We chat with Anil Dash, CEO of Glitch and board member here at Stack Overflow. He breaks down the tech behind Glitch apps, explains why the company is launching an online magazine called Glimmer, and talks about the fight to keep the web weird, fun, and open to all.
Glitch, a platform that makes it easy for anyone to create or remix a web app, has seen over five million apps created by users. You can read more about how it works here. If you want to learn a little about how it works with Docker, check out this piece here.
If you want to know more about the shared history of Stack and Glitch, you can read up on it here. TLDR; Glitch was born out of Fog Creek software and counts Joel Spolsky and Michael Pryor as founders.
Glimmer is a new web magazine from the folks at Glitch. It focuses on creators and makers, with a special emphasis on unearthing the human stories of people building today's software.
While you're here, don't forget to take 15-20 minutes and share your opinions in our 2020 Developer Survey. Whether Stack Overflow helped you during your journey as a programmer or not, we want to hear from everyone who codes.
Some fun background for younger listeners:
Geocities - a popular platform for building and hosting a personal website and linking it with others that share similar themes.
BetaBeat - a website launched by The NY Observer that covered the SIlicon Alley tech scene. It was how Ben first met Anil, Joel, and many others.
Docker
If you have comments, questions, or suggestions, please send us an email at podcast@stackoverflow.com
Today’s episode is brought to you by Refinitiv. Unlock new possibilities with consistent, high-value market data from Refinitiv. Try the Refinitiv Eikon Data API for the largest breadth and depth of data and community tools with native Python support. Check out refinitiv.com/stackpodcast to try the Eikon Data API today. Refinitiv. Data is just the beginning.
The Stack Overflow Podcast - Anil Dash - Transcript
00:00 Ben Popper: Alright, everybody. Today's episode is brought to you by Refinitiv. Unlock new possibilities with consistent high value market data from Refinitiv. Try the Refinitiv Eikon Data API for the largest breadth and depth of data and community tools with native Python support. Check out refinitiv.comcom/stackpodcast to try the icon data API today. Refinitiv, data is just the beginning.
00:26 Sara Chipps: I was going to say you've rebuilt geo-cities.
00:27 Anil Dash: Yeah, yeah, yeah. I wish. I hope. I mean, and I think like for those of us who are old enough to remember geo-cities or MySpace, it was like millions of people used to just make cool stuff on the web.
00:37 SC: Yeah.
00:37 AD: And it wasn't like we spent all day hitting refresh on like three different websites that made us miserable.
00:42 [INTRO SONG]
00:51 BP: Hello everybody and welcome to The Stack Overflow podcast for the week of February... Who knows? Post Valentine's, hope it was happy for you. Hope you got a telegram, a Postmate, a paperless post.
01:05 AD: A Valentine?
01:07 BP: Yeah.
01:08 AD: Could it be a paper Valentine?
01:08 BP: It could be a paper Valentine.
01:09 SC: Yeah...
01:10 BP: We have a wonderful guest on today, Anil Dash, the CEO of Glitch. Hey Anil.
01:14 AD: Hi!
01:16 BP: Nice to have you here.
01:16 SC: Welcome!
01:16 AD: Thanks. I was here in the olden days, back when this podcast was in black and white.
01:21 SC & BP: Wow!
01:23 BP: That's right, when it was on the radio.
01:24 AD: Yeah.
01:25 BP: It was a live radio show. And you're also here today for the board meeting cause you're on the board of Stack Overflow.
01:30 AD: It's true. Both things.
01:31 BP: Yes.
01:32 SC: Is the board meeting here?
01:32 AD: It is. It is.
01:34 BP: It's going to be in the same room.
01:34 AD: Well, what's funny is at Stack Overflow HQ, the podcast studio and the boardroom are the same room.
01:40 SC: Wild!
01:41 AD: And I think I'm probably, other than Joel Spolsky, the only person who regularly was in both.
01:46 BP: Right.
01:47 SC: Yeah, that makes sense.
01:49 AD: It's weird.
01:49 BP: You have a special connection to the company. You also are the CEO of Glitch, which kind of was the Phoenix that arose from the ashes of Fog Creek. Is that a nice way of saying it?
01:58 AD: Yeah, well it was to reboot.
01:59 BP: Yeah. Reboot.
02:00 AD: Yeah. It's a new generation. Grittier reboot.
02:03 BP: [Laughs]
02:04 SC: That's great.
02:06 BP: It's like one of the middle, fast and furious.
02:08 AD: Yeah, it's actually, it's less gritty. It's like if we went back to like the cartoony Batman
02:11 BP: Uh huh.
02:12 AD: And he's like, yeah, my parents are dead but I'm a little happier.
02:14 BP: Right.
02:15 SC: Yeah.
02:17 AD: Colorful, bright colors.
02:18 BP: Yeah. Okay. Got it.
02:18 SC: I just punched people.
02:20 AD: Yeah, yeah. I put it like more Robin. More like jokes.
02:23 SC: Yeah. Yeah.
02:24 BP: So tell us a little bit about Glitch. I know you guys have hit some nice milestones, 5 million sort of apps created, which is pretty cool.
02:31 SC: Amazing!
02:30 AD: It's wild. Yeah. So we like to say Glitch is the friendly community where you build the app of your dreams and it's, and it's basically where everybody can make the web. So you go in the browser from a coding standpoint, it's like almost like Google docs for writing code. You just go in there and you can type it as you type your app is instantly deployed and it's really fun because you can send that link and be like coding along with someone if you're pairing or something like that. And then if, you know, whether you're a coder or not, you can go and browse. There's like 5 million apps and find lots of fun, interesting stuff and it feels a little bit like... it's open-ended and surprising like the good parts of YouTube where you're just like, wow, somebody made a whole documentary about this one thing they're obsessed about.
03:06 AD: And like one of the apps that went viral on Glitch the other day was from somebody who's a fan of BTS...
03:11 SC: Yeah!
03:11 AD: The stars of K-pop. And they used Spotify as API and they built a system to track how each of the songs in the entirety of BTS catalog is doing in terms of plays on Spotify. And it's just like see which songs are most popular among the fandom and what's getting played. And it's just like apps is a mode of expression for K-pop fandom was like very exciting to me.
03:32 SC: Yeah!
03:32 AD: I was like that is it. Because back in the olden days people would make a fan page on MySpace or something.
03:37 SC: I was going to say you've rebuilt geo-cities.
03:39 AD: Yeah, yeah, yeah. I wish. I hope. I mean, and I think like for those of us who are old enough to remember geo-cities or MySpace, it was like millions of people used to just make cool stuff on the web.
03:48 SC: Yeah.
03:48 AD: And it wasn't like we spent all day hitting refresh on like three different websites that made us miserable. It was just fun things. And so there's a lot of that spirit and then it's cool cause like alongside that there'll be, somebody made a really nice integration of a Slack bot that just like reports on stats in their companies. So you can do both the fun pop culture thing and that really practical thing for work. And it's just a set of tools that let you do that. And it's amazing to watch.
04:12 BP: There was an app that went viral, you'll remember the name of it for me, but basically you could make a doodle and then it would sort of come to life and wander around this area with the other doodles.
04:19 SC: Oh I loved that one.
04:21 AD: Yeah, it was called doodle something and I forget the name, but it was incredible because it was using this sort of generative animation framework to add animation to what you had drawn and it was just like one of those where I realize like at a tech level. I was like, wow, what you're doing with Web GL and SVG and these things is like, I was like, I didn't know canvas could do this.
04:38 SC: Yeah!
04:38 AD: So yeah, tech level is amazing. And then you sort of switch off that part of your brain. You just look at it, you're like, this is fun. And these are lots of people just goofing off and sharing their art with each other. And it feels like I said like the good parts of the internet used to, which is like these are people playing.
04:51 BP: Yeah.
04:52 AD: You know, with strangers, yeah, some of the drawings are butts and like it is...
04:55 BP: Right. That was the fun part, I showed it to my kids and they immediately got it and they did doodles to get in there and then their doodles started wandering into these huge penises and then my wife was like, why is it only penises? She had to draw her own. All those we kind of added to the chaos. But yeah.
05:09 AD: Ah, thank you for bringing your art to Glitch.
05:11 SC: You're welcome.
05:11 AD: Yeah, exactly. [All laugh] But I think that's exactly it, right? It's like that sense of like a playful space and it's not, you know, I don't want to over-index on like how much of this is purines content, but like a lot of it is just where is the space that you go and play and do goofy things with friends or kids or whatever.
05:27 BP: TikTok.
05:28 AD: Um, yeah. Well, and, and you know, TikTok I think is great and it really generative. I used to love Vine back in the day and now one of the creators has launched Bites, which is sort of a modern version of that. Those are great, but they're video and I'm like, my medium is the web.
05:39 SC & BP: Yeah.
05:40 SC: Yeah, my makeup is not always ready for TikTok.
05:42 AD: Right? Yeah. Like I am not presentable at all times, but like I can, I can type some tags.
05:48 BP: I mean, I do think it is interesting that if you work hard at it and you sort of establish a culture from the beginning and a vibe, then you know that becomes ingrained awhile. And like you said, I feel on Glitch, it's playful, it's creative, it's inclusive. And the same thing I think to a certain degree is true of TikTok. It doesn't feel like it's a place to go and argue about politics, you know, or stuff like that.
06:09 AD: No, yeah.
06:09 AD: Given the ownership, I don't know if they...
06:11 SC: Yeah yeah, they don't like, the [inaudible]
06:14 BP: But yeah, no I think you know, and that can kind of, you know, become this flywheel, which is a nice thing as opposed to some of us...
06:19 AD: Well yeah people say I'm going to go to this website or I'm going to load this app and I, I know it's not going to make me miserable. But that goes a long way these days.
06:25 SC: Yeah.
06:27 AD: Like the bar is that low. And then you know, in the case of Glitch, so you can talk about the like setting that tone. Some of it, it's the DNA, Glitch used to be Fog Creek.
That is a company where Stack Overflow was born, where Trello was born. There is this 20 year history of like let's give people tools to share, collaborate, communicate that it's not going to do weird, creepy things behind your back with your data. Like all that kind of stuff was, you know, I get to walk into a place where that had been set 15 years prior that that goes a really long way.
06:54 BP: So tell me a little bit about what the underlying sort of technology is. You mentioned canvas. If I don't know anything about code and I show up, what tools can I use? And if I do know stuff about code, what code?
07:04 AD: So Glitch is an architectural marvel, technically it's kind of amazing. So each of those 5 million apps, and we call them apps and yet it can be websites, they can be bots, they can be whatever. But each of the 5 million apps on glitch is its own full stack app running in its own Docker container. That's instantly gets spun up as soon as you remix an app or create a new project on Glitch. And just the orchestration of that. It's kind of mind boggling.
07:27 SC: Wow that is kind of mind blowing.
07:28 AD: And like you think about people managing big Docker fleets and things, they're probably all like the same where you have like 10,000 workers that are all like a database or whatever. And these are like 5 million apps and no two are the same.
07:38 SC: Are you all Kubernetes?
07:39 AD: We are not.
07:40 SC: Crazy!
07:40 AD: It's really mind bending. It's definitely like I'm like, we are fortunate that we have a lot of very, very clever smart people working very hard because Kubernetes, incredible infrastructure optimized for you know all these containers are the same. And also you keep them running. Right? In Glitch apps if you're not using them, they go to sleep after five minutes and then we restore them when you go back to the app. And that is not a workflow that any infrastructure optimizes for cause it's very weird behavior. Nobody else wants their database to go to sleep when you're not looking at it. Right.
08:09 SC: Yeah!
08:09 AD: And so there's a very, very interesting set of constraints we have. And so the team spent, I mean it was really the first year, year and a half of working on Glitch was kind of deep work on like how do you solve this class of problems as a different optimization anybody else is doing.
08:24 BP: Yeah.
08:24 SC: Do you charge folks?
08:25 AD: We're just about to, so that's least surprising reveal. But like guess what we'd like to make money and still exist!
08:32 SC: What!
08:32 BP & SC & AD: [inaudible]
08:35 AD: Well it actually, you know again like it goes to the DNA of the company is like, you know I love Stack Teams and we use it every day but like Stack Overflow Teams or Trello like there were these products where it's like in exchange for providing you a product or a tool of value, you will give us money and everybody understands how it works. And I'm like, I don't understand how that became old fashioned.
08:54 SC: It is old fashioned.
08:55 AD: Like everybody like else has this very elaborate complicated and even like there are companies like that, like I love Slack and GitHub and whatever where it's like I'm going to give you money and you're going to give you these tools. And it's all good.
09:04 SC:Yeah.
09:04 AD:It doesn't have to be complicated. And I think because Glitch is a little more social and seen kind of a little bit like a social network, they're like, are you gonna slap ads on and do weird creepy stuff on therr data. It's like, no, no, I don't want to do any of that. Like take lunch to work, make cool apps with it. We'll charge you a couple bucks and then your app will go to sleep all the time.
09:20 SC: So you're going to do that?
09:21 AD: Yeah.
09:21 SC: Oh that's fascinating. I was thinking that what you're doing is kind of like the Roku model. Early Roku. They used to take stuff down when you weren't using it. So if I pay then that won't happen.
09:30 AD: Yeah. Your app keeps running and maybe you have some more storage and compute and all that stuff. And then, I mean honestly what people want to say, like I want to, my it guy says I have to be able to sign in with Google or something. Right. Okay, fine. We'll give you that. Sure.
09:42 BP: Yeah. Yeah. I was reading a really fun article about this ML train and it's called an AI Dungeon or whatever.
09:47 AD: Mhm.
09:48 BP: And so, you know, it's that word one where it's like you walk into a cave, what do you want to do? And they've gotten it to a point where you can say literally anything and it'll play with you. But they were saying eventually, you know, this is a university project. The cost got too high.
10:01 AD: Yes.
10:01 BP: So what they had to do was come up with a system like yours where it's like, okay, if you haven't played in a day or so, we're going to shut you. You know, we're going to close you.
10:08 AD: You close your session.
10:08 BP: Yeah, exactly. So kind of toggling them on and off while they're at that moment of saying like, we want this to be free and open and have as many people as possible, but we can't support all these instances, you know, or you know, 24/7.
10:19 AD: And that stuff seems like it seems pretty fair. Like I think more of these models, talk about Slack. I mean freemium has been around a long time where it's like, here's a business model where try some, if you like it, you pay some money. And everybody is square and I just wish more of the web ecosystem work that way instead of the like some weird creepy thing is happening behind your back and that's the monetization model. I think it's nice to be able to understand things and how they work and then you know, to that point about the AI, it's so fascinating because text adventure games have been around for 40, 50 years on computers and it was always people writing scripts and I'm like, yeah, this is a Corpus of texts now it's like being able to train up, you know, an AI on that is like a dream. Like it's really exciting and we have this huge community on Glitch of people doing work, especially in TensorFlow, but all the different AI and ML frameworks and the TensorFlow team at Google, they use Glitch for all their examples. So it TensorFlow, JS is all...
11:08 SC: Cool!
11:08 AD: Yeah, I was definitely, which I think a lot of people are. There's always the like I ought to learn that. Like I definitely have that.
11:14 SC: Oh yeah, TensorFlow is on that list for me.
11:15 AD: Yeah, I mean I think, I think it's like, whether it's like this is the year I'm going to look finally learn View because I'm a react person, you know, like whatever it is. You're like, this is that framework that you feel that I always felt like guilty of I haven't like kicked the tires on this yet.
11:26 SC: Yeah.
11:27 AD: And it was always, what am I going to have a couple hours to set up a dev environment and pull down the code and to build it and blah guys like a nightmare. Right. It's like a child that I would be neglecting. Like it's a lot, you know.
11:39 SC: Dad setting up another dev environment.
11:40 AD: Yeah, exactly.
11:41 BP: You gotta you gotta do it with them. You're gonna be like, Hey, let's learn this together.
11:43 AD: Well he's more of a Python guy. And I'm like, Oh, I don't want to be like a nine year old that knows more than I do. I was like, I don't want to compete.
11:50 BP: Yeah.
11:50 AD: But then I could go to in one of these TensorFlow examples on Glitch and was just running and then I was just playing with it. It was already trained. It was already there. I could just go and it was like I could do it as a coffee break or I could do it like over lunch or I could do it after the kid had gone to bed and instead of like was literally like while watching Netflix, I can be tinkering at this thing. And it felt like the fun part, the exploration part, it really brought me back to like why I love coding, why I love creating, I love messing with stuff on the web and being able to pull apart real AI apps that people would build Glitch and be able to just view source on them and see what they did. And I was like, this is the joy of coding. This is the part I loved.
12:24 BP: That's the power of like owning a really great synthesizer, is that there's all this stuff built into it and you can hit a few buttons and all of a sudden it's playing this amazing thing and then you can sort of tinker in there and then deconstruct it a little bit. But yeah, you don't even have to know how to play yourself.
12:37 AD: Right, right. Yeah. The presets and the things that are there. I was always a person that loved like the DVD extras as much as the movie, like how did you make it? Like I could not get enough of footage of like somebody in a blue suit with dots stuck on them in front of a blue screen getting me that. I was like, yes, like show me more of this and that thing. I think it was sort of that same impulse as like how did you do that? How'd you make that? Or you know, growing up on like being a Star Wars fan, I was like how did they do the stop motion and I think it's the same impulse which is like that app is amazing. I want to see what you did.
13:05 SC: Yeah, like open source and learning. One thing it sounds like, so Paul and I have a segment scheduled to complain about package managers as long as possible.
13:13 AD: Oh, okay.
13:14 SC: It sounds like what you've done is circumvented the package manager. Like the whole is not compatible. You do not have Virgin 2.9 installed. Now let's start over.
13:24 AD: No, it's so it's kind of amazing on Glitch we detect your package, and we have basically cached out almost every common node module.
13:32 SC: Yeah!
13:32 AD: So like if you just sort of added dependency into your project, it's just there. There's no bill, there's no nothing. And it's shared across these 5 billion projects. So like essentially every node module, it's used across 5 million apps like it. At some point, everybody's used to everything. And so once they're cache they're cache for the entire infrastructure. And so that was a huge part of it was like there were all these like image libraries where I was like, I remember spending hours being like, this won't build. And I had like a stock standard Mac, like I'm not doing anything weird. I'm going into Homebrew and I'm like banging my head against the wall. And I'm like, but I just wanted to make the picture, you know, like I wanted to make funny, cool things.
14:05 SC: Every time.
14:06 AD: And it was like all of them. It was like, yeah, if I were like baking cookies and I was like start by grinding flour. No, I want the sweet delicious treat. And so yeah, that, that was a huge thing for me that felt really empowering in the early days of Glitch was, and I take it for granted now it's funny cause we, we use it to build Glitch, right? But I don't think about dependencies.
14:24 SC: Yeah!
14:24 AD: I don't, I don't think about package management. Like it just happens magically in the background. And in fact like we have a little thing that'll sort of pop up, Hey this one's out of date. Do you want to see it? And I think about like, you know, how do we even help you understand what changes are going to happen at your app if you upgrade upgraded dependencies. That's definitely a path we want to go down.
14:39 SC: Yeah. I had that happen to me recently where I wanted to fix a bug on an app that'll go unnamed and three days later I had it installed and I was like, well now I don't have time.
14:48 AD: Yeah, well and it's also like you only have so much energy in your soul, right? And like your soul is [inaudible] You're just like, well, you know what, I don't like, I don't have it in me anymore.
15:00 BP: Right. You're the character in the question. You've got your sleep meter.
15:04 AD: Yes. Yeah.
15:05 BP: Your my soul is crushed meter.
15:06 AD: Yeah, exactly. Right. And I have spent all my soul points on builders.
15:11 BP: And so along with Glitch, you now have a publication called Glimmer. Tell us a little bit about what the impetus behind that was. Where people can find it and sort of, yeah. What you're trying to do with it.
15:20 AD: You know, the high level is, I think all of us who are in tech, well you've heard me complain about, Oh well, you know, these sites are doing creepy things and these networks are full of content I hate. And yet there's still an optimism. I think for a lot of us who make tech that is like what was supposed to be good for us and the web was supposed to be good at tech, was should be good. And we all know those stories. We all know moments where we had a success. I made a thing and it was cool and somebody else liked it and it felt good. Or I saw this app and it reminded me that people were out there being creative and there wasn't a place to tell those stories about, you know, we can still be critical. We can still say apps shouldn't be doing bad things with your data or whatever. But then to say, and where do we point out what somebody created that is inspiring, where somebody who's unexpected has been able to make technology where technology is shaping the world around us.
16:04 AD: You know, every, every aspect, right? So like the idea of a way of expressing your fandom for your favorite boy band is software... is new, right? People are like, yeah, you make a poster. Yes, you get a tattoo. Yes you have a poster on your wall or you wear a tee shirt. But the idea of like, I'm going to express my excitement about this thing by making technology is new. And so Glimmer is a place to tell the stories. And I've been really, it's brand brand new, so we're just getting launched. And the first wave of stories are really thoughtful, I think in that balance of you don't, yeah, we can do better. Some of this stuff that we're gonna point out at what's broken in tech and can be implemented better. But also I think any of us who have ever written a line of code, read a regular news story about tech and kind of roll your eyes, they're like, they don't understand this.
16:47 SC: Yeah.
16:49 AD: They don't get, I mean, they understand the social issues of the business issues. They're really great at writing about, you know, this company raised this many dollars from this investor because they know finance.
16:58 SC: It's a number and it makes sense.
16:59 AD: Yeah. But then you're like, no, no, no. You've misunderstood what happened technologically, you know, we have the recent debacle in Iowa at the caucuses. That data happening there and you know, there's a lot of cultural and social underpinnings as to why they ended up in that situation. But at a tech level, it was a couple days later, somebody wrote a story, I think it was VICE magazine had written a story. And then Buzzfeed or somebody did a kind of a recap of it, but they mentioned the app was distributed through TestFlight, right? This mobile app was distributed through test flight and all of us who've ever shipped an iOS app are like, Oh my God. [inaudible] like that would never work, in the field, like low connectivity. And they're sending around, they're like device ID. Like, no, that's never gonna work. And so there's the thing that we have that reaction because we know what that means, where you're like, somebody was down to the wire in X code fighting with iTunes connect and just being like, Oh please, Oh please work.
17:50 SC: We're imagining that person with the sweat.
17:52 AD: Exactly!
17:52 SC: I guess we're just going to have to do tests like, Oh my goodness. Not ideal.
17:57 AD: And the funny thing is people who've never coded, like I think about a cooking show and they have watched the like down to the wire and the person's like, ah, I couldn't plate the dish. I didn't have the sauce ready. Or you know, whatever the competition show is. And they're like, I don't cook. But I can see, Oh, I understand what that person went through. And that it wasn't done. And they can understand that panic. And actually if it had been they were trying to serve a Bearnaise sauce to everybody in Iowa, they would actually be able to, the writer would be able to talk about that, you know, like the souffle fell and they weren't able to get it done and therefore it didn't get onto the table. But if they don't have that fluency knowing, Oh this is an indication you weren't ready to ship, they can't tell that part of the story. And that is key. That level of technical knowledge of being what only coders know is key to actually telling the story in this case about what happened in politics but is true in art. That is true in education. That is true in every aspect of society. And you know, this is some of this is what we want to do in Glimmer, but I think just across the board and you know Ben, like how you and I met is the, you telling these kinds of stories. And I think going back to like how many people in the world are telling stories that involve technology from a point of view that is fluent in technology.
19:03 SC: It's a tough thing. I was, I was talking to someone the other day, a relative actually about their work and the developers at their office and people were frustrated because they have an app and they have one of those pop ups that say, how do you feel about our app scale of one to five. And they were frustrated because if you answered three or below, it sends you to a feedback form that goes nowhere.
19:26 AD: [laughs] That'll show 'em.
19:26 SC: And then if it's like above three it goes to the app store. So you could do it there. And the developers were frustrated because they felt it was unethical, but the business was like, come on, we want good ratings for our app. So that's one of those situations where like you're building the tech, you feel like you have a good ethical stance, but it doesn't resonate with the people driving the business.
19:48 BP: Yeah that's a dark pattern I'm going to have put in my pocket. [all laugh]
19:53 SC: We should use that for the podcast.
19:54 AD: All of us have encountered those things where somebody made a choice and we don't actually have that many places to talk about that stuff inside the industry. Right. So they go as folklore and almost all of us have probably been at like companies or worked on products in the past where there were things like that. And where would you go? Because it's not going to get to the level of outrage where it's going to be a headline in the New York Times.
20:16 SC: Yeah, no ones reading that article.
20:17 AD: You know inside the industry there is a culture of choices in our technology. We make choices about where the app goes, where the form survey results go. Well you know how we encourage certain behaviors and we don't talk about them and we're like what are the incentives? And so I think those things are very, they're really interesting and that they're, you know, you move a button by a couple pixels and has this huge impact on maybe millions of people's lives and yet it doesn't get discussed. And so I think about, I mean that's actually a bad example cause UX at least there's some conversation cause people can sort of see it. You know, if I have Apple changes to dark mode, everybody talks about that. But the choices about like how does this software work and who can access it and what does it do to the people that use it and with their information.
21:01 BP:Yeah. Well it's like the unseen things. What is being surfaced in your feed and what's not and why, you know, or like what is trending for you, you know? And those choices are completely opaque to you. And when you wake up every morning they shape your mentality for the whole day going forward.
21:14 AD: Yeah. Yeah.
21:16 BP: So we have been doing a lot of work here at Stack Overflow, you know, on the community side. I would like to get your take on it as somebody who, you know, who's been involved with the company for a long time. You were saying before about how sort of you go through that transition of, Hey, we've been providing something for quite a while for free and now we'd like you to pay for it or whatever. I do think there's a deep sort of held sense within our power user community of, Hey, you're getting away from your roots as a community created thing and now you're just all about business. It's about the money. It's all about the suits. It's all about the lawyers, you know, to which the response is, well, we have many, many employees to pay and we have changes that we'd like to make to the product that would, you know, you've asked for, how would you like us to do that? Should we go to the Jimmy Wales Wikipedia donation model or you know what? Well, you know, the whole idea was to get away from the experts to exchange, right? To make it all free. You don't even need to log in. You know, we don't even need a user account. So I wonder what you think about that challenge, especially with 12 years of history on it?
22:12 AD: Yeah. So, you know, I'm lucky in that I kind of got a front row seat to watching all this and yeah, I should be clear. I'm, you know, I'm on the board of Stack Overflow, but I'm not speaking as like here's a corporate spokesman and kind of like, this is literally just me. Well exactly what I was, I was somebody who knew Joel Spolsky and Jeff Atwood and read their blogs. I was a working coder. I thought I could never pass a coding test to work for those guys cause I'm not that good a coder. But there's a lot to learn from what they write. And this is something where I don't think that, well the majority of coders won't remember it because it's been so long now. But you mentioned experts exchange. There were a series of sites that had some answers to some coding questions and they were all pay-walled. And it was weird because as the early era of SEO and Google and so you knew the answer was there because it would show up in search results when you would get there and immediately lock you out and be like, you got to pay. And it got more and more expensive. I mean, it was literally one of the things where you're like, the answer to the question that I need to do for my job is being held for ransom and I can't expense it because there was so much stigma too, which was like if you tell your boss you're going to expense expert exchange...
23:14 SC: Yeah, why don't you know?
23:15 AD: Right, exactly. What did we pay you for? Right. And we bought you two O'Reilly books. Like shouldn't you know everything, right?
23:18 SC: Right, they were sixty dollars.
23:21 AD: Yeah, exactly. And, and um, and it's very hard to sort of, to picture that world. And then there would be like, I was a mostly PHP coder for most of that part of my career and there would be a couple of mailing lists. And you know, you had the, what's the nice way to say rules Nazis? Like you hit the most strict.
23:39 SC & BP: Rules lawyers.
23:41 AD: Yes, exactly. Rules lawyers. And um, sorry I'm an old man, but you would have these, these, they were so rigid about, well you top quoted your email, therefore you don't get to know the answer to that question about PHP and I would just be like, all right fellows, like I just wanted to know like I'm sorry.
23:59 SC: I wanna leave at five!
23:59 AD: Exactly. Like I'm sorry I have transgressed against you with my plain-asky email here, but like can we just talk about what the code is that we're writing? And it was, at an emotional level, it was humiliating.
24:12 SC: Yeah...
24:12 AD: It was degrading because it was like you have to perform in these ways or you have to have like unlimited disposable income to pay for three different pay-walled sites. And the overall effect was this is a secret priesthood and you can't be in it. You are not allowed to have access to this information. And I had this experience which was like child of immigrants. I didn't go to college, I just wanted to like have... I loved the web and I love tech, but I wanted to be able to build stuff. And it was like if you don't do these things and these rituals you've never heard of in this way, then you don't get access to the information that is like economic opportunity and career advancement and self-expression. And I didn't know until I saw another way how much I resented it and how hurtful it felt and how exclusionary it felt. And this is like, I mean this sounds like I'm like kissing ass, but like, and then I saw what Joel said and Joel was like, there's a moral imperative and I'm paraphrasing because like there was a moral imperative to not lock this information away.
25:11 AD: This isn't a path to access for. And now I look at this today, there are millions of people around the world who their path to economic stability, opportunity is access to the information that is in the Stack Overflow community. Right. If they can see how to answer that question, how to write that line of code, they can transform their lives and transform their careers. And you know, and the kids, like my family is from a very rural part of India and Aboriginal tribes live around our family village live on about seven or $800 a year for a family of four. And kids there go and go to the like bigger cities and take a class on learning Java or learning, you know, like whatever the framework is. And they can permanently upgrade their entire villages standard of living by learning these skills that the Stack community has been generous enough to share with them. That is the center of what Stack Overflow is about. You know, and it's funny because at the same time, like, you know, I know Jeff Atwood a long time and to the point where, you know, he and I have butted heads many times. He's not the, he cares deeply about building platforms that people can share knowledge on. And he's super opinionated in a way that I always found really obnoxious. Right. And so I was like, and Joel is no shrinking violet either, right?
26:30 BP: You're pretty opinionated on Twitter.
26:33 AD: Oh no, no. No I haven't. Yeah, I've never. Yeah, alright. And, and so you have these very strong personalities and it is both the yin and yang of that, I think is what the catalyst is for Stack Overflow, which is like the impulse, the underlying motivation is one of the most generous acts of knowledge sharing in history.
26:49 SC: Yeah.
26:50 AD: It is akin to Wikipedia. It is one of these things, which is like we're going to throw open the gates to this information and not have this. Yeah. And that, that is powerful. And then you say, what are the choices you have to make about incentivizing rewarding people to participate? I think people come to Stack out of a sense of, you know, when they're answering out of a sense of altruism. I think I really do want to help people. I think for when you're asking a question, you're looking at that information, there's a vulnerability to it. I don't know this, I don't want to be seen as a fraud. I don't want my boss to say, why don't you know this? I don't want to have my entire career be precariously in the balance about whether I could actually get access to a certain kind of information. There's a vulnerability and I think that's one of the biggest balances over the last dozen years is we're balancing people at a moment of maximum vulnerability about their future, their career, their life, and people who are genuinely altruistic but playing a game that was made by a strong headed people who wanted to reward a certain competitiveness and a certain, you know...
27:50 BP: Girls not sand or whatever.
27:52 AD: Yeah, exactly. Right. And it wasn't a bad motivation, but Stack Overflow historically has been a single player game.
27:58 BP: Yeah.
27:58 AD: There is a global leaderboard, but it's been a single player game and I think we're in the transition to making it over, you know, years and decades to come. A multiplayer game where we advanced by helping each other. That was always the impulse, but that wasn't the score keeping that wasn't the rep. Right, right. And I think it's a hard transition and I think it's no surprise that people feel stress and confusion and anxiety about that. I think we have to be, and Prashanth and I have talked about this a lot. Joel and I talked about this a lot. We want to figure out how to navigate that, but that thing makes me so excited about the future of stack because there's this chance to rethink the fundamental principles. What brought us all here in the first place.
28:38 BP: Awesome.
28:39 [OUTRO MUSIC]
28:42 BP: Alright, we got to wrap it up. I'm going to read two shout outs, the lifeboats and then we'll say goodbye and where people can find us on the internet. Good zookeeper. Hello world program with Java client, awarded. Yeah, to Depok Singhal. So thanks for answering that question. I don't really know what the question is, but what does the return value of GC.collect actually mean to Tim Peters yesterday? Thanks for those lifeboats to both of you. I'm Ben Hopper, the Director of Content here at Stack Overflow. You can find me @BenHopper on Twitter.
29:09 SC: I'm Sarah J. Chipps, @SaraJChipps on Twitter, and I'm the Director of Public Q&A here at Stack Overflow.
29:14 AD: And I'm Anil Dash with Glitch, find me at Glitch.com and thanks for having me.
29:19 BP: Yeah, thanks for coming on!
29:20 [OUTRO MUSIC]