move, leaving, and he was very sweet about asking if I would stay. I remember him making this comment: "Why would you go build software for one head of state when you could stay here and make software for all the heads of state forever?"

He meant it so sincerely as a "I'm giving you a cooler job." In many ways, for a lot of people, he would. But he's such a competitor.

Jackson: (1:38:40) Yeah.

Josh: (1:38:41) The thing I learned from Zuck—I actually learned from Zuck that I would not be a good public company CEO. There was just something about the way he thought strategically—3D chess scale, what's impactful or not—that I was like, "I just kind of like President Obama."

It seems like a fun experience. I watched The West Wing when I was a kid. I kind of want to know what the Situation Room looks like. Maybe it's not rational, it's not game-theory optimal.

Jackson: (1:39:05) Yeah.

Josh: (1:39:07) So it was definitely a hard decision. I'm so glad I did it. It definitely changed the course of my career in so many ways in retrospect. But at the time, definitely people were like, "You just got here. Why'd you work so hard to get here?"

Jackson: (1:39:17) What are you doing? I mean, I guess Mark was right in the end about the scale.

Josh: Mark was, yeah.

I remember around that time—actually, before I started working closer with them towards the end of my tenure—but at the beginning, The Social Network came out in 2010. That was my sophomore year in college. So I really wanted to meet Mark Zuckerberg.

It was late one night, still when they were in the old Hacker Square campus. Zuck was walking out, and a few of us were there late, and he came over and chatted with us. People were asking questions, and I forget what the question was, but basically Zuck's prediction for computing at the time was that one day, sooner than we think, computers would understand human personalities from an algorithmic perspective.

I actually just looked it up the other day. We were acqui-hired right around the time they started FAIR with Yann LeCun. So secretly in the background, I think where that was coming from—though it seemed random at the time—was he was deep in 2013 thinking, "One day we're going to have these AI, and it's going to be..."

But I thought what was so prescient about that is he focused on people's personalities. He didn't say computers would be smart—I'm sure he believed all those things—but the emphasis in his answer was human personalities. That's the thing I think people still underappreciate about large language models: our computers can now make us laugh.

They can now give us constructive feedback that really burns a little bit sometimes. They can hallucinate kind of like humans do and come up with wacky ideas. To me, that relates more to personality and sensibilities and preferences.

Once again, you can like him or dislike him, but Zuck, around the time he was like, "I think the more impactful thing is to make software for every head of state," by a lot of measures, he was correct. He was also correct in saying the really profound shift that would come in computing is they're going to start being algorithms that resemble people and can understand people in a way that we don't comprehend.

Jackson: (1:41:35) Sometimes it's good to do things because you want to and you feel the need in your gut.

Josh: (1:41:38) I have no regrets. I made the right decision for me.

Jackson: (1:41:41) How would you—if you're going to give a title to this chapter of your life, post-acquisition, what feels... are there any words or ideas that feel descriptive of this moment?

Josh: (1:41:51) Well, in this exact moment, which is different than if we'd done this podcast six months ago, I feel a profound sense of listening and curiosity. I can run my mouth, as you can probably tell on this podcast, and there are good and bad things about that authenticity.

After the deal closed, I felt this almost manic energy for all these creative things I wanted to do at The Browser Company now with Atlassian, and just sort of this explosion of things that were possible. I still feel that, but I just want to go back into that kind of learning, observation, and curiosity.

You asked how I developed my sensibilities—like going to a museum tonight with Valerie. Even at Atlassian and even post-deal, I was so inspired. I got to meet with this woman that runs the software for one of the largest grocery chains in Mexico that is a big user for Atlassian.

Just hearing about them—they're super early on AI and they're trying all Claude code. I just feel like I have so much to learn, and in some ways, my world was so small.

Jackson: (1:43:03) Sounds like things are going great at Atlassian. As I understand it, The Browser Company is very much still its own thing. I'm sure there's a lot of room to run.

If you were to try to imagine what the next era or phase is, whenever that phase is done... Is it another startup? Are you finally going to be the artist?

Josh: (1:43:26) Yeah, I feel like hopefully the input tokens in this conversation make the output tokens clear that I don't know, and I've been wrong when I've made plans.

But I'd say anyone that knows me really well is going to say, "You have to tell them." I do want to make a public park one day. I've always wanted to help build a public park. I don't have any ambition of how big or small it is. I have some ideas, obviously, but I would at some point love to play a small part. That's what people know me for.

Jackson: (1:43:54) You ever seen the Noguchi park designs?

Josh: (1:43:57) Yeah, I actually just stumbled across them the other day. Someone I really respect said that's a great example where it looks good on camera, but actually is very inhumane as a parent with kids.

Jackson: (1:44:07) Oh, interesting.

Josh: (1:44:08) I excitedly told her about it and she was like, "Come on, you're a parent, you know this sucks." But that's the sort of exploration.

Jackson: (1:44:15) Not a parent.

Josh: (1:44:16) Okay.

I have to say it, because anyone who knows me knows that's on my list. I do come back to my mom, who is a big force in my life. And my godfather, who's passed away — seeing them take their immense talents and EQ and IQ and put it into where we lived was influential.

It was almost like that same aspect of the web browser. It's a web browser, but it's really a rectangle that we spend eight hours a day in. And so shouldn't we make the rectangle feel nice and look good? Why would we not want to live in a good rectangle?

I think there's some part of me — there's this quote from an artist I love, and she says, "Art is what I give back to the world for all that it gave to me." And I definitely am sitting here going, "Wow, The Browser Company, if it had gone five degrees that way, would have been nothing." I wouldn't be on this podcast here.

I've gotten so much luck and good fortune to be here that I feel this indebtedness and sense of responsibility to somehow channel it in some way. Again, I have no idea what that means, and it may just be something local, but I feel the need to go local in some way. Very, very local.

So I think it might look more like what's around the corner than what's on the internet. But that might just be me exhaling from six years of getting punched in the face every day, as Hursh said, and saying, "Maybe we don't get punched in the face again."

Jackson: (1:45:44) Maybe there's something to what we were talking about with Gabe Whaley. Especially in tech, scale is virtuous, and there are things that don't get better with scale.

And things that actually get way better. I also can't help but think of, as you were talking — granted, it's more on the scale end of things — that old Steve Jobs thing. An old email to himself about how, "I didn't make the clothing... I didn't do any of this. And so I'm just trying to add something back."

Josh: (1:46:17) And I think the one thing that I'm now proud to say is that if you do look back at my career, there's not a lot of patterns between sociology and then here and then there. But one of them is I followed my people.

Yeah, I was at the Obama White House from Facebook. What you don't see is Jason Goldman took a bet on me, and in some ways us on him, when I was 20 and dropped out. And then I followed him to the White House.

Or you look at Hursh and The Browser Company. I met him and we followed each other. I loved working for Josh at Thrive. I hope — and this is a reminder to myself — to follow the people. Follow your people, because that is definitely what has brought me so much joy so far.

I hope that I've still got the crew, whatever that is, whoever they are, hopefully some of the same faces, whatever comes next. But I'll say I have a strong prediction: I will be at Atlassian longer than I ever thought I would be. I definitely am not a lifer at Atlassian — sorry, Mike, I love you — but I will do something else at some point.

1:47:17 Closing Questions

Jackson: (1:47:19) I like to ask people, when we talk about things we regret, what they're most glad that they did.

Josh: (1:47:25) I love that. I mean, it's got to be Valerie. I know that's so cheesy, but it's got to be marrying Valerie.

I think a thing that I've always bonded over with other founder CEOs, especially folks that have been successful — even Henri, who's on this podcast and I was with him when you asked me to join — we've spoken about how Valerie and Chloe were co-CEOs of these companies, and you may not have known their names.

So I mean, that's just at a professional level. I'll spare people all of the why I love my life partner so much and what she has done for me and us. But in many ways, Valerie and I are very "opposites attract."

Though it's obvious in hindsight, I think there were a lot of friends from college that were like, "Mostly, what is she doing with him?" There was a lot of, "Is that pair going to work?"

Valerie's been right there the whole time, out of frame. So I think the best thing I ever did was marry this Swiss woman that made me promise not to live in California ever again. It was the only kind of prenup thing we did. My heart is in California, so there were a number of things that were like, "Ooh, wow," but it worked out. She doesn't like sand and the beach. The beach is a red flag. Sand, beach — the beach is tough.

Jackson: (1:48:46) How do you think she's most influenced you, at least in the professional, creative part of your life?

Josh: (1:48:52) It's been the same answer since day one. It's why I was first attracted to her and what I still struggle with to this day: she keeps me honest.

Valerie's always known — I mean, there's a connection to Hursh, actually. They are so in touch with who they are, and they are so comfortable in their own skin, and they are so driven by what's internal to them.

I am someone that, at my worst, has been reactive and looking for approval and love and the things out there in the world, like, "I want to be successful."

I think Valerie's the person that always — I mean, put it like this. When I'm on the car ride home right after this, I'm going to think, "Would Valerie, if she listens to what I said, say that was the real me, or did I show up?"

Again, it comes from a good place. But did I, out of social anxiety or wanting to sound smart, put it on? It's always a bit of both. It's one of the things that's interesting about the behind-the-scenes era of The Browser Company: you're always performing a little bit, even if you try really hard not to, if there's a camera right in front of you.

Jackson: (1:49:52) yourself a little bit.

Josh: (1:49:52) But Valerie's the one that has always and will always ground me in, "Wait, who am I? What do I really want?" because she is so grounded that way. I'm very lucky.

Jackson: (1:50:01) It's great to have somebody who can check you on that. Yeah.

Josh: (1:50:04) Oh, and she checks.

Jackson: (1:50:04) Remind you who you are.

Josh: (1:50:05) Very Swiss. We're rule followers.

Jackson: (1:50:07) No California. One last thing — obviously not rehashing something we've talked about plenty. But I believe it's possible I made this up, but I really believe this is something you said to me several years ago that I wrote down: it all comes down to the crew.

Who would you be fine staring at a blank wall with once all the other stuff washes away? What do you hope — for so many of these people that we've talked about — what do you hope they say in the future, if I were to ask them about their time in the weeds with you? What do you hope they say about you?

Josh: (1:50:49) I hope that they, more than anything, remember my smile and energy on whatever it was we were working on together. I think I was just born this way. I'm an optimistic, excitable, genuinely "let's go do something fun" person.

I'm most proud of the ways that people have shared that I have, in moments, renewed their sense of possibilities, a sense of hope, and a sense of creativity.

Obviously there are many things I hope, but I really hope, especially for the sorts of people we've talked about today and the things that I've done, the people around those crew carry some of that with them.

I get it as much back from them. But yeah, I hope it's the smile and the sense of, "We can do new. Let's do something new. What could we do? Why don't we go? Let's go do it."

Jackson: (1:51:45) Thank you, Josh. That's all I got.

Josh: (1:51:46) Thank you.

← Back to all episodes