Charlie Deets - Designing the internet computer

00:00:00What's the special sauce that designersbring to the table when everybody canexecute?>> I can pull things back. I can make itsimpler. I can make it moreunderstandable for the user. And yes, Icare about making sure that button worksexactly right or whatever. And I couldhave went in in there and fixed it. Butif I keep staying like on this end ofthe equation where I'm thinking in thisrealm, I'm more novel over there than Iam in the execution realm cuz likeeveryone can kind of execute right now.How do you create these little momentsthat separate good products from greatones?>> Impulsiveness in decisions can beproblematic in the way that you can makebad decisions. But impulsiveness in yourdesign is actually something I think youshould facilitate all the time becausetrying out like one random thought popsinto your head. Go down that path. Tryit out. See what happens. If you're notplaying or you're not like messingaround, you won't come to those moments.Welcome to Dive Club. My name is Rid andthis is where designers never stoplearning. This week's episode is withCharlie Deets, who's a designer at thebrowser company where he's working onDIA. So, he's going to give us a littletour of some of the finer craft details.But we're also going to do a deep diveinto Charlie's

00:01:15personal designprinciples and all of the ways that theyshape his practice. But before we getinto all of that, this episode actuallystarts at Apple because before Charliewas designing Arc or DIA, he was one ofthe designers working [music] on Safari.And it's fascinating to hear him talkabout the differences between the twocompanies. I was part of the humaninterfaces design team at Apple. And itwas kind kind of in a slightlytransitional period where you know atleast for me I was like working on abunch of projects across various teamsand my manager kind of came to me oneday and was like it's probably time tofocus and like go deep on something kindof told me some of the options that wereout there and one of them was Safari andI've always been like in love with justlike pure utility driven apps and thingsthat aren't necessarily like flashy butwork really well. That's like I think mycomfort zone and you know Safaricouldn't be more like that in a lot ofways. It's like>> most people that use Safari likeliterally don't know they're usingSafari. Like it's kind of like that typeof an invisible surface. And I was likeI wonder what that's like. So I like Imet the team and uh started working withthem and it was just awesome. Like Iloved like it was technically extremelydeep. Like the

00:02:30things that they'redealing with were always rich. it waslike multiplatform, you know, like allthe different platforms have Safari onit and it's like extremely highly used.I was like this is a perfect scenario.So, and I I just clicked with the peoplethat were on that team and then ended upjust doing that for like the entirety ofmy time at that company. You say likethe invisible interface in many ways andyou don't even know you're using Safari,but it's like what were some of thekinds of things that you even work on assomebody who's designing a browser? likecan we go a little bit deeper into whatmakes designing this type of productunique because the vast majority ofpeople listening myself includedprobably won't actually get thatexperience.>> It's interesting like the positioning ofsomething like Safari which is likemeant to be ubiquitous versus somethinglike Dia or Arc which are more craftedto be an experience. This is the samerealm I've been working in but like fromtotally different perspectives like withSafari doing something that likedisrupts prior existing workflows or youknow like Safari has been around fordecades. People rely on their bookmarksbar in a very particular way. Takingrisks there comes with like hugeconsequences and also with such like anincredibly wide user base.

00:03:46You can'tdesign for any one particular type ofuser. You have to always be designingfor all users. Your decisions are thismatrix of complexity. You you can't justsay like this is the better experience.We're rolling with that 100%. Becausethere's always downsides. It's always agradient of usage. Like for example,let's just think about tabs. Like somepeople use like one tab at a time. Somepeople use like four to five. Somepeople use a bunch more. Some people aredoing multiple windows. Some people areusing profiles. They want to have abunch of windows open because they knowthat window that space is where theylike want to go. Exactly. Like you can'tmake a decision for tabs withoutconsidering literally every possible usecase. And that is always a veryinteresting way for me to think cuz itit takes me out of myself. Like I couldsay like this is how I want to use it,but then I basically have to immediatelydismiss that as that's just one tinyspeck in the spectrum of like differentways it's going to be used. And that'salways challenging. real quick messageand then we can jump back into it. So, Igot a sneak peek of DES's new privatebeta release and let me tell you, theyare taking the infinite canvas to awhole other level. It's not just [music]random mock-ups or one-off prototypes.You can run real product pages

00:05:01directlyon a [music] canvas. I mean, just thinkabout what that unlocks for parallelexploration. You can branch as much asyou want. Get the agent to exploremultiple ideas side by side, which isperfect for aligning the team becauseyou have one space to review prototypes,leave comments, and even get the agentto act on the feedback you receive. It'sa big deal. But here's the thing. Destinis only onboarding a few [music] teamsevery week. But if you go todive.club/desen,you can book a chat with theirco-founder, Gab, and tell them Ridd sentyou to jump the queue. So head to[music] dive.club/desen.That's dsn.If you still haven't tried paper, eventhough it kind of feels like everybody[music] on this podcast is talking aboutit lately, they just made it way easierto get started because you can now copyand paste from Figma directly intopapers canvas. All of the layers andtheir properties, including images andSVGs, are imported in an instant. It'spretty big deal and you can try it outtoday. Just head to dive.club/paperto get started. Now, on to the episode.When you're designing for a gradient ofuse cases like that, does that mean thatyou just have to be

00:06:16more data driven andeverything because you have to see allthe different perspectives? Like, talkto me about how you even figure out whatthose use cases are when you'redesigning for a billion people using aproduct a million different ways. Youwould think data driven would be a wayto do it and it probably is but in myexperience it's mostly feedback rightlike people will be like that's notworking for me you hear it you hear itquickly too cuz another element ofbrowsers is that people rely on them somuch like if their workflows areinterrupted in any way like they areextremely quick to be vocal about how itinterrupts them because they really likerely on it as a tool but at the sametime like they're mostly not thinkingabout it. So it's like you don't findout until you kind of make the changeand start testing it. So like at thebrowser company, our entire method ofdoing this is as soon as we have anidea, we put it in an internal build andeverybody's using it immediately becauseyou don't know like you just have noidea what's going to work or not untilyou just have it in people's hands. Solike this extremely quick iteration likeextremely quick that's failing, let'sstop right away, start again. likethat's the way we work and the companyhas always worked that way and it islike I think an extremely effective wayof doing it cuz I think we're likepretty public company in the way that welike express the work that we're doing

00:07:31but the amount of things that we do failand just drop and start over like it's alot more than what we ever express inthe world. So like the decisions thatend up making it to the app in public,they've always been like through aprocess. It's a very different processand like the very deliberate nature ofApple's design, it's um it's reallyspontaneous and like driven around likefeelings. That's something that is likeextremely important in the way that wejust think about it. Like did it makeyou say like this is better? You know,is it a feeling? For example, we've beenthinking a lot about clarity, like abreath of fresh air lately. So we haveum something called the morning brief atthe beginning of your day in India likeit'll like kind of go out there and seewhat's going on in your world and thendeliver you this brief about the thingsthat are most important to you. But theway it's designed is like when you openit, it's like the first thing you see isa painting [laughter]and it gives you like a little subheadline kind of just like describingyour day. But it's always like I'd sayon the optimistic side of things likeit's like you know your morning's busybut your afternoon you have time to workkind of things and then you have toengage with it. There's open space toit, right? So it's like I'm when

00:08:46Iscroll up I'm going to see like my firsttwo tasks maybe the the three but I'mnot going to like overwhelm you, right?If I open another type of tool I mightsee like oh god I've got like 12 thingsto do right now. But this is like heylike let's like just give you a littlebit at a time to make you feelcomfortable the whole time. So that thatthat design for feeling thing is reallyimportant.>> Let's go a little bit deeper there. Likeare there other mental shifts thatyou've had to make as a designer movingfrom a world where even though you'redesigning the same surface areatechnically it's no longer just pureutility. You are creating what you calllike an experience because everythingyou're describing with DIA here is it'sso many levels past the raw invisibilityof something like Safari. You'resupposed to know you're using DIA tosome degree. Like it's it's definitelylike toned down in its overall presencefrom Arc. When I started using Arc, youcould feel it like it was just likethese people are having fun. End ofstory, right? Like>> I had a denim brain on my sidebar, youknow, like Arc was there, [laughter]>> right? Especially for that moment, I wasjust like, "Wow, they are having so muchfun making this thing." And it resonatedwith people because like if you careabout product development with eachlittle update, you could kind of be partof it and

00:10:01be like, "Wow, that was funfor me. Like I can't believe you thoughtthat." There's the downside of whathappened with ARC, which is that itdidn't have a totally defined like itwas an huge experiment, a hugeexperiment that ended up working in alot of ways, but like it was overflowingwith ideas and at some point you had toquestion like well what is it? And Ithink like Dia starts from that pointwhich is I think we actually definedwhat we were trying to create here fromthe very beginning. It's just we wantedto have an easier on-ramp. When youfirst got into Arc the thing was like ohmy god there's so much new here which isgreat if you're willing to explore itand look through it. But for someone wholike switched from Chrome that morningthat they'd be like I don't know if Igot all this in me today to like learnthis whole new system. Whereas with Diayou start you could like keep using it.exactly like you were using Chrome 30minutes earlier and just go with it andthen like introduce the complexity overtime. Maybe be like, "Oh, there's asidebar." Like there's a little popupthat fires off once you start usingenough tabs. It's like, "Did you knowyou can use a sidebar? Oh, that might bea good time to start to use it." Or, youknow, like introduce the complexity thatexisted in Arc, but over the course oftime in a way that's like meaningful tothe way that you're interacting with it.So if you're the person who

00:11:16uses fourtabs all the time, you're never going tosee that and it's probably not reallyrelevant to you. So like making a lot ofdecisions like that from day one hasbeen I think a success in DIA.>> There's a ton to talk about around thistopic. So, what I kind of want to do isuse your personal design principles as alaunching point because you've writtenabout these and pieces on your websiteand I'm wondering if you can kind ofjust give us the rundown and then we canuse those as a lens through which toview some of your work and talk aboutsome of the design process and thingsyou're wrestling with with DIA. This isan evolving type of thought, but likewhere I'm currently at is I thinkthere's like three values that work forme that have they they interplaytogether in a way that creates like alittle thing that I believe in, which islike I think that the concept of likesimple design is always attractive tome. It comes with downsides. I thinkthat aiming for practical decisions endsup being really valuable in a lot oflike you can be extremely whimsical.There's all sorts of different types ofdesign that you can do that arebeneficial for various situations. Butwhen working on utility style appsespecially, sometimes the practicaldecisions end up being the most valuableand cause the least interruption of likeuser

00:12:32experiences if they're coming fromsomething else. And then finally, it'sjust doing these things at like anextremely high level of quality like acraft bar that's really high. So you canadd back the personality through thatstep a lot of the time. But really whatit does is if you have these simplepractical decisions that are executed ata really high level, I think it givesyou a feeling of confidence often withlike the experience you're interactingwith. And it's like the type of appsthat I'm drawn to like you know thingsthree or something like that is like ait's like wa it's a to-do list.>> Yeah. But at the same time it reallyworks like it's been thought through allthe way through and it it just works.What's interesting about these things atthis point in time is like I think it'sreally important that people have like aframework of thinking for themselvesdefining what they want to be doing withdesign because like as you know toolsare changing rapidly around us and it'svery easy to get pulled into the tooland the tool guides your path of likewhat you're doing more but if youconstantly have like this framework ofhow to think behind it then the toolsyou'll use in the way that you're liketrying to execute that frame frameworkof thinking rather than beingpotentially more affected by a tool andexecuting

00:13:47things that don't exactlymatch what you probably want because youknow like well if I can do this quickand it's awesome that might create anoutput that doesn't exactly match likemy goals. So I think it's like moreimportant than ever right now tobasically have these like systems.Whenever I bring up like values orprinciples like everyone like is likenah like I I see it in everybody's eyesthey're like nah. No one really wants tolike think or talk about those things,but I do really think over the course ofa lot of different work that you make ifyou really have something like that, itstarts to be thematic through your workand then like people know what you wantto do and you'll get the type of workthat you want to be doing. So, I thinkit it ends up benefiting you a lot inthe long run.>> I'm going to put a pin in simplicitybecause I want to come right back to it,but let's add a bit of clarity because Ithink you were talking about beingpulled by the tools in an interestingway. Is there like a hypothetical or anexample of that failure state that youcan share? Like what does it look liketo not be rooted in values and let thetools pull you toward a place where youprobably shouldn't have ended up going?>> I can write code and I can writeproduction code now in a way that Icouldn't before. It seems veryintriguing

00:15:02sometimes for me to just belike, you know, I can just go make thosechanges. What I find a lot of the timeis the scale that I work at the browsercompany. It is more impactful to kind ofkeep creating designs and sharing thosewith engineers rather than going toomuch into the work specifically. If I amdistributing more designs, I amdistributing the problems that comealong with those designs. Whereas if Itry to do it all myself and fix theproblem, sometimes I will just not Iwill be wasting time where I could beenlike executing other designs which arehelpful on other ways and I'm justtrying to think about like what myspecial sauce is in the equation and Ithink what I because of these values orwhatever it's like I tend to like reducethings back. I I can pull things back. Ican make it simpler. I can make it moreunderstandable for the user. And yes, Icare about making sure that button worksexactly right or whatever. And I couldhave went in in there and fixed it. Butif I keep staying like on this end ofthe equation where I'm thinking in thisrealm, I'm more novel over there than Iam in the execution realm cuz likeeveryone can kind of execute right now.>> Yeah, that's a good totally right.Everybody can execute. Even yesterday, Ihad this idea where, you know,

00:16:17we'relike spinning up local preview builds inthis Mac app that we're kind of playingwith. And I had this idea for thisloading state inside of the tab. Iactually worked on something that feltsimilar to dia tabs in one of myexplorations. I ended up going adifferent direction, but I was like,"Oh, I'll just bring that in here andit'll allow me to delete this entireother primitive." Like, wow, that's sucha good idea. I still think it's a goodidea. But I started building it and as Istarted building it, I started realizingwe're not handling this type of air caseas well. And then I'm like all of asudden I'm looking at all of thedifferent air cases in the Aurelius andall of these different failure statesand I had this moment of like oh this isa trap. Like I have went too far. I'mnow 90 minutes into designing for airstates that frankly I have no businessthinking about. It's not my specialsauce. My special sauce was the idealike the system level idea for what thiscomponent could be. And if I would havestopped there, I probably would alreadybe working on something else where Icould deploy my special sauce and Iwouldn't be lost in these air states.That trap exists every single day for menow. And figure out where that line isis like kind of the new challenge for adesigner like when you said whereeverybody can execute now. Like are areyou really differentiating on theexecution level? Maybe, but maybe not.>> I feel embarrassed saying this, but it'sit's like almost in my personal lifelike I write more code than ever. likeI'm

00:17:32in cloud code kind of like everynight as like my downtime or somethinglike just working on my website orwhatever. But in work I've actually likeprobably shifted more towards Figma. Andwhen I say that I mean I'm shifting evenmore away from prototyping which soundsinsane cuz like I've always consideredmyself to be a prototyper but it's likeI actually am like hey this is it. Let'sget it going and then we'll like figureit out down the road. Here's the visual.Here's how it basically works. Let's getit going. And then like we come back toit and like add in the magic [laughter]slowly over time. But what I find islike yeah it's like it's actually kindof almost novel feeling right now. Andnot everyone works this way. This islike you know I'm I think I'm likenervous about it because it's like itsounds so radical for this moment butit's been effective because like peopleare just cooking up code so much fasterthat it's like well they can actuallyhandle more designs. maybe that's what Ishould be doing because like I do knowhow to do that part for sure and noteveryone has that particular skill set.So,>> it's funny that we live in a day and agewhere this is like an incredibly noveltake, [laughter]>> right? Like, yeah, it feels backwardscuz it's like, well, yeah, of course, Iwould want to be more involved in theoutcome, but kind of relying on biggercompany experience.

00:18:47Communication isalso a skill of design more thananything. And being able to communicatequickly, get it through, like set theboundaries of what we're doing, whatwe're not doing, like how far to takeit. That actually is something I feellike I've been investing in more aswell. Just how fast can I get you theinformation you need? How quickly can Imake a decision that's like sturdy here?This year I've been doing a thing whichis in my Obsidian I have like a filewhich is just called decisions. Andanytime I make something I consider adecision, a design decision, I write itin there and then I go back a coupleweeks or months later and say like wasit a success? Was it a failure? or whathad to change. And then I'm likeessentially going back and being like,okay, where am I actually making baddecisions? Because that's the thing thatI think is the highest value of me inthe future is like if I could make moresuccessful decisions, then I thinkthat's kind of what design is going toturn into.>> Are you able to see places where you'vetweaked your practice as a designerbased off of those decision logs andreflecting on them? Yeah, for me thebiggest weakness is like impulsiveness.So it's like if I don't take a beatsometimes I might make the quickdecision and it might

00:20:02seem totally finein the moment. But often decisionsrequire an extra level of awareness thanyou even expect, right? Like it may seemsimple in the moment, but you at leasttaking a beat before the decision andlike reconsidering the deeper set ofoptions every time prevents yousometimes from pigeon hauling down intoa bad path. I just like to be quick andlike respond quick and like act quick.But the fact of the matter is is that'swhere I'll make my mistakes often if I'mjust being a little too rapid and notmaybe just reconsidering. Andreconsidering often like slows peopledown and they like don't want to hearlike we're deep in this thing and like Idon't like we think we're pretty muchalmost done but it's like sometimes justtaking a beat and being like okay are weactually on the right path is this theoutcome we want and just going throughit one more time before saying like yeahthat's what we do that could have savedme a few times.>> Yeah [laughter] totally. It'sinteresting too even to connect withwhat you were saying before. It's thatextra loop and practice of considerationthat I find often is the only way that Ican get to simplicity. I might have theright idea as my knee-jerk reaction. Weput in a lot of reps. Like that's notuncommon for me. Sometimes I return backto the idea, but then it's like, okay,knowing that this is directionally

00:21:17correct, but then going through it onemore time and be like, what can Idistill or cut or combine? Yeah, this issomething I think Apple just does betterthan anyone else. And I think it'spartially because of sort of this likeone-year release cycle thing. It's likeat any point in that cycle, you can kindof be like, is this really the rightway? And you can go back to thebeginning and it's like people kind ofalmost expect that disruption as part ofthe process. Like it's very much part ofthe process to just be like actuallywe're on the wrong path. Let's startover from this position. And I thinkthat requires a lot of confidence and itrequires a lot of low ego or somethinglike you can't like just to be able tolike and just emotional awareness andcomfort. It's like you know it it sucksto be like deep in something and be likewe're starting over and by the way nowlet's do it next year like cuz we justdon't have time. So you know that likethat that can suck. So it sometimes itreally is the right call and it itrequires Yeah. like a lot of confidenceto do something like that cuz itconfidence that you're going to sticktogether through it.>> Let's talk about simplicity through thelens of the type of product that you'vebeen working on because I read yourwriting on simplicity and you referencelike old calculators and how they'revery purposebuilt. They do you know

00:22:32really kind of one thing you're justdoing calculations.And my question then is like how do youthink about evoking the benefits ofsimplicity when you're working onsomething not only just like a browserbut also a browser in a world where AIis just woven through the very fabric ofwhat the product is. Therefore it cankind of do anything you know.>> Yeah I think it's often about justmeeting prior expectations on one level.So for like example like working onWhatsApp was I think where I got intothis mentality more than ever which islike it's a chat app you should be ableto chat in it. Anything beyond thatactually dilutes the fact that that is auseful interaction in your life. Right?This was kind of from the founders. Theyhad just this like incredible drive tomake it like simple, fast, secure. It'slike they had just really strong valuesabout what the thing needed to be. And Isaw like, oh, if you hold on to thosevalues through a product's life, peoplewill believe in it. They'll like trustit, right? like WhatsApp is very popularand people use it to communicate allthat got me really excited. But thenyeah, like any sort of like utilityproduct like if you're relying on thething to do a task for you and you wantit to do that one thing like it's justgot to be like that's the focus.

00:23:48And Ithink like in that writing that you'rereferencing, I think I kind of describedlike, oh, if your calculator startedbeing like, I have a software update orlike um you know, like there's thisother feature that you can turn on nowor something like that, you'd kind of belike, I'm just trying to make acalculation. And I think that softwareis so, you know, so beyond that pointwhere like accept all of these behaviorsin our software that probably wouldn'tbe acceptable [laughter]in in that world where you're thinkingabout just a hardware device that youbought with your own money. But I thinklike a paid product is kind of likethat. Like if you pay for something,you're expecting something out of thatthing and you should have that type ofexperience primarily. Like yeah, there'sstill room to be like try out theseother little features on the side orwhatever, but like the focus should belike whatever value you're getting outof it. Also, in that piece though, Ikind of like show the contrary, which islike no product ever really fullysucceeds that way. like you could like Ithink I don't know if I said it in therebut like notion kind of comes to mindwhere it's like notion does kind of likeeverything in some ways the way Ithought about it originally was like ohthat's a nice simple like textapplication right but it's like it'sevolved so much from there and I thinkthe counterpoint to simple design isthat people want less tools a lot of thetime if

00:25:03I can do it in notion I'll do itin notion cuz I already use notion>> totally>> and so there's like a value to addingcomplexity you see this like I likemaybe Microsoft is a kind of the bestcase of this ever. It's like I don'thear often like oh like Microsoft designis like the pinnacle of like simplicity,right? Um it's it's kind of theopposite, but like oh people got thevalue because you had all the differentthings you could do out of it, you know,and it's like they don't do all of thosethings perfectly well or whatever, butyou could do all the things out of itkind of like was the vibe. We'reexperiencing the pull of that asdesigners more than any other point inhistory too because we get to outsourceso much of that new functionality to themodels themselves. So it's like whereveryour entry point is to the model, youkind of can do anything, right? Like ittakes a very barebones set of primitivesto allow someone to literally doanything. And that temptation to grow inevery direction simultaneously has neverbeen more present. So, I'm going to likepush on some of the things you're sayinga bit because most people today, I mean,by most it's probably like 99% of peoplearen't actually really using AI as acore part of their browsing experienceother than maybe like a pre-generatedanswer by Google at the top of Chrome,right? So, like the value proposition ofDIA, at least as my interpretation ofit, is hey

00:26:18actually like AI should be alittle bit more integrated into yourbrowser. Use DIA instead. I wouldimagine you you'd feel a little bit morepressure to pull people into some ofthese more interesting AI use cases thatare potentially tangential to the corejobs to be done that you associate witha browser. How much does that resonateand then what is that like as adesigner?>> I think the evolution of DIA has been westarted from this idea of like what isan internet computer and now you havethis new technology as part of it. thelevel of the LLMs that we were dealingwith early on made us think of it askind of like a browser with chat. Thatwas kind of like our first, you know,that's where we kind of landed. Then theLLMs got better and we realized thebrowser can make things. It can make youreports, can make you uh presentationslike it can do all this nice stuff. Sothen it kind of evolved into like abrowser where you can also make thepages essentially, right? So that waskind of the next thought. I think whatwe've come to realize now is like thetruly valuable state of this thing isthe browser is context and it'seffortless context. You're just doingyour work

00:27:33in the browser, but thebrowser is capable. So like I think themorning brief like kind of defines thisas the first step, right? Like it's likehere's all the things in your life and Ican report back to you on them in a waythat helps you move forward and youdon't have to do anything like and I'mjust adding value because you're usingthe tool that you are already using.It's a constant evolution as thetechnology changes underneath us of likewhat we're capable of doing. But likethe end goal is an internet computer. Iknow that sounds like really bombastic,but it's like it is a thing and thisshould be your interface into that typeof computing and the more that it can dothe better cuz like you know you spendso much of the your time in the browserthat why not have the features there. Soit's like going back to the simplicitything it's like how do you make thatfeel nice and simple?>> There's one question that I can't stopasking myself. What if companies appliedto talk to you rather than the other wayaround? And that question is thefoundation for the all-new dive talent[music] network and it's working. Likeright now, I'm helping many of the mostexciting startups that I know to hirethe designers and builders who listen tothis show. So, if you're curious whatmight be

00:28:48out there, and maybe you wantto get on my list, or maybe you're evenlooking for your next design hire,[music] head to dive.club/talentclub/talent to join today. I remembersitting on a plane reading Josh'sinitial post about the internet computerand finding it very [music] compelling.I also remember interviewing NateParrot, the first design hire at thebrowser company probably a couple yearsago now. And one of the things that hetalked about as call it a prototype thatyou all were wrestling with internallywas is there a new tab home screen likeis there if you hit command T do you seeanything in some ways like the dailybrief is it's an extension of that sameidea right like that strand has beenforever I I almost anyone who's designeda browser has thought of this right whydid it work now like what did it takelike what were some of those internaldebates in terms of what the morningbrief became different designexpirations, what didn't ship, anythingthat you could do to add a littleclarity around what that design processwas?>> The first thing we kind of did was weintroduced something called new chat.And this was, you know, this higher tierof LLMs in the browser. We were like,"Wow, it does all this stuff." And itwas like kind of like overwhelming, butwe just had like an internal

00:30:03discussion.We're like, "It'd be so much cooler ifit just did this automatically.">> It was that simple, right? So, anddeliver to you. Like, I don't want to doanything. just give me the value of thecommon things that I'm doing with it andlike everyone was kind of being likewhat am I not caught up on you know likethose were the type of things thatpeople were pulling out. So the morningbrief just came to you answered thosequestions without you doing anything andthen we immediately saw like oh thatthat's hitting with people like theythey found it valuable and so that thatwas like a good starting point for aproduct. [laughter]Can we talk about the practicality piecea little bit given everything thatyou're working on? I don't know. I thinkdesigners everywhere always feel thispush to do the novel thing, right? Tocreate the pattern that's changes theindustry, you know, what is the pull torefresh for the browser kind of thing.I'm sure that you're experiencing a lotof that. Like how do you stay grounded?What does it take to know when to pushpast the obvious solution versus when tojust pick the thing that people arefamiliar with? This team happens to be ateam of people that are all willing totry to come up with those like newparadigms or new experiences and pushforward. And then I think we also havethis like counter opinion which is likesometimes the obvious solution is the

00:31:18right solution. So like I think in DIAfor example the bookmarks bar like wetried everything. We tried a littlebackpack where you throw stuff in. Wewanted bookmarks because we wanted tosupport like if you're coming over fromChrome like you got to have access toyour bookmarks and we were like you canmake those amazing you could make themlike I think we have the design articleand it shows like a small screenshot oflike eight different directions orsomething we chose but it was like wetried everything. We tried it for monthsand in the end we shipped a bookmarksbar that's like nearly identical infunctionality to Chrome and that waslike choosing the obvious solution. Itwas not that we wanted to make it abasic bookmarks bar. That didn't soundlike what we wanted to do as designers,but like at some point the egos kind oflike just go away and you're like,"Yeah, people actually who want thisfeature just want to use it the sameway." And you can turn it on, you canturn it off just like a normal bookmarksbar. You can have folders in it. You canopen them the normal ways you would. Andlike it just met expectations. It hadlow downsides, right? like other thanyou know us wanting to do somethingmagical with it but it wasn't an easyprocess. It was a very difficult processto come to that kind of a conclusion. Wejust couldn't find something that was aremarkably better like for like one wehad

00:32:33was like anytime you took a tab youcould just like pull it up a little bitand the page would move down and youcould like put in these little stashes.It felt awesome. It was cool but it waslike less discoverable. It was a longerinteraction. You know, it's like it justcame with downsides like animationstates, like stuff's moving around. It'slike the bookmarks bar. It's like if youwant the thing, all your bookmarks arethere. It's really dense. You can pressthem. Let's just make it look as prettyas possible. Move on to the next thing.[laughter]>> If you think about having like a fixedbudget for novelty as a designer, wherehave you been more excited or willing tospend it? The first stage of DIA waslike, can we keep our novelty budgetalmost entirely to the new features ofthe technology that's being developed?We shipped only with a top bar to start.We shipped with a bookmarks bar. Weshipped uh you know just basic pin tabsand stuff like that. Like it was like itwas bare bones to start. That wasn'twhere the exploration started. We likedecided at some point like the noveltybudget is going to be completely on thenew things that you have to learn usingthis app. and we wanted room to grow inthat area. So the more comfortable youwere with the pre-existing UIs that youwere used to, the less you'd have totake on mentally.

00:33:49When VIA came out,like I know this is silly, but it waslike it was a little novel to have likechat just built directly into thebrowser as like you know, we tried somany different ways to like integrate ittoo. And like we kind of kind of wentwith the one that would be the mostexpected which was the sidebar is kindof like the primary interaction becauseit's like I have the context of the pagenow I have my chat and it it works rightso even that was like um kind of like areducing the novelty as much aspossible. You have kind of been spendingit in little places by bringing in someof the arc functionality and almostre-evaluating what needed to existagain. Can you talk about that process abit? Like maybe you can even give us abehind the scenes of how you're thinkingabout bringing the arc spaces into DIAand what's that like? Like how do youfind the right implementation for thismore refined vision in DIA? We have achecklist of all the things that arcdoes really really well that people arepassionate about and we kind of havejust been moving down it one by one onlike the order of what people are sayingis the most important one. So like youknow at one point there was no sidebarand that was kind of like a moment forus. We were like okay we are going tolike it was always a possibility to dosidebar top bar but

00:35:04one is we wanted todifferentiate DIA enough in V1 that itwasn't our final cut pro 7 I think wasout and then like Final Cut Pro X cameout or something like that and it waslike they tried to kind of match all thefeatures but it didn't match and peoplewere like really upset or this happenedwith like Lightroom CC and likeLightroom classic or whatever like itwas like not a total match. And when youhave that and they're so close but theydon't match, you just get like kind ofpure disappointment out of the new thinguntil you can build it up.>> But like we were hoping that DIA wasdifferentiated enough that you're likethis is a different browsing experience,right? But yeah, like we love all thesefeatures in ARC and we want them in DIAcuz we want to feel satiated by thestuff that we enjoy at Arc too. So it'slike just listening to people, seeingwhat's right. So we started withsidebar. I mean maybe we started beforethat but like cybar was a big one. Withevery single one of these features thatwe introduce from arc we do take itthrough the full process ofreconsidering like where could it beimproved? How does it fit in with thedifferences of dia? Can we just do itbetter in any particular way? So withthe sidebar for example like one of themain differences between dia and arc isarc is really like a single windowapplication

00:36:19which is kind of like alittle unusual for a browser. Like ifyou open another window, it's likeeverything is essentially synced betweenthose two windows. You can have them indifferent states, but it's like it's onething where DIA allows instancedwindows, right? So it's like I can haveone window that's completely differentthan another one. You know, there's uhconsiderations that need to be takeninto account with that like if this oneis in profile X and this one is inprofile Y, what what is the overlap?What is shared between the instances andwhat is not? So yeah, every feature thatcomes through, we try to like run itthrough the whole process. Also designreally widely. Think about like maybesidebar's on the left right now. Whatelse could you do? Like what are theother ways that you could show verticaltabs? We tried other things. Sidebarstill ended up like being pretty muchthe one that checks the most boxes. Thenwe went forward with it. What's funny isa bunch of other products in theindustry went through the exact sameexercise with the sidebars and everybodykind of just arrived at like, yeah, ourkind of got it right.>> Yeah. And now there's like like a lot alot of browsers that offer a size>> and they all look exactly the same.[laughter]>> Okay, let's talk about the quality piecea little bit. I'd love to see what thatlooks like in practice in DIA and whereyou all are intentionally trying to like

00:37:35go above and beyond and push past thebaseline from a craft standpoint. Ithink a lot of the quality moments inIndia come from really like small ornearly invisible moments. Like Casper onour team added these like really niceinteraction states for the littlebuttons like when I reload it spinsturns into the X and it pumps pumps backout. If I like uh the arrows move whenyou tap them or whatever. These are justlike small moments in the app but it'sthings you may do a lot and you kind ofget like a sense of the refinementthrough them. something I worked onwhich I still feel really excited about.I don't think people probably thinkabout very much but like uh the URL barwe changed how it works. We show thedomain and the page title rather thanthe domain and the URL but at the sametime you're not blocked in any way. Soas soon as I hover into here I haveaccess to the URL but when I'm not onthere I have the page title which cansometimes be useful. This was like aextremely small moment, but to me italways it adds value. Like maybe becauseI paid a lot of attention to in thefirst place, but I'm like it just it's agrounding sort of thing and it you don'tsee complex dirty URLs with a bunch ofuh numbers in them ever, you know, youjust have a sense of the

00:38:50page you're onand that felt like the next evolution ofa URL bar to me. I really like thatexample because you only arrive there ifyou are going through a browser's UXfrom first principles with like a veryfine comb, you [laughter] know, like I'mnot sure how long it would have taken meto arrive at that as even a potentialthing that you could do. It's very cool.>> Um, another one's like if you copy apage link. So, if I hit command shift C,um, it'll just quickly change that URLbar to give you the status of whathappened in the position that it shouldhave happened. And actually, what's alsonice in here, like another nice littlequality moment, is if you have a URLthat has like trackers associated withit, um, and you copy that URL, it willremove those trackers and give you aclean URL, but then it also alerts youto that in the text. It says like, uh,copied a link without trackers orsomething like so. That's like another,you know,>> gesture of quality.>> Help someone reverse engineer some ofthis stuff. Maybe it's going back towhat we were talking about earlier wherelike the execution level details are notthat crazy for any of this stuff. It'syour ability to see them and have thecreative idea that like, oh, wouldn't itbe cool if X Do you have any advice fora designer who

00:40:05wants to stop listeningto this video, turn to their product andfind some of those quality details?Yeah, I would say that this is thecounterpoint to my impulsiveconversation earlier, which is likeimpulsiveness in decisions can beproblematic in the way that you can makebad decisions. But impulsiveness in yourdesign is actually something I think youshould facilitate all the time becausetrying out like one random thought popsinto your head, go down that path. Tryit out. See what happens. Like I thinkCasper is like a really good example onour team. like he's just looking aroundlooking at the buttons like what are thebuttons that are here all the time whatcan I do with them and then like ideasjust come to you and you try them andbecause of the browser company's justability to facilitate this try anythingwe don't know attitude he'll just throwthem in the app and then you start tofind out if these things are valuable ornot I think it's it's just about wastingtime I guess I I would describe it I canstay on task I have these things thatare really important and I have to getthem done because you know the companyhas goals and like all these things aremoving things forward, but if you're notplaying or you're not like messingaround, you won't come to those moments.Yeah, I like that. Like the question,what can I do with them? Is sointeresting to

00:41:20me cuz even looking atthat URL bar, I think that's animprovement. I really like it. Nobodyever is going to come to you with aproblem and say, you know, I have thisproblem in this piece of software I'vebeen using my entire life. problem>> the URL bar, you know, like, God, if youcould just solve my No, nobody's ever,there's not one person on planet Earththat's ever been like, I have a problemwith my URL bar. It's just the way thatit is. You only arrive there throughplay.>> Here's another one that doesn't solve aproblem, but it kind of does, but I'llshow it. Um, when I open a new tab page,there's this animation and you see thissort of like prism effect move up intothe input one. It ends up focusing youreye to the correct element. So, like youcould argue that wasn't a problem, butit actually helps a bit. [laughter]>> Yeah. And um and and two, there's not alot of moments where you can actuallyfit animations in. Like for example,moving between pages. I should likethere should be a crossfade, right? Thatwould feel better, but it always feelsslower. Like that's like I've tried many[laughter] times to have like some sortof state change between pages. It justhas to be a snap. But you can do thingson the back end of like loading a newtab page. Like the page is loaded. I canstart taking action while that animationplays out, but it still like gives me asegue into the experience.>> Mhm. Mhm.>> Here's a one

00:42:35that doesn't need to exist.Adam, who's an engineer on our team,totally made this thing a fidget spinnerfor no reason.>> This element is also used throughout theapp. Like when I'm getting chatresponses like, "Oh, me about it. I'mtalking to him right now." [laughter]Um, you'll notice that this thing goesout and it's it changes from the new tabpage into here and now it's going togive me status, you know, about um,what's occurring and it it acts as sortof like the focal point. Um, so it'slike that continue that continued UIends up being just a way to drive youforward. [laughter]You guys have done some work on theseinteractions. I'll be honest, like I'musing Arc. I'm recording this in Arcright now. I'm on the fence. I'm gettingreally really close to coming back toI'm really close spaces, you know,that's like one of the big things forme. This level of interaction detail hascome a long way in the last eight or somonths. Like that is gorgeous stuff.>> Arc in the state that it's in wasdeveloped for like half a decade. And Ithink that was a bit of a confusingtransition moment for users with DIAbecause we had only had like a year intoit when people really started likejudging it. And um the idea is thatwe've just been like pumping non-stop

00:43:51effort into both the micro interactionsas well as the main features. And I Imean like I paying a lot of attention toit cuz I'm so invested in it. But it'sreally at the level now where you canappreciate these things and the overallexperience. About maybe 6 months ago, wewere like, "Oh, people love Ark." Likewe want them to love DIA. So how do weget them there? Like how do we get themto have that feeling that they lovedabout Arc? It's like it's a differentapp. It's like a different experienceoverall. So, it wasn't like it was aplaybook that we could just like doexactly the same. It's like we had tothink about these things from some newperspectives, but trying to bring backthat feeling, emotion, and getting youinvested in the thing more than just uhpure utility. Before we switch, anyother craft details that you wanted toshow? So, another interaction that wehave both in Arc and Dia is being ableto split tabs within your window. Andwith that, we we wanted to make it likemore clear how to do it because it's notmaybe the most common interaction, butit's it's it ends up being really usefulin a lot of situations where you justare trying to compare information or doone thing in one tab and see theinformation in another tab. So, um, webuilt this paradigm where you could, uh,drag out the tab and then,

00:45:06uh, AdamStern built this, I should say. He's anengineer on our design team, and it justkind of like guides you to what willhappen. And like basically just thelanguage on the drop target then allowsyou to know more clearly what you'regoing to do. Like, it actually comeswith some downsides. There's like a lessspace for you to drag the tab, whereasin the arc paradigm, it's like just thewhole half of the window. But it theeducation element we thought was likeimportant to this um making it moreclear like what is going to happen anduh yeah it was like a cool littleinteraction he came up with and reallylike that one.>> That was amazing. Charlie I appreciatethe tour and we're approaching the endhere but I I have saved the deepest mostgripping most important question of all.That question is who is your favoritedesigner named Gabe?>> Gabriel Vald Valdivia. Like this isn'teven a question. I um he might be myfavorite designer. End of story.>> Oh, and>> his ego his ego is going to explode whenhe hears that. [laughter]>> Well, I uh I was texting him last nightand I asked for a few quick hittingquestions that we could ride off intothe sunset with and the there's somesome practical and educational andothers are a little bit more playful.The first one I

00:46:21do think that weeverybody listening has a lot that theycould learn from you on, which is whatare some of the less obvious tips thatyou've added to your practice as adesigner when it comes to presenting andgetting people excited about your work?>> Yeah, I think this one I actually feelreally strongly about. So, greatquestion, Gabe. You're doing wonderful.>> It's got the context. [laughter]I have a pretty particular strategy forpresenting information, which is Ireally aggressively guide theconversation to a certain point. Like Iwant people to be on the same page whenthey're having the conversation. Andthat often requires me giving context asfast as possible, but as clear aspossible to get people up to speed withwhere I'm at. Cuz like in general, Ioften think that designers havesometimes the most context on a problembecause they're in it, you know, likethey're they're like thinking throughall of the various things from and it'sjust it's repeating in your head, right?But like maybe I'm in another positionand I'm coming to this to kind of likehelp make a decision and I've thoughtabout it from just one particular angle.So like one spreading that context out,giving everyone like the same startingpoint, but then also framing veryclearly what the problem is. we'retrying to solve

00:47:37so that everyone can belike, "Okay, we all agree to this pointthat this is what has happened and thisis where we're at. But this is the nextstep for us." And like, yeah, my job isto provide a solution and be cool aboutit if people don't like it. [laughter]But like, but it's more important that Iprovide a question to have an answer toso that at the end of whateverconversation we're having, we can allagree that we answered that question.And then like in practice like I oftenthink you know it's like I don't showthe design necessarily right away. Ilike kind of overexlain it like thisfirst. I try to reduce the informationthat I'm presenting down to just likethe essential points because people aregoing to ask whatever question they needto know to fill in the gaps. Like nobodyhas a problem really with that. Likepeople will ask the questions but it'smore so like trying to make itdigestible so that everyone's on thesame page and keeping it really highlevel to do that. I think that that'sthat's kind of like my strategy forconversations in both big companies andin one-on-one environments. It's like Ithink it feels pretty effective ingeneral.>> Is there an example either real orhypothetical that you could share aboutwhat a good like foundational questionto answer looks like?>> The type of question I would try tofocus like a meeting

00:48:52scenario on isoften hopefully lower level than thebigger problem, right? like hopefullyI've done the work that we're in theright space already and I'm solving forthe right general problem. What I oftenwill present as the problem is somethinglike okay if we take this path forwardthese are the downsides if we take thispath forward these are the downsideshere's the advantages you get on thisside here's the advantages this sidethis is like a matrix now that we cankind of point our fingers at detailswe'll deal with those but like if we gothis path I have confidence in my nextsteps and I can move this projectforward people will always feelcomfortable like breaking out of thatand being like we're not in the rightproblem space if that's the case. Butlike for the sake of a conversation,limiting it to whatever needs to besolved next to move forward. And that'soften sometimes like bringing it down alevel, which is sometimes different thanlike if you're going to like bepresenting up to like an executive orsomething. And they like they reallyneed the context of the meta decision.Like that's the type of decision they'remaking, the like really hard complex onethat may be more vague or whatever. Butif it's like, you know, design crit orsomething like that, it's like, let'sjust like solve this one right now or>> Okay, I have two

00:50:07more Gabe questions foryou. The next is, how do you manage thepanic that comes with high pressuredesign environments?>> I have a lifelong anxiety disorder. It'sjust natural for me to feel like tunedup. What's cool about design is somehowI've turned it that anxiety into justlike motivation. Often I like getexcited. Working with Gabe often doesthis to me a lot, which is like when wework together, we're both probablyreally anxious, but we get reallyexcited and like push each other andlike we're just like do it. It becomeslike really frantic or something. I justthink that turning the anxiety intoenergy and motivation, it keeps me goingand excited. And I just associate designwith being excited. That's how I dealwith it. And like I try not to thehardest part is always like trying toremove the personal element a littlebit. The hardest part about coming up indesign is like you're going to bepassionate about your ideas andsometimes people are just going to thinkthey're bad and you might even be right.Like it's still a good idea or it's likea good execution and people might bedown on it for all sorts of reasons thatyou can't expect. But like being willingto go with the flow, try out what otherpeople are suggesting. I was an onlychild growing up. So, it's like it oneof

00:51:22the challenging parts for me was kindof like being truly collaborative in thedesign process. But at some point whenyou're like internal confidence comes upto a level, it gets way easier to becollaborative because I'm like, "Yeah,I'll try it." You say it like Gabethrows me something, I'm like, "Yeah,I'll try it." I don't even question it.Even if I don't like it, it's like,"Let's try it. Let's find out." I mean,people really appreciate collaborativedesigners, people who are willing towork with you and whatever you'rethinking about, but also, you know,coming to it with like an excitement. Ifthey see it back and they're like,"Nope, that's not what I was hoping itwould be like, then boom, we're on thesame page now. You're probably going tobe maybe more into the thing that I wasinto and we're moving forward." It'slike, it's just a way to keep it goingreally fast. I always struggle to makesure that I'm dedicating the amount ofcreativity and mental resources to thethings that I kind of know are the badideas. I'm like, I'm going to put in thereps, but I don't want to sabotage theidea, you know, for my own selfish gain.I've been>> It's just a sample like this is what itlooks like. You wanted to see it. Here'swhat it looks like. We're going tohopefully be moving on quick.>> Looks just as bad as I knew it was goingto look. All right, last question beforeI let you go. When are you bringing backbread time?>> Okay, there's the secret episode. So, werecorded I think 2 years ago,

00:52:37but weboth like we we recorded it video stylelike all modern. We both were like, "Notnow." We finished it and we both said,"Not now." Like, it was it was very fun.For the 6 hours after we recorded, I waslike, "I can't wait to do this again."And then I think we both realized it'sreally important for us to focus on thethings that we're doing in our primarystuff right now. And it would end upbecoming a pretty serious distractioncuz like when we did it we were all inlike it was really like took up a lot ofour mindcape and took up a lot of timeand I think we would want to do it withthat same level of passion again. Butdoing this type of stuff, like being inpublic, it's a stressor. Puttingyourself out there, like for anyonewho's ever like going to be talking on ashow like yours, you know, it's likeit's it's dangerous and that that's partof like, you know, it's going to be onyour mind in different ways to putyourself out there in the world. So, Ithink we just decided like now is notthe right time, like maybe again in thefuture. You just give me a heads up whenyou're going to come for the uh thedesign podcast game so I can figure outother ways to feed my family. I justneed like a few months heads up.>> We will not be a competition, but wewill be a weird

00:53:52other thing that peoplecan appreciate hopefully. [laughter]>> Well, I I'm It's always fun. Like it'sfun to listen to both of you, but also Ijust really appreciate you coming ontoday and you've given me a lot to thinkabout. I I still even have some of thesecret sauce and creativity versusexecution ideas bouncing around in mybrain. I really appreciate you bringingthat perspective and sharing a littlebehind the scenes. Always good to catchup.>> Yeah, thank you so much for having me. II told you off fair, but it's like Iwatch every episode of the show. Ireally think you're doing a great job.So, it's very much an honor to be hereand I appreciate it. Means [laughter] alot. Thanks,