 So welcome everyone to the Civic Lunch. We're going to get started. Those of you who are sitting way in the back there, you know, move up a little bit. There's very closer chairs. You want to be able to see better. I'm pretty excited to introduce Evan Henshaw-Clath, also known as Rabble, also known as Anarco Geek. So Evan, let's see, was created protest.net back in the 90s, mid to late 90s. An idea that was instrumental in the development of the Indian Media Network, the Indian Media Centers, went on to work at Odio, a podcasting company that was before its time, and was then involved in the Genesis of Twitter, and we're going to hear about that today. Evan later went on to do a number of things, including starting cooperative development shops in Uruguay, working at CTO for NIO, one of Joey's companies, and a number of other things. When I first met Evan was in Cancun in 2003, and unfortunately we weren't relaxing on the beach in Cancun, that would have been nice. We were there as part of a protest against the World Trade Organization meeting which was in a sort of walled off hotel zone that's not accessible to the people who live in Cancun, and we were setting up an Indian Media Center to produce people's coverage of the trade talks and mobilization. I remember sort of being in the media center where we were setting up computers and networking them, and organizing things for the action, and I was having trouble getting one of the computers network to set up, and I said does anyone know, can anyone help me with this? And this guy was sort of under another computer working on something, and he kind of looked up and he said, well, is it, what's it running? And I said, oh I don't know, it looks like it's running Windows, and he said well there's your problem. And then he came up with a procedure to install Linux and connect it to the network, and then we had another active machine for the Indian Media reporters for the action. So I'm really excited to hear Evan's talk today, and click the way. Yeah, so the funny thing is the thing I've been doing right before that was attempting to smuggle carloads of computers to the Zapatistas and failing to sneak them across the border of Mexico. Because you were too honest to bribe the official who was trying to get you to keep it from the bribe? Yeah, I couldn't figure out exactly how much you bribe. We even went into the bathroom and now you give me the bribe and I just, I've gotten better at bribing Latin American officials to smuggle shit. But that is not what this talk is about. Turns out if you want to smuggle radio transmitters to India you put them inside desktop computers, and then they think it's a normal computer. So this talk is about Twitter. One of the things I find fascinating is that the origin story of how something comes about and how it gets created isn't the story that's useful for companies when they create their own narratives. Particularly for me it's fascinating to be here at the Media Lab because Twitter's origins really come in large part from the Media Lab, but most people at the Media Lab don't even know it because I'm not sure that it was a project that people, like it wasn't the normal commercialization route and it wasn't entirely like the project that Tad was supposed to be working on. And so that history got lost, but hopefully in telling some of the history of how Twitter got created and where it came from, I can revive some of that history. So, as Sasha said, from 2000 to 2004 I was traveling around the world setting up activist media centers and trying to build ways in which we can build technology for all the media stuff. And by 2004 I had run out of money selling my first company and got a job, started doing contracting in San Francisco and there were some folks in San Francisco at the time, there was no big dot com boom, there was no job and everything else, so it was easy to find interesting people. And I ran into a guy who said, I want to reinvent radio, a guy named Noah Black. We're going to democratize radio. And I'm like, well I've been doing media activism, democratizing radio sounds really fun. So, we started working cafes, we started hacking on stuff and we created this thing called Audio. It was sort of a standard startup, we got a little bit of angel funding and we worked really really hard. And we just kept working and working, you know the sort of like moment which you never have been doing the work. And then we launched it and we launched it in a really traditional startup way, which is at Tad X before there were any Tad X's and in the neary times of the same day. We showed it to no one before this, we've been still, we got no feedback so it looked really pretty and everyone said this is great and cool and the interface was really nice but it lacked some serious user experience stuff beyond the visual interface. And so, not realizing that our problem was user experience, we just kept adding features. We built a tool for recording and editing podcasts in the browser using Flash and we just kept working. And then we sort of got hit with Apple integrated podcasting in an ugly way in the iTunes but had really good user experience. It's sort of the best example of ugly good UX. They set everything up in nice lines and they set everything up so that you can get the history and you can pause it and you can be started and none of it was pretty and you couldn't find which podcast you wanted to listen to but it really worked well. And we had gone down to meet with Apple and shown them everything and they were like, oh great, that's neat. Podcasting is neat. We're not going to work with you guys and then they launched it really quickly. And so, we had this kind of oh shit moment fuck. What do we do now? At this point we've taken about $5 million in venture funding. We had a team of a dozen people we had an office, we weren't working in cafes anymore but the product was never going to be like that big home run. And so, we ended up in this light of like startup zombie purgatory in space where we had $7 million in the bank and we had a great team with interesting people and stuff but no viable path to like, you know, being what it's called the unicorn now. And so, you know this is one of our investors and they like everything was perfect. Like everything that should have been lined up and what we were doing was lined up. And yet, we couldn't like get that hockey stick of massive growth. Besides it was profitable. Like it made, it worked. It had growing traffic but it was never going to be that turret thing. So what we decided to do was sort of go back and say well rather than shut down the opposite way everyone off let's start doing hackathons. I started this thing back in 2005 in 2006. I was very active in the Ruby on Rails community and I said with some friends let's do a 24 hour hackathon to show that you can build applications in 24 hours. It ended up converting into something called Rails Rumble which is two days now and people consistently build applications in two days. But you know, ten years ago the idea of building an application in a day or two that was a viable web application was a crazy idea. So we started, we organized this 50 person hackathon remote and we did it in the office and other people in the office were like wow, you know, maybe we can build things not six months in stealth with a bigger team. Maybe we can build things really small and quick. So what we did is we set aside every Wednesday in the office to not work on ODM, not work on the thing we were supposed to work on. And we did a brainstorm in the morning from like 8-9am and we said all the teams have to be cross functional. So you can never work with someone who does the same thing as you. You're designing, you can't work with other designers, you're a product person, you know, we're gonna mix up the teams and there wasn't a like order of it, it was just like pick someone you don't normally work with. And then we just all went to our office and we just hacked on stuff. And moved around and built stuff. Everything was like things in the back of our head. And then at 6pm we did a demo. And it was just like what's the fastest way we can get to trying things out. And so we built some things. We built a little app that you could post on MySpace, the social network of the time. What you would record videos of yourself through your webcam and post it online. And get a series of them. And the idea is you would post one every day. We added a social level on top of the podcasting. So at the time there hadn't been a social sort of graph level. And now you can have friends and you can sort of see the teeny icons that eventually became in Twitter. And at the time there was your name under it and things like that. One of the things we created was a collectivized voicemail. Where it would look at your address book. And when you called in you would record a message. And then you would hear all the messages of people who you have phoned in your address book with. And so it was this sort of almost like this. It was the way the Twitter feed works except it was recorded audio. Playing with asterisk. And then one of the things we built was this thing called Friend Stalker. And for a couple weeks Friend Stalker was the name we actually used internally. Before Noah, the guy with the original podcasting idea, sat in the office with the Oxford English Dictionary looking through trying to find a name for this service. And he got all the way to T before he discovered that Twitter actually kind of reflected what it did. And so that's how Twitter got created. That was the process by which we in San Francisco said, oh, we have this failing podcasting company we're interested in media. A bunch of the people on the team had worked on Blogger, Blogspot, Pyra at Google. And they're like, oh my god Google doesn't understand social software. They hate us. So part of the reason we did podcasting in the first place is because they commute from San Francisco down to Palo Alto for work. And so they had all this time sitting on the Google bus. And they were really bored and so they did podcasting. And then once we moved the office to San Francisco and everybody biked and walked to work, no one listened to podcast anymore and we stopped sort of dog-fooding our own products. And then all of a sudden we got bored and we built something else. So the particular hackathon project that made Twitter in this case was the one that did the thing that became FriendStalker. I went back in my emails and I think the same day we did three products that were indistinguishably similar. Like one was called Ketchup and one was called Linked Messages and one was called FriendStalker. It's one of those things where people start rebuilding the same stuff. Jack didn't call it FriendStalker. It's kind of out of it. Before Jack was just a developer on the team, before that he'd been a bicycle messenger in New York and he kind of got obsessed with those messages that they give on the radio of where they are and what they're doing. And we'd had this idea in the team, in the project that every time there's a media shift so every time the medium shifts from dumb phones to smart phones, from desktop applications to web applications, from static web applications to dynamic every time there's a major media shift, you know, what you can do is look at the social software, look at the social connections and the way people work with the previous generation of technology and you can try and think about what creating those modes of interactions would be with the new media. Because almost always the old technology can't transition to it. And so a lot of what we were looking at is like what would be the future of smart phones if smart phones came out? Remember if there was no iPhone. So we were using like, you know, Symbian Nokia phones. And then what in the past had we all worked on that made sense? And so through this, one is a, I'm an old computer, I'm a big geek, and in Linux when you have a shared system, which people still use sometimes you can have this .plan file in your home directory and there's this command called finger and you can finger people and it sends you back there .plan file, which I thought was really awesome. And on some of the systems I've used and some of the other people used this had become a social network. People would regularly update their .plan file that they're working on and then people would finger each other to see what was going on. And it said, oh, there's a way in which you can sort of have a distributed system when you do status updates. And one of the guys on the team had gone to Carlton in Minnesota and in Carlton they got so obsessed with the finger and .plan files that they had actually built a web version in the mid 90s of it. So you had a web interface to do it. And because a bunch of people had worked on blogging there was this idea of, you know, the feed reader, what if there was this flow of information that came to you and what if there was asymmetry about who you read versus who you spoke to. And what I believe sort of was the most important influence into it was this project that came out of Ted Hirsch who was here at the Media Lab called TextMob and it was a social network group messaging system only for protests. And he had worked on it over a while and perhaps would just show the video of it so you sort of get a sense of what TextMob was. Since 1998, the Institute for Applied Autonomy has developed technologies that enable average citizens and political protestors to functionally parently to defend their way of life. In the summer of 2004 IAA researchers produced TextMob, a cell phone text message broadcasting system. TextMob allows activists who are distributed throughout a city to remain organized during chaotic street protests. TextMob users are able to use their cell phones to send up to a minute text messages to hundreds and thousands of other TextMob users containing vital information regarding needed medical assistance, protest meeting points and police blockade locations. TextMob's utility was proven during the 2004 Republican National Convention in New York City when protestors were able to bypass preemptive police tactics by sharing information on undercover officers and police movements. Demonstrators use text messages to convert protest targets with little or no advanced warning, catching law enforcement off guard and temporarily seizing control of the city. Tipped off by TextMob messages, protestors are then able to disperse before police reinforcements arrive only to reconvene around a new target moments later. Such tactics often referred to as swarming by defense analysts at DARPA and the RAND Corporation are considered by security experts to be the most effective tools at the modern activist's disposal. TextMob is also an effective tool for safeguarding the integrity of Democratic elections. TextMob is used in the Orange Revolution when allegations of widespread motor fraud during Ukraine's presidential election mobilized hundreds of thousands of people into the streets of Kiev. TextMob was also used by demonstrators protesting the January 2005 inauguration of George W. Bush in Washington DC. TextMob, now more than ever. The Institute for Applied Autonomy would like to thank the following individuals and organizations without whose continued efforts this type of work would not be possible. So essentially the video was 2007-2008. It's a bunch of years later. The RNC is 2004. Yeah, so this happened 2004-2005 and then what happened was Tad was here but he was a really terrible program. Like horrendous program. Did everyone touch that Tad versus a Media Lab student at the time? He's at the University of Washington now. And we in San Francisco had found out that this TextMob thing was happening and it was really cool and he's like, I want to do this thing and we're like, Tad, come out to San Francisco and we'll spend a weekend hacking on stuff and we'll fix your software. So he flew out and there he is and we spent a weekend basically rewriting the back end of the software. And at the time SMS delivery was really tricky and so we're like, oh shit, SMS is expensive and we don't want to spend any money. So what we did is we built a dispatch server and we built a Java applet that we hid in a bunch of activist web pages at the bottom of the web page and the Java applet would then connect back to the distribution server and it would run SMTP and it would send through the public email gateways for all the delivery for all the SMS to email SMS gateways and because we did it through this sort of distributed network and we had it up on 100,000 web pages or something we didn't have to pay for anything. We could do about 40,000 messages an hour through this thing and it didn't get blocked because that was immediately what happened to Tad is before we had that for the DNC here in Boston a few thousand messages went through and then all the providers blocked it and so in the two weeks between the two we had to write this other software. Tad was part of the smart cities project and worked with a bunch of other people. He didn't actually talk to me much about what he did here. We were very focused on the code but it was one of like a dozen projects that he worked on. In sort of preparation for this talk I went and looked Tad and I both gave keynotes at the O'Reilly Emerging Telephony Conference in January 2006. So about a month or two before the first Twitter prototype got built and I talked about TextMob and I talked about a bunch of other activist projects and Tad didn't mention that he worked on it all. He talked about another system he built here for doing simultaneous translation for Chinese language speakers in Boston. So how to use TextMessages. Very similar stuff to the stuff Sacha still does about technology for community organizing things. Then after TextMob got a lot of attention at the moment at the RNC and then sort of disappeared. And then a while later years later, like four years later eventually there's a grand jury and they want to figure out who sent what messages and how people were using TextMob. Because they finally sort of come to realize that there was a lot of useful information analysis and figuring out who had done what. Because there were actual TextMessages about the burning of a giant paper dragon in the streets. So Tad had to deal with a bunch of legal fun. I think the data was gone by then. I mean four years later the chances of the servers still being around so I don't think they ever were using the data. And people were using disposable mobile phones and things like that. They go after this data a lot. And Twitter has this all the time. And before that with Indymedia we used to do jurisdiction hopping. So we would never put a server in the same country where it was covered. And we moved the domain name registration from country to country every like six months. Which means they had to restart another legal process in another country. Until eventually we put it in Brazil. And it turns out that Brazil's courts are really really slow. And there are lots of lefty Brazilian lawyers. And the Brazilians aren't very keen to follow the first world's demands for handing over activist information. And so domain name registration if you want to put it in Brazil it's a good place to do it. It's pretty cool. But so this is the first version of Twitter. The version that got built in that one day. It's really simple. You update status. You get a list of status of everyone. I think the very first version you had to default everybody following everybody. And then eventually there was sort of this friending mechanism that was created. When we launched it all of our friends thought it was really stupid. And we would never use this. Yeah I remember you sent me a thing and I was like what? Why? I'm like look try this out. This is awesome. Everybody's like why would I want to do that? Everybody. And the tech crunch and the press and our investors and everybody had about exactly the same reaction. This is the stupidest move of any startup that's happened in Silicon Valley. The investors Charles River Ventures their answer to this was fire everybody shut it down, delete the servers. There's still a couple million dollars in the bank. We can get 50% of our money back. This is worthless. A couple years later they go back and say oh shit Twitter's becoming successful and they bank their way back in so they can claim they were the original investors in Twitter. And so they ended up making a lot of money at it. But at the time their answer was kill this and kill it as fast as possible if you can't sell it to someone. We were lucky. One of the guys on the team who was the original angel investor Evan Williams. He was our angel investor and we worked out of his like old apartment for a bunch of months and then he had worked on blogger and sold it to Google and left Google because he decided they were antisocial and Chris Anderson from TED gave him like a phone call and said hey yeah what are you doing? And he didn't want to say I've been traveling around the world and taking cooking classes. So he said looked at the only thing he had he done one angel investment. He's like well this audio thing I invested in and I'm helping make this podcasting thing. And Chris from TED is like oh podcasting that sounds cool it sounds hip. And so after so instead of saying yes I'll give you the invite to attend TED which is all that wanted. They're like we'll launch it at TED. And so that's how we ended up launching a pitch. And after that he actually joined as an active person. He's really good at product as you know he made medium after that. So at the time Charles River Ventures is like shut it all down. This is an embarrassment and he's like well I'm going to take almost all my money and buy out the investors and keep going with this thing because I think it's interesting. And at the time Twitter had really really few users. The original homepage would show the most recent tweets and you could reload it and even six months a year in you'd like reload it and like there'd be a new tweet and there'd be like the other three tweets and then there'd be a new one. There was one time we were at a conference and there was a panel at the conference talking and Michael Arrington was sort of this asshole who works on TechCrunch You know you're being recorded. Yeah I know but I have no problem saying that Michael Arrington was an asshole. I don't think he had a problem either. So he had this thing where he was saying oh Twitter's not real it's not a big deal and they were fighting back and forth on the panel and so I started a live tweeting panel and then he brings up the Twitter homepage and of course because Twitter gets so little traffic at this point my live tweets giving him shit about giving the panel shit are the only things on them. So it didn't get a lot of traffic but the people who used it in the first month, the first three months there was only about 100 users but there were 100,000 messages delivered. So at the time you had to pay to receive messages but the people who used it that first 100 users were paying $300 or $400 a month to be able to use Twitter in SMS feeds and that showed you something really important which is you really want something that says love or hate relationship. You wanted people to say what the fuck is this? This is dumb or oh my god I'm going to pay $100 a month so I can get these messages and that was exactly how Twitter grew and so they had these super fans super hated it and everyone was really passionate about it and it started to grow and this is sort of a couple months in what Twitter looked like. There's a couple things that you'll notice it has friends and not followers so the process of making it sort of blog like took a while to develop the other thing you'll notice is the logo it's TWTTR very mobile and SMS driven so the only way to sign up for it was sending an SMS to the thing. We used a little Nokia phone and USB cable and some pirated Windows software to send and receive the messages so that you could send and receive them and that worked for a while. It was mobile only for a while it was just showing you 24 hours of tweets and you couldn't go back to the future beyond that. This was how many text messages this person had received, you know 700 text messages and a lot of people complained about that. And to send the tweet after we went from the phone to a service you had to prefix it with things so there were extra codes and the codes were more complicated than they ended up being. So that existed for a number of months but Twitter didn't take off it had this small group of users who really loved it and then about this time which is you know two years into the company I actually quit and I said I'm going to go off and work on activism and change the world again. This Twitter thing is you know some dot com thing. But Blaine who was the person who gave the ETel keynote with me and a bunch of other people stayed and kept cracking on it and Blaine who became the main programmer on it for a number of years. He didn't have a cell phone because he didn't believe in having cell phones and so he had no way of using and testing the software he was building and we had another developer this guy Florian who lived in Germany and he couldn't receive the text messages either. And so the two of them without telling him went and hacked up a little aim bot so that you could chat with Twitter and just get the messages that way. And so Twitter went from a almost accidentally a mobile phone driven thing to something you can get on your desktop. It worked on a mobile phone but you can also handle that stream on the desktop and then you can start building on top of it. And then because it was easy to be on rails an RSS feed was added so that people could build on top of it. Well, yeah. And probably the most important thing for Twitter's growth was the fact that all the commands, everything you could do in the web interface you could do via REST. And there was no authorization. There was no process by which you needed approval to access the Twitter API. In fact, Twitter had no idea who was using the API. It was just, you know, HTTP off basic, username and password and every command in the system was available to everything which means with one page wiki of this is how you use the Twitter API, hundreds of applications got created in all sorts of ways. And so that created the ecosystem. It was really vital for its growth. Yeah. There's, you know, after a while we bought some vowels and the applications started putting other display applications and sort of ambient discussion steps started appearing and being possible. And about one year after the hackathon people were still saying that Twitter was this crazy dumb idea but they were using it. And at the South by Southwest Festival of 2007 was the first time you started to see real emergent sort of blocking and crowd behavior using Twitter. And it was around finding which bar had free alcohol and didn't have long lines. And because of that Twitter ended up being sort of considered the darling. And after that moment in 2007 after a year of sort of struggling, which the API and the aimbot and all these other things got built, then it started to take off. After that, Odeo got sold off. The podcasting thing in Twitter today is sort of very similar to that. I mean this is actually several years old. But that's a sort of a rough history of how it got created and where it was coming from. Nope, thank you. Questions or discussion? I'm just wondering what, so I was, I knew about TextMob previously and I talked about it. I wonder what was the influence of TextMob on the eventual actual product and it sounds like you guys were very connected but was it just kind of a congenital thing or did you borrow ideas back and forth? So we worked on the Odeo office on stuff and then after we gave the keynote detail on TextMob and social activism using telephony we spent a week in the company where every single person signed up for TextMob and we made a bunch of groups and we actually did a usability and use analysis in the company. So everybody in Odeo used TextMob for a week and we also used a couple other services. We used UPOC and a couple other services and then we had a couple hour long whiteboarding session where we analyzed what didn't work about TextMob socialism. So that was a week worth of work of analysis of TextMob that happened a couple weeks before we built TextMob wasn't the only thing that went into Twitter but it was probably the piece of software and interaction model we studied most leading up to the creation of it. I'm going to jump in a queue here to ask you part of what's interesting about this talk and the story of TextMob on Twitter is the contrast between a very complicated story that you're telling which is that there's a lot of different things going on there are people coming in and there are applications that people are using as examples and analyzing and that's all feeding into the creation of this tool then there's this there's a different narrative which is the corporate origin story in various books that various founders and journalists are writing about it which tell the story of well some really smart Silicon Valley people were like sitting on a swing you know looking up into the blue sky and the idea popped into their head fully formed and then it suddenly exploded. So I'm wondering if you could, is the takeaway here or the key story here, is it about the direct influence of a particular technology on what then became this blow up tool that so many of us use or is it about the incredible complexity of the creation of any new technology or is it some of both? I think that it wouldn't have existed without TextMob. It wouldn't have existed if we hadn't built TextMob but the environment at Odeo was in some ways similar to what I think the media lab is in some ways. We had lots of people who weren't employees who came and hang out at the office. We had lots of people who had other jobs. We had lots of people who were doing all sorts of projects. You know we weren't very good at focusing on what we were supposed to be doing because we did lots of side projects and played with things. You know there was this guy Duncan whose official job was to redesign the Apple homepage and the Apple homepage just doesn't need to be done often. Between that he would spend weeks just hanging out at the office trying to build cool stuff and re-hacking our CSS just for the hell of it. We had half the blogger team who were like I'm tired of hanging the bus down to Google. Let's just hang out here and then they would stop doing their day jobs and we'd build shit. There's two reasons why the Twitter story is so narrow. Jack Dorsey who is sort of most widely known as the creator of Twitter was an engineer on the project and the rest of the team thought Twitter was going to be sold really quickly and the most likely acquire was Google and half the team were like dear god I don't want to go back to Google and so they picked Jack as CEO and gave none of the people who had been at Google any title in the organization so that the Google people wouldn't try and convince them to come back like tie them up with an earner. So that's really how he got named CEO then they worked on doing it. There was a fight between them and Jack was fired from the organization this is well known but he kept his Twitter email address and then he was pissed that he was fired so he went on the media when they asked him and said these guys didn't create Twitter I created Twitter and everybody has this ability to create this memory in their own head like the more you tell a story the more that story becomes real in your own mind and so I don't think he was malicious in it I think he was pissed off at being pushed out of the organization and so that story is an easy story to tell it sounds great the lone genius who has the idea and so the media sort of eat it up but real innovation is real messy the reality is that there were 12 people who participated in the hackathons at the organization to make Twitter and that's not an easy media story it's not an easy story to sort of consume and define who it is if anybody is the one inventor of it it's a guy Noah Glass who came up with a name he was the sole founder of ODO he was the most passionate person about the whole thing and he was completely written out of his name so I'm going to go back to the protest because for me the most interesting bit of this was not that innovation is messy because we all know innovation is messy and it's never very simple the most interesting bit to me was how community activism actually led to concepts which while you know TexMov couldn't like get off the ground as like this thing was used by millions and millions of people somehow the combination of transferring concepts from TexMov into this crazy process that ODO and Twitter was able to reach millions of people and I'm going to just mention one thing that you left out which is you know when we saw the 2011 use of Twitter in Arab Spring I was like well of course because I've been in the RNC in 2004 and used TexMov and it worked pretty well honestly blocking stuff and Twitter in 2008 against Republican National Convention and you know despite various FBI ratings and stuff it had been exceedingly successful in organizing protests so I was sort of like slightly not surprised when this capacity was picked up by people all around the world but to me it seemed like maybe part of the reason why that happened was not just hey it's Twitter but also somehow the genetic code from TexMov that had actually you know because TexMov had been as we've seen clearly used to organize protests and it was sort of not surprising everyone had a smart vote and everyone had Twitter that the same dynamics could be used again I mean there were things that were built into the technology sort of a philosophy of the open web, the philosophy of anyone connect the philosophy of you just need a phone, you don't have to take your real identity the ability for memes and ideas and links to spread quickly and activism and social movements and Twitter does those things well and I think that's some of the holdover it never surprised me but it wasn't the narrative that Twitter incorporated they loved being associated with the protests but they didn't want anyone to think they were taking sides so it was a narrative that was downplay it's more of a general question, I'm curious about what you see the current state of TexMov in some way I used to be like both in the media but I used to see those kind of things we are getting older I mean I see a lot of the ideas that were relevant in time and products and other things and things that we fight with became social media, social media in the late 90s and early 2000s that didn't exist, like the idea that people would cover their own news and everything else was a radical concept and now it's sort of seen normal well so the total obsession is around crypto stuff and security and privacy but there was a problem with indie media and a bunch of the tech activism stuff it was much easier to build software and scale it within the context of a start-up than the context of a social movement and so you saw the core development team of Craig's day for the media people the architect and core development of Flickr the guy who ran and built the way communities work on YouTube the lead developer, CTO of Etsy you know a whole slew of these organizations that are stuff where the activist going into the dark and I think that's a lot of what happens people attempt to dig into their same ideas inside this brokerage actually designed to activate the mind to stop activating my social movements but nowadays I do it every time so that's fascinating so they're inside Twitter that was a tremendous amount of fight over the direction of the platform at one point in 2007 or 2008 we had a bunch of talks on federated social networks using XMPP this was when PubSubHubbub was developed and so this sort of amorphous crew of folks inside the organization who don't the line between who was Twitter and who wasn't in the early days was very blurry I think a lot of people had root on all the servers who weren't employees or had no formal connection to the organization just because they wanted to help but so there was actually a point that was a service called Jaiku that was bought by Google and there were a lot of people who didn't know about it and at one point the software was written so that micro blogging could be a medium and there were a federation between these services so you could follow someone on another service and Twitter supported that on a protocol level and that direction basically that was abandoned partially because it was built on XMPP instead of PubSubHubbub XMPP is a little bit of a pain in the ass so when Twitter lost the person who was most knowledgeable about how to build and scale XMPP systems the federation was abandoned and then a couple years later the business people got afraid that TweetDeck was going to take over the company and steal their users and so the app ecosystem was abandoned and so there were all these points in which the development team said let's get more open, let's make it a medium like email and Twitter could be a big fish in a much larger pond and people said no, we need to control it for us, let's be the only fish in a very small one and so the Twitter we see today is the company that said we don't want the larger medium we're afraid of it, we're going to do the thing we control and so that like if there's anything that I think was a bad change of direction I think it was a lack of creativity in short side decisions because they could have kept the app ecosystem and the revenue share with the app creators so that the app creators didn't make money off the ads that were displayed on people, like there's no reason why they had to walk away from that except for sort of rushed decisions made in the theater That's the time they closed down the exact plan Yeah well they didn't turn it off completely but they started shutting down after they had more than 100,000 registered users and the narrative that was the Twitter success was because of the API was also because they listened to the users so retweets, app replies, hashtags, those are all created by users even Twitter search is all created by third parties and users and then incorporated into the platform so like the lesson of Twitter's innovation is look at how people use your stuff and support it and embrace it and then Twitter stopped its massive growth when they got afraid and started trying to control things and stop listening to it so like the tweet streams is a great example if you've seen tweet streams or this is an app that lets you post a whole series, like that's a user innovation of people using the platform that Twitter as a corporation should embrace and support because it's actual use and so Twitter used to be as a company very good now they've had many generations of complete turnover in the organization like no one I knew there there was no overlap between someone I worked there and someone else you know I think there's so many generations that the people who work there today don't know how they got there they've lost their own corporate history I'm curious it sounds like both in your organization and these other activists who were talking about it went to Etsy or went to YouTube like it seems like kind of a loss for all these people to go in with these wonderful ideas and then see it kind of get corrupted through the process and I'm wondering if you think how do you think these groups should be structured to prevent that from happening it sounds like there's a bunch of great people with great intentions and yet still under the corporate structure just like push things in a really negative direction yeah it's taking on a big believer in startups do a licensing or licensing their stuff under open source licenses you know I don't have a strong opinion of the MIT first license person of GPL but you see an organization like whispered systems which makes you know tech secure whispered systems was an angel backed startup they got funding they got acquired by Twitter they incorporated their technology in there but because they had dual license it was all open source licenses after they earned out moxie and the folks went back and just kept working on whispered systems so it let them play both places and I think the companies which are trying to capture users and capture value they don't actually care that it's dual license or it's open source like they care about the existence of it that there's an alternative path that's an out they don't care about it so I think that when people go into these organizations it's really important that the core technology get license under open licenses so that when companies make bad decisions you can keep going. I sort of want to build on Sasha's question one of my hobbies is industrial archaeology and there's a whole history here of the company that and if the company doesn't even know its own history anymore how can anybody else and how can this how can it be captured in something other than the self-serving biographies of the founders are you even thinking about that you talk about going back to your own email finding a way of keeping the original documents somewhere somehow. There's two things one, yes a lot of it is just in my gmail account the book Hatching Twitter Hatching from the New York Times which is sort of the widest read book about after it got created. It's very focused on the personalities doesn't really talk about why or where things came from almost all the original source material like the direct material is shit I forwarded. But the other place you get this is because we created social media. And all you got for it was a short passage about how you were the anarchist who wouldn't stand up during the stand-up and also I think he used the justification that there were anarchists involved in the organization to say that's why Twitter had a disorganized was disorganized and had bad management culture I did do things like we were talking about when the launch and I thought the software was ready and no one else thought it was ready so I just turned the site off we've been talking about it too long tell me I wasn't saying that I wasn't a difficult person to manage in the organization but you should ship software if you think it's perfect you waited too long I had a plan but the tweets and the social media and the documentation that exists publicly that in some ways we've taken pre-Twitter ideas, pre-social media ideas of how you document your own lives and now we document it online. And so now when I go back and find my own participation in the old history I'm going back and finding okay is there a recording of the keynote that I gave talking about how we can use text messaging and telepathy for activism the months before the Twitter talk that was created, that kind of stuff we are creating a history of it online part of the reasoning of this talk is the same kind of thing there's a documentary on my perspective on the creation of Twitter and how the role of it coming out in Al Jazeera English in November I spoke about talking about the way in which the interactive between activist and social movements and technology stuff but it's actually taken me a while to say I'm no longer upset about what is and isn't in the history and trying to figure out what the right way of not stepping on people's toes and not taking credit for other people's work but telling the story as I see it. One of the problems is that Twitter now knows the history like some of the official Twitter accounts saying Bravo was never there and it's odd because I have a Twitter account that dates from like 10 minutes after Twitter got started and there's all these photos of me there and there's like books so like the Twitter's own PR people they don't know the history that's actually not uncommon and they had a new museum director who started going through their archives and found things like I think it was John Wayne's hair and then had to trace back the corporate acquisition history that got them John Wayne's hair so the fact that companies don't know where they came from even companies as big as Dr. and Gamble that are big enough to have a museum and historians who work for them so that's not uncommon but again as you said what was invented here is the process of self documentation and my question to you is what are you thinking about in terms of your own archives a book sure but beyond that the original source material is what's really interesting I don't know if someone wanted my emails I'm not going to just post all my emails well no Andy let's just take one or two last questions is anyone who hasn't so the decentralization point was brought up did you bring up a little bit about like OAuth as well created by Twitter as a way of attempting to protect users privacy it's the way people log into like you log into Facebook you log in with there was a years of attempts to build OpenID and identification systems and login systems and they were all complete failures and OAuth was an attempt to standardize the way Flickeroff works in a standard open format there were a bunch of open meetings and a bunch of people got together and everybody got together in the Twitter office and said well we should have a way in which we do log in that we're not sharing user passwords and so OAuth was created written into libraries everything else adopted a bunch of companies the really ironic thing was OAuth was about redistributing power and protecting users and OAuth is what made it so that Twitter could shut down their own app ecosystem and OAuth has been one of the most destructive centralizing controlling technologies that we've adopted in the web and it was built by a bunch of anarchists who wanted to protect users privacy and decentralize the web so sometimes you can build and deploy technology that's massively successful and does politically and ideologically exactly what you didn't want to do so just to bring this point up for those who want to spend more time with Evan tomorrow at this time for lunch with W3C World Wide Web Consortium which is like the web standards body we're also brainstorming how to fix some of these issues around centralization OAuth definitely comes up and stuff like open standards for things like Twitter like streams and staff messages or something we're working on so that's at the stop center on the I think it's the fifth floor yes it's the G30 the gates building on the fifth floor just come out and take a immediate right and we'll be doing a lunch there this time tomorrow noon tomorrow at W3C thanks Evan and thanks everyone thanks