 Welcome, everyone. I know we've still got a few people trickling in, but we thought we would get moving for the panel this afternoon. So just so you know, you're in the right place. This is the Tech Data Frontier discussion. So thank you for coming and thanks for your interest in this particular topic. We're going to do this a little differently than the other panels that you've seen today. So bear with us. It's a little bit experimental. So all the Alliance for Peace Building folks, thanks for hanging in there with our experiment. We're going to run this a little bit more like a talk show. So we're looking for a lot of engagement from the audience. And so instead of doing, you know, each panelist sort of gives discussion, what we're going to try to do is frame out three different issue areas because the Tech Data Frontier topic is pretty large. So what we've tried to do is to pick three really relevant focus areas for peace builders and kind of spend a little bit, a few minutes, maybe 20 minutes or so on each one, and encourage your questions within those topic areas, you know, as we move along in the discussion. So without further ado, Juan and I just jump in. I am Nancy Payne. I'm the deputy director for the Peace Tech Initiative here at USIP. We've lost our slides, so I'm sure we'll get those back in just a minute. But we're not going to... I've got some pictures that we'll show a little bit later. So just make sure that those... Oh, there we go. We'll get to those in a minute. But the Peace Tech Initiative, for those of you who aren't familiar with what we do, it's a collaborative effort at the intersection of technology, media, and data to find effective ways of reducing violent conflict around the world. So why? Why do you ask that? Why now? Why this? Think about it really for three reasons. Data. Think about the number and amount of data that our digital devices now generate in a single year. They can do that in a single year what we've done over the course of human history, one year. And it doubles every year. So our ability to collect that data also continues to grow. That's an enormous challenge for anybody. You hear the term big data. It's a huge challenge, but how do we actually harness that and make it useful for peace builders? Mobile. Everybody knows these things are proliferating. And in Africa alone, there's about 63 mobile subscribers per 100 people. And in a lot of Arab states, a lot of Arab countries, there's 105 subscriptions per 100 people. So you've got people actually having more than one device. But how does this relate to peace building? More data isn't good enough. We need different data. So my boss, the director of the Peace Tech Initiative, Sheldon Himmelfarb, who couldn't be here today, he often says it this way, we need data to tell us not just what is going on, but also what people think about the things that are going on. So we at USIP are actually studying this. We're looking at this and doing projects in this field. We have experts in technology, social media, data and curriculum based media, all working here in the building and also in the field. And we work closely with USIP's country and regional experts on different projects aimed at using all of these different innovations in technology and media that are transforming the way that the international community confronts deadly conflict and promotes security. A lot of the tools that we're using are empowering people all over the world in the places that you work as well. Countries like Kenya, Nigeria, Burma, Iraq, to name a few, they're taking concrete steps toward countering violent extremism, tackling inter-religious tensions, preventing election violence and promoting local securities in a lot of other ways. But technologies are often created in isolated pockets with little chance of being brought to scale or applied to different conflict zones. You heard Liz Schraer earlier talking about needing all of us to feel like a community. We need to look at ourselves as a community and that's really what the Peace Tech Initiative is also trying to advance. Some of you may have actually heard about the Peace Tech Lab that our team is currently standing up that USIP is supporting. We've been working toward getting this lab set up and hoping to have that done within the next year. The lab is going to bring together engineers and data scientists along with other experts like many on this panel in peace building from USIP, other government agencies, NGOs, and in the most importantly in the conflict zones where we operate to design, develop, and deploy new and existing technology tools to prevent and resolve violent international conflicts. This hopefully we're working toward getting it up soon but it should be the first facility of its kind and it will be located here at USIP and we're going to work in close collaboration with the institute as we formulate the lab as a standalone entity and we're also going to be working with US and international agencies to be able to bring these solutions and actually scale them up. So with that let me introduce my panelists here and we can maybe jump right into the discussion. Here on my left is Michelle Gregory. Dr. Michelle Gregory is a research scientist with and she's at this she uh sorry I'm trying to read too too too far ahead social media and data sciences and you are you were at your University of Colorado and now I'm totally lost you were at the lab and now where are you? I work in innovative analytics a small consulting firm out at DC innovative analytics and training. So Michelle obviously specializes in data science and she also has a real focus area in social media and how that applies to the peace building field. So welcome Michelle glad you're here and Dr. Michael Best is joining us as well and he is an associate professor at the Georgia Institute of Technology better known as Georgia Tech and he also he he's kind of it has a has a foot in two fields he's in the Salmon School of International Affairs and he's also in the School of Interactive Computing at Georgia Tech so thank you Michael for coming and Michael has done a tremendous amount of work in the election violence violence space and has a lot to say on on that topic as as well as others. We've got Noel Dickover my colleague at the Peace Tech Initiative he's here at US Institute of Peace and Noel is a senior program officer and he has really focused a lot on this question of big data and how do we harness it. And Rachel Brown Rachel's here she is the CEO of Sissy Niamani and Sissy Niamani is based in Kenya and Rachel is thankfully here joining us now on this side of the ocean and they have also done a lot of work they're they're known for having a group of local Kenyan peace builders and leaders and has a lot of insight on again election violence and a lot of really good stories to tell. So with that I want to jump in what I'm going to do is pose a couple of questions and then I'm going to actually invite some questions from the audience on the topic of election violence. Elections become a real touch point for possible violence and actual violence in a lot of countries. A couple of recent examples like Kenya's election last year and Afghanistan's election cycle now underway showed some tentative success in managing this violence. Later this year and next year we're going to see a lot of elections in several places that carry with it a potential for increased conflict. We've got Egypt, Turkey and Ukraine coming up this year. We've got Nigeria and Burma and others in 2015. So I'm going to throw this out there Rachel and Michael specifically to you. You both have experience in monitoring elections using interesting new technologies and using mobile phones and using social media. Can you each share a little bit about your work in Africa and in Afghanistan and any insights that that you you want to share? Okay I guess I'll start. So Sisi Nyamani used a variety of different tools around the elections and just to give a little bit of background context in Kenya. By the time of the recent March 2013 elections there were a lot a lot of efforts in Kenya to monitor election-related violence and tensions. Probably some of you are familiar with Ushahidi. So there was an Ushahidi map called Uchugusi that was being run by Ushahidi in Kenya. There was also another crowdsourcing effort and map being run by a Kenyan government institution called the National Steering Committee on Peacebuilding and Conflict Management. Again they were collecting information and data crowdsourced through text messages and putting it into a map. Sisi Nyamani was actually sort of complementary in my in my view to these approaches in that we were not actually collecting information via SMS we were distributing it. We in the run up to the elections and I'll talk about this a bit more later had created a database of text message subscribers. We had more than 65,000 people throughout the country in over 20 target communities that we had reached on our platform and we were able to send targeted messages to different areas. And our goal was really to understand and interrupt communication flows that were leading to violence. So I would say our monitoring was a little bit unique and that we weren't looking for verified factual early warning. These people are moving from here to there and they're armed and they might be ready to attack. If we heard that that was happening and we verified that it wasn't true but that the rumor existed our goal was still to interrupt that information flow because what we understood was that even things that weren't factually correct even communications that had no basis in reality had the potential to influence reality. A rumor is a very potent thing and it can really rumors can often be self fulfilling and they can cause conflicts in different ways than what they're originally predicting. So our goal was to really look at how different behavior chains and information flows that could lead all the way from a rumor being started up until someone decided to take action and arm themselves and participate in violent conflict about that information or perception of conflict or electoral rigging or any other event. So when we were monitoring the situation we were monitoring a whole host of factors including sort of the mood on the ground the types of information and rumors that were going around as well as what was actually happening. So our response was very technology based and that we were sending these mass text messages that were targeted to different parts of our subscriber base but the monitoring was really simple. It used phones but it was actually a trusted network of partners that we had worked with 20 to 30 people in each of our target areas who we talked with about these information flows to conflict and who would alert us of any issues and then we had a team of about 10 coordinators I'm sorry 20 coordinators verifying that information by calling around to the different partners in the area reporting it to a central location and initiating a message. So we were using technology and this is something that's important to me the crowdsourcing efforts they brought I think they have a lot to offer and they can really target the masses but our purpose was to have a specific type of response and we built a monitoring system for ourselves that was really geared towards verifying what we needed knowing what we needed to specifically have that response and because of that our monitoring system itself while it did use phones to communicate over broad geographic areas it wasn't so technology based as it was people based. Thank you Nancy and Rachel and I think that it's really important what Rachel was mentioning because we too look to bridge these gaps from data to knowledge through analysis to action through engagement and I think it's that last step the knowledge to action leap that I would like to interrogate maybe even as the panel goes forward in 2011 my lab at Georgia Tech the technologies and international development lab partnered with a group of youth activists called enough is enough in Nigeria enough is enough is a consortium of youth activists they call themselves the Facebook generation of Nigeria they're young they're networked enabled and they're a little bit pissed off and they want change and we partnered with them to develop a system and system of processes to monitor the 2011 election and respond or act upon actionable information that we picked up across the social media we built collaborating with the consortium of youth activists and my lab which is made up of political scientists computer scientists visual designers and lawyers a software technology and process called Aggie the social media aggregator this technology vacuums up across multiple social media platforms everything from Facebook to Twitter to Ushahidi instances to blogs to anything you can syndicate over RSS during the Nigerian election we which was a three week election we brought together about a half a million different reports across these various social media platforms and we're able in real time to provide some analysis and response not dissimilar to the type of workflow that Rachel described we went through processes of relevance testing to reports as they were coming in in real time to then verify ability to then action ability in one particular instance some of you may know in the April presidential election in Nigeria in that year violence outbreak broke in the north of the country in particular as the opposition party candidate called into question the early results putting good look Jonathan back in the presidency during the election thousands of people were killed the country truly was on the brink and so we quickly were able in this instance to reconfigure the searches and queries that the Aggie system was performing in real time to focus not just on electoral irregularities which had been what primarily we have been looking at say for instance a polling place that have run out of ballot papers but instead to focus on these acts of violence and potential acts of violence occurring primarily in the north in one I think you know useful anecdote though not to be taken too far we picked up tweets that were coming from a pair of sister students at a polytechnic in Kano a major city in the north they were tweeting about mobs amassing around their particular dormitory we had an embed as part of our escalation process at the national security coordination facility in Abuja the capital of Nigeria after testing for relevance verific very terrifying or veracity and action ability we then escalated this particular series of tweets to the security center in Abuja who dispatched law enforcement and troops and were able to secure the safe passage of these two students coming new for 2014 and I'll foreshadow this a little bit but perhaps return to it later we're beginning to look at new ways to link a broader set of data inputs into this process of data to knowledge to action and in particular what we're integrating is formal observer data based upon a parallel set of projects coming out of my lab working in collaboration with the Carter Center to develop appliances and software for platform for mobile handsets and for tablets that enable their formal trained observers to obtain election based knowledge so we're now going to deploy next month at the akiti state election and then in 2015 February for the national election again in Nigeria a technology that will integrate across social media integrate bringing in and aggregating in addition the formal observer data coming off of the Elmo technology excellent you know it's interesting you talk about the verification process and each of you took a little bit of a different you know have a different take on it and I'm just curious you know if you if you would sort of advocate for one method or another I think Rachel you you made the comment about verification and not you know not being able to necessarily verify everything but but also being able to counter rumors and the importance of doing that whereas Michael I think you're taking a little bit of a to be able to go in and actually try to stop very specific actions I don't know if you know if you want to talk a little bit more about how the verification process works and and maybe what is on the horizon on that front for this this upcoming election cycle really quickly the verification we use mostly is two fold one is triangulation across multiple sources or media types and the second is what we would call out of band verification meaning verification via our own partners who might be on the ground in the right location or by retweeting back to the tweeters and saying is this really going on can you provide some more data that kind of thing but I think Rachel's point where it's not just if it's true but if it's something that could influence is of is a very important insight and and there's nothing I would say particular to our processes or workflow that would would disallow for that kind of important addition yeah and I would definitely say I mean we were verifying to find out if things were true so if we did find out that something was happening we had a different type of response including a message and usually contacting relevant responders but I think my point was more that we also cared if it was just true that something was being talked about and I think what this really gets at is designing monitoring based on programmatic goals and so I think what was important what worked for CC Niamani our goal was to be able to send messages and interrupt flows of information that could contribute to violence and because of those goals we knew we'd also find out about information that could be acted on and responded to in other ways and we did set up structures to do that but our entire monitoring effort was created to enable us to fulfill that goal and I think there's different purposes to different programs so I would think that probably different monitoring approaches can be taken depending on we also had a we had a big subscriber base we had 65,000 people and made a conscious decision not to ask them to report message to report information via SMS by a text message because we knew that we didn't have the capacity to deal with that huge amount of information so we also made decisions based on what was our capacity to filter information and respond to it and we thought that our best chance at really being able to take action on the issues we were finding out about and to find out about the issues that mattered in our target locations was to train and work with about 20 people in each area where we were working but then I think there's different value to different approaches the approach that Michael was taking was looking at a much larger data set and obviously the verification procedures for that is going to look a lot different because he's also not working with pre vetted people for example on Twitter or Facebook where they're collecting the data and the people that we were working with we pre vetted them so we just had to make a couple calls to see how far our rumor had spread or if someone had witnessed it firsthand I'm going to turn to the audience and and as I have audience questions I see somebody right here in the middle we've got one here and then we've got another we've got three so why don't we take these three and then if there's more when we start with the the woman here in the middle and then we'll go from there okay that's you're great go go for it and then I'll take yours and we'll we'll we'll try and answer them in a row so I think it's so cool because you're working in very very real time and responsive one of the things that and I don't do this work myself but I have friends that are doing it with the defense department and especially around disaster response just wondering because one of the struggles is the effects of time so you know you get something that comes up you know and then how do you do you have like algorithms or have you kind of come up with some best practices for you know this road is blocked well 10 days later it may not still be blocked well so how do you start to you know change the storyline because otherwise you just get a map full of things that or you know an information space that's full of stuff that hasn't been able to accommodate the time thing so great I've got somebody in the middle here too and then we'll just we'll take them one by one thank you go ahead thank you hi Jessica Murray with search for common ground so I have a question for Rachel how did you get the 65 subscribers and how early did you start kind of campaigning to get those subscribers okay and I think there was one more here on the side so so I like what I like about these these stories is it's beginning to close the classic gap between early warning and early response but I I still don't quite hear what the early response mechanism side of the thing is um you talk about actionable information but then who is in place to actually take the action so have you preset we have local organizations local partners and some of them might be in a position to do something whatever that is but in at least one case Rachel you you actually were sending you're getting police to come in are they pre you know have you had a relationship or communication with them ahead of time to know that you may be getting information and you may be passing it on to them I know that in some places people are setting up you know various forms of local peace committees whatever who are preset to to be able to take this so far so connecting that the information that you're talking about the information flows to those those response mechanisms at the other at the other end and then what are the variety of responses that they might be able to undertake and have you know is there some sort of protocol that decides well what is the appropriate action okay thank you why don't we start why don't we just go online here um I think what I heard is the the challenge of on the effects of time uh what are some practices for dealing with real-time data versus data that is minutes or hours or days old um just looking at humanitarian assistance disaster response that that whole field did this this dramatic change in innovation with the Haiti disaster you had most of the the formal response mechanisms incapacitated so there is both this need and opportunity for for all sorts of different approaches and in the problem there is you probably know is how do you get to valid information in most now recognize the first week of a disaster you don't have your assessment teams on the ground you're not going to get it and really the best you're going to find is social media and mobile data and so the change happened that that you know previous to that you still had formal you know crisis response organizations waiting for their assessment teams to to to make findings but that just it couldn't happen there so you now have a very different way where um you and Ocha even has you know there's this digital humanitarian network where volunteers around the world they did things in Haiti like actually map Port-au-Prince and the Haitian diaspora translated messages from mobile and there's this whole process that was called mission 46 36 that got into it but the bottom line that field has really evolved in three or four years the peace building field is is almost like a before Haiti time I mean both of these instances are amazing things happening there's not a lot of them it's it's a relatively new thing to think through how peace building and technology work and in most peace builders aren't really tech savvy right and most civil society participants across the world aren't very tech savvy so there's there's this you know almost cultural change that peace builders have to do to to apply to that but thinking through real-time data in a peace building context we're experimenting now I mean we we have some ideas but I think if you you talk to the HADR folks are gonna have a much better idea now how it works than say three years ago right so yeah so I would add to that there is a very large effort in algorithm development but looking in social media in particular on on the issue of we do have upcoming elections so it's not just monitoring in real time it's how much can we figure out beforehand what can our algorithms predict that's going to happen and then you can have a very good baseline and have a better place to shift from and then your algorithms can work in real time to update that information so there is quite a bit of research in this area and you'll hear me talk about it later today as well but there's definitely lots of algorithms for determining real time the geolocation is a huge part of that determining that in real time as well and getting that information out to people who need it I'm gonna hop to the last question then we'll come back to the Rachel question and that was because I think that there is a little bit here related to early response and I think that's a really key question because you all know better than anyone that early responses it continues to be a challenge especially around election violence so you know are we closing the gap and the question was who's in a position to take action and what are response protocols Michael do you do you want to jump in sure so I'm trying to answer both of them a little bit at the same time we also are very much engaged in this algorithms for immediate response we work on two time frames micro and kind of macro so we're doing things in real time receiving quite a lot of data so in a normal election we might be saving 50 reports a second and we're prepared to scale that to a thousand a second I think right now and we have algorithms to try to make sense of that both human and computer based and then our response algorithm both human and computer based really is based on these embeds and the way it's been done so far in our intervention so far which are have been in Ghana Nigeria and Liberia all three countries we've had our own team members in the national security coordination facility in the national election center the independent commission in the traditional media center so all these places have had a place where broadcast media folks are at print media and in the coordination center amongst the international observer missions and sometimes domestic observer missions so we've had our people in those three countries successfully embedded in those places and then if it's a security issue we go to our embed who has relationships in the security coordination place if it's you know the polling places has a logistical problem we go to the electoral commission where we have a relationship I'll add we were at the Kenya election and we failed to establish relationships with in particular the government actors who were not interested in the relationship and so we were not able to escalate uh through to a security response or to the electoral commission so it requires the government buy-in and I can share a little bit of our experience which I think um if I back up a little bit about the sort of goal and purpose of CC Niamani our goal was to support local peace groups and peace actors on the ground to better use strategic communication including new types of technology so that's where our sms platform came in we weren't set up as a monitoring organization and so like I said our monitoring efforts were all based at being able to support local peace builders to have this informational response to send messages um so before the election we spent a lot of time preparing ourselves for that response um and I think probably I'll talk about this a little bit under the social media side of things but actually some of it connects to the panel that we did that we just heard from previously on marketing piece um we actually had someone from the marketing industry from Ogilvia mother come work with us for two weeks pro bono um we had laid out different situations that we could predict might happen different types of situations that would lead to violence and done co-creation workshops where we invited members of different target demographic groups this included young unemployed men or men who had participated in violence before but it also included a category of people we called information spreaders so this was people selling things by the side of the road or in the transport industry who we know had a really big role to play in the spread of rumors and information in their community so with each one of these groups we walked them through the situations that could lead to conflict and had them create messages they would send to appear with a specific behavior change goal so one goal might be stop a rumor from spreading we want this person to decide are there not to spread the rumor or just to ask a couple questions about where is this rumor coming from and interrogate it another specific behavior change goal might be a lot of situations would happen especially in urban informal settlements where people said you can tell that there's violence young men start grouping um and talking with each other and the conversations get defensive and they start making plans so when when we were monitoring we were monitoring for things as specific as young men are starting to group and talk about potential rigging and our message that we would send then was a message that had been created specifically to encourage people to go home we did a lot of work on risk mitigation so that these messages wouldn't be alarmist but our preparation for response was a lot about creating messages message templates that would be appropriate for specific situations as well as message guidelines and some understandings of our audience part of what we pulled out from these workshops was what were incentives for people on the ground what did we need to remind them of to change their mind it wasn't usually be peaceful it was usually something like if we're violent we're not going to be able to run our businesses and make money um so it was really pulling out these learnings and being ready to respond on that front and I will say um Kenya was a little bit unique in our experience I think we were there for a really long time since 2010 and we did develop relationships with a lot of government responders and Kenya actually had one government institution that was supposed to be responsible for coordinating all the responses to electoral tensions so we were really escalating to them and we did wish that there was a network of local mediators we could escalate to for sort of an in-between response but yeah one last thing I know I've been talking for a little while but I do think in terms of closing that early warning early response gap I think that there's maybe a need to not design those two things so separately I think when you design early warning over here and you design early response over here and you try and bridge the gap especially at a large scale it gets messy and confusing and really hard to coordinate and I think the two need to be designed more closely together not just designed as totally separate entities did you answer the question the last question about the subscribers how did you get to have so we worked on that for a really long time we started piloting an approach where people would subscribe in late 2010 and adapted our mobile technology and experimented with different types of subscription at first people signed up on paper and then we developed the technology further so people could sign up by phone so really we started as early as early 2011 doing major subscription we did our biggest subscription drive actually in the months before the election just because that's when we got financial support to do it and we subscribed that was about 30,000 people that we subscribed in the three months before the election until then we'd sort of been bootstrapping along but I do think it was really important that we were in the community and developing relationships all along and I think it was important that we weren't just coming in for the election and because our goal was really response and having a relationship and influencing the community having done that work for the years prior and built relationships is what let us get information about exactly what was going on it's what let our partners trust us and what made community members I think more responsive to getting messages from our organization because they were familiar with us and again back to the former panel that was the messenger right we spent a lot of time building our brand and getting people to know and trust us in the community and now which it was door-to-door it was local peace groups going door-to-door subscribing to people there we go um there's there's just a few pictures here I think of Rachel's colleagues in Kenya and and talking illustrating I think some of the work that the great work that she's done I think we can move on to the social media topic so this is this is great and I think this kind of morphs right into the picture showing here and what a revolutionary tool social media has has known to be has has really evolved into it has become a tool for activists and journalists and citizens trying to get news out or get news in about different revolutions and and different things happening in in conflict areas so you know in many ways it's also become a tool for organizing and engagement you know we we heard again from the previous panel about the you know the viral videos and how that's a tool for raising awareness and delivering a message I'm hoping that all of you can give some specific examples of how social media is actually now being applied to the peace building field so I think why don't we start with Michelle and share a little bit about that okay so social media and for peace building is used really in three ways the first is how these two are talking about it where you have a real-time situation and you're organizing online this is also true the recent case in Brazil where protests were organized on facebook sort of crowdsourcing we all know it a flash mob is and so those techniques can be used by local peace builders to to quickly organize they another way that's that's the local real time another major way that social media has been used for peacekeeping is in informing the international community and I would put monitoring political elections in this category we all know about the Arab Spring that would be in this category there have actually been instances where you can find locals who will be interacting on facebook say in their local language and say hey we need to get on twitter and start using english so we get more attention from the western world there's lots of documented cases of that then there's I think a major way that it's used and and especially in academia is post-conflict analysis and in fact we learned from post-conflict analysis and this is I think a usip study in 2010 done with bitly that that it turned out that it wasn't so much in the Arab Spring that people were organizing themselves it was mostly used as as an information place to go for the international community it was mostly used by the international community not so much locals to organize so those are I think the three major ways that social media is used in peacebuilding and all of those different techniques require different questions and different tools that you have to use in different capabilities to be able to accomplish those. Anybody else want to take that? A really interesting example of one of use of social media recently in Burma. Anybody here are the pensigar movement? So Nathan Lott who's a inactivist from the saffron revolution he's started this ICT organization there called MEDO in April they had their their water festival which is you know associated with the new year and he did a really interesting example of sort of online organizing to offline action right which is where you see a lot of these even social media you know collectivism isn't isn't going to make a difference but if you're able to do coalition building online that leads to offline action it's a it's a really powerful approach and the neat thing there I mean Facebook in Burma right now is pretty much what AOL was in the 90s here right I mean that is the internet for the very small select group that have it and if you're using Facebook in Burma and contacting somebody chances are you're contacting somebody in a city but they they did this this this effort on Facebook that led to people offline putting flowers in their mouth and going to public places posting posters one of the real challenges there is some of the hate speech is driven by Buddhist monks and their place in society is not such that you can challenge them directly so this this flower power approach of putting these flowers you know and sharing this these pictures and and broadcasting it wasn't a direct you know response to the hate speech as much as it was do this instead right and that's a really interesting campaign it's a positive campaign and it seemed to have you know certainly got a lot of buzz I haven't seen you know the the impact of it yet but in a place where only seven percent of the folks are online that's I thought was a really great way to get the youth to take that and then to start having offline action that you know is going to show up where the bus stations are and so forth so that's that's from April I thought it was a great example just sort of to write on that point of online to offline action I think that's something really important to remember is that social media impacts how fast we get information from how many people how much information we're able to get and it really influences people's perceptions and behavior but it's only powerful in so far as it's actually influencing people right it's not the tools or the social media in and of itself it's is it actually having an impact on people and on society is it translating into something offline as well is it changing social norms is it changing the way people are acting and I think this is one really big challenge for the peacebuilding world in terms of specifically using social media to counter calls to violence and speech leaning towards violence and and one thing I I also want to point out is that this isn't just an opportunity for actors in the peacebuilding world the opening up of social media and different mediums of communication is also an opening for anyone who wants to incite violence to do so really effectively the reason ccmani chose to use mobile phones in kenya is that we saw in 2007 and in 2008 during the post-election violence violent actors use mobile phones really effectively and they influenced a lot of people really quickly across geographical areas um in manmar that's facebook in other areas that's twitter um it's not always one medium but when there are these mediums available to influence a lot of people really quickly violent actors are taking advantage of them and they are at an advantage because usually someone who's trying to influence people towards violence has a very clear action we talked about this earlier right do we have something concrete that we're trying to sell I think the peace message often gets lost because what does it mean be peaceful a lot of people take that to mean do nothing right the um flower talk example is something where people could take an action to symbolize what they were standing for having really clear goals of what we want people to do is going to be more and more important because violent actors also have a bigger more influential arena where they can be calling people to action I think that's right um I actually think as social media relates to peace and conflict that it is jannis faced that on the one hand it can be used as a tool for peace building but on the other hand it can be seized as a tool to enable conflict what I you know and and this morning I mean today at lunch miss mark said it's not about the technology it's about transformation and you know I believe that but as a technology I guess I want technologists I guess I want to shamelessly have it both ways and really think about how to architect social media to privilege peace building and to detract opportunities for conflict and so here's a modest result that might suggest these kinds of architectures for peace uh in our analysis post talk as we're talking about the temporal issues the macro analysis uh across african election data we find that some social medias seem better for the peaceful response rapid response now social medias in general are architected for connection and for disclosure you connect say on facebook by forming your group of friends their intimates and quasi intimates and then you disclose to them for instance I had the turkey sandwich for lunch these architectures can select between connection and disclosure and what we find in the african context is that the selection more for disclosure which we find in twitter than for connection amongst intimates which we find in facebook may support peace building and I say it's modest and may because these I don't think we should run off and and we shouldn't over learn this result that we all want to support twitter or we all want to support disclosure over connection but what I does I think point to is a real deep engineering and human question which is what are the technologies or architectures for social media that we want not what we're given not what especially very large corporate interests are giving us like a facebook or a twitter what are we wanting and then how do we architect them designing the social media platforms to allow us to privilege peace building over conflict so I'm just going to play devil's advocate for a second here we have to maybe we all don't remember but some of us remember that facebook actually didn't start as a big corporation worth billions of dollars right it started in a dorm room and and in fact there's a lot of a lot of the social media that's out there how it takes hold is because it's inherently a social device and and we've seen that in china with the advent of tencent this was a bit like snapchat but the government's monitoring sign away bow and so it was the actual users that came up with this new platform here's the way we can get by the government monitoring right and so these technologies inherently we all know twitter now is huge but there will constantly be new platforms that people can because they're really they're individual driven which is really has been the power of web 2.0 technologies is that they really start at the localized level for use so we don't have to pick what's out there right we can also sort of go with what's organic sort of like choosing sms in the kenyan example because that is what people use and in pakistan it's a combination of sms and twitter so there's also a corollary in there from the hadr space where there's now a recognition that that people in the affected crisis zone aren't just victims they are truly the first responders and to the extent that you can empower them with better situational awareness of their environment and actions that they can take they can really make a difference and i gotta think there's a corollary in the peace building space the real peace builders are not people in this room right it's it's people in the conflict zones and so how you know like rachel's approach they are getting 20 people in each of these areas those are the peace builders so social media enables that process but but it's not magical it's hard work right there's there's there's a lot of things you need to do to make it successful but we see that this is possible and it's it's again it's in a time of experimentation to see really how far that that approach can go one thing to add on that actually um that i think is interesting talking about how we can support local peace builders one sort of unexpected and i'll put this out as a hesitant result too because it's mostly anecdotal um that we saw from our work is that um violence tends to be really visible because if there's violence going on you see it right there can be five people running around in a community wreaking havoc and people feel unsafe even if the majority of people want to be peaceful and often in situations where there's violence i think we all realize that it's not an easy choice between peace and conflict it's often people at huge risk um you're up against people really being afraid for their lives for their families lives and making decisions based on that and often peace looks pretty invisible and what we saw from sending messages because we had several thousand subscribers in each community and because we'd worked with these different peace groups to go out with these branded t-shirts and messaging into the area they were telling us that essentially peace began to look more visible so some of the subscriber feedback we got when we called people and did um and surveyed them was that they felt safe for getting the messages and just knowing that there was someone looking out um about violence in their community or that getting a message gave a lot of people said the message gave me the courage to preach peace in my community so um i do think social media can also increase the visibility of peace which is often seen as the absence of something it can really create an identity that people can latch on to but we have to also be really smart and strategic about how we're doing that and in turn that identity can give credibility and courage to people on the ground who want to speak up um do we have questions anybody on this particular topic all right well i'll ask a couple oh we've got one right here jjope okay no no please go ahead this is a it's a collaborative effort you should be tweeting live thank you excellent um i'm doris mariani i'm ceo of nonviolent peace force um we've got about 225 people on the ground in places like south sudan Myanmar and the philippines and um in the philippines the early morning early response mechanism has been very important to us um and and i stressed the response part because this came up and we're now trying to figure out um how to make it better now that the peace agreement has been signed by the milf and the government the new banks of morrow sub-state that is going to be established is going to need a statewide early warning early response system and and we just recently learned what what you shahidi means and you know we don't have a lot of tech people so i'm most curious to hear from you uh where does one start this process uh i mean we can articulate maybe what we would like to see but the journey from here to where we want to be is where where you all live and and and then um the other thing what we don't do is we don't name and shame so so the disclosure part as you put it is i mean whatever whatever information we report first of all gets verified and the idea is the reduction of violence to create a more peaceful world but but how does what one go on that journey if if one wants to serve the communities and keep one's staff safe uh and use honest technology that's an excellent question and definitely i think the safety question is definitely is one that you have faced in the past but it's one that i think every day we we do have to think about so just anyone want to jump on the where do you start the process and then talk about safety i can know so so the part about learning the technologies right this is this is a real challenge and there's sort of two levels to it i spent the last three years of my life connecting local technologists with local civil society folks first at the state department and now here we've we've done a series of three of these in iraq uh around anti-corruption it turns out the real innovation in most developing countries is with the local tech community especially if the local tech community is tied to the open source movement in most places i've been to and michael offers me a counter example so i'm not going to say all the open source community has some real power it's it's it's really based on a meritocracy in a lot of ways and if you're in a country that has endemic corruption you're really not exposed to to meritocracies is is a normal thing but getting tied into the open source community it has these ways of local meetings called bar camps and hackathons and you develop some level of trust relatively quickly and you do it over time and and they start to know each other and they innovate what doesn't always happen is this connection to social good but it turns out technologists in developing countries love doing social good projects just as much as they do in new york or paris or anywhere else so sort of a first order would be if you can connect the folks you're working with with the local technologists in the philippines they're going to be able to tell you what technologies work what services are available on the sms gateway and come up with innovative solutions to do what you need you need to come up with a format that works for that i developed this at the state department called tech camps we now have them here called peace tech exchanges and we've done this really nice back end thing but the idea is to come up with this two-day very interactive small group discussion format where you bring these two groups together more peace builders than technologists so you don't get conversations about ruby on rails and you know linux and so forth but it's really about the issues that you have but but there's a process for doing that this this you know it's online and it's shareable although there's there's still you know a lot of work putting in practice but bring those two groups together that's going to give you a level of innovation you're just not going to find in the peace building uh field the other step is for for us in what i would refer to as this nascent peace tech community to start laying out online some of the resources so you know what is this crowd sourcing thing and i now hear about crowd seeding we're actually you know pay people to give me specific results or event reporting platforms what does that mean i mean there's a lot of these in the challenges they rapidly change that's critically problematic for security right so skype is safe today in some places or not what is it going to be in three months from now when something you know when microsoft bought it what was the implications right uh who owns your telecom who owns your isp you know these are all questions that that the local tech community is going to have some level of knowledge at and if you could find the security person in that country or set of them they're going to be very helpful in answering that question for you know what can you do how can you interact safely online there are also other inherent risks you know whatever you can do as a peace builder and a peace building network and community the opposition can do as well and whatever you can see they can see and there are obviously i mean we've all heard a very real examples one of these was last december where the syrian opposition was in their twitter network was infiltrated by government supporters who are able to download some malware then the government then could recognize through this malware who was part of the opposition and that's always going to be a problem and the technology does change quickly and in cyber security this is you know it's always the race we have as soon as you can patch something up you've got the other side with just as good technological skills getting around that patch so something to be aware of there's also geolocation is great but you can be geolocated then so it's sort of really understanding how the technologies work and what they do and what information is out there is really important to keep people safe and i would add to that also i think people because these things are relatively new and they're changing all the time i think um someone who decides to go to a rally they understand what the risks are it's a normal behavior and the risks and benefits are pretty clear to them people don't always know if they're texting into a crowd sourcing platform what the risk is and i think that there becomes a really big responsibility to make sure that anyone who's participating understands the risks not through some legalese agreement that they clicked when they agreed to have a facebook account or something like that but really understands what are the potential risks and can make that decision for themselves because people in complex situations do decide to put themselves at risk and they decide what degree of risk they're comfortable with and i think what the thing that we can really do is make sure that people understand exactly how much risk they're exposing themselves to two other quick things i would say on the community side would be one i think it's really helpful to observe what are people already doing what technology are people using what's normal for them what's comfortable and building on to existing behaviors so not setting up a platform that's totally different from the way people are interacting or calling their aunt or uncle or brother or sister to say this is going on right how do you build into existing behaviors because i think that's much more likely to be successful um and again i would say almost designing response-based monitoring so knowing these are the types of response options we have and as we get information in our goal is to filter it towards one of these responses right so really looking i would suggest working both ways looking at monitoring to response but also looking at possible responses and how would you get the information you needed to have them i think we've got question down here christin i think we need the mic for that's all right thank you christin lord usip the panel is making me think back on a comment i once heard from a really brilliant engineer at incutel and he told me uh that when you're trying to use technology to address a problem more than half the issue is not the technology but surfacing the problem and articulating the problems in ways that technologists can then address you're all nodding okay this is very well good um but so the question i have for you all is how good a job is the peace building community doing at surfacing the right problems and then articulating them in a way that the technologists can address and if we're not doing a good enough job how do we get better so this is what i spend my days doing um in peace building communities and other government communities uh there is this real lack of understanding between technology providers even when they're being paid by the source to build the technology and the people using it and it really takes someone who can bridge all of that um as peace builders you're not you're trained in languages you're trained in political science um you're not trained in technology and xml and json formats and what it means um if you estimate geolocation versus um have actual geolocation you know that's that's not your job and it shouldn't be at the same time you have the technology providers thinking we know how to do this we can do this but they really don't understand those questions um and and you need to find people or hire people that can really bridge that gap um i i can tell you so many instances of um of clients thinking they have some technology and it turns out they don't they they thought they were paying for it but the technologists are like oh no that's our special sauce of course we don't do that um you know i mean the funniest quote i have from from one of the peace builders is you know it's that's fine if you're building toothpaste to make that decision but uh we're talking about people's lives here right so so we need to know exactly what's in the data not your version of what's in the data um and that that is a gap that is getting better with the younger generation is as people are just more tech savvy coming out of high school and college um but it's it's really understanding those questions and what it's not just the technology but it's also the data that would actually answer that so do you want to go okay go ahead so so the peace tech exchanges is is really geared towards answering that exact question and it is to me a combination of both bringing international technologists to fill gaps that maybe aren't local but but having the local technologists who may not understand the peace building world at all but i i almost have to take as a given and sometimes this isn't true but the peace builders should understand the world and what they're trying to get accomplished and if they do you have this nice overlapping venn diagram and in sort of the progression of the day is to expose people through you know this fire hose approach of brief technologies then give them a deep dive in an area then to completely step back and talk about the problems and challenges of say corruption in iraq independent of that and then to do this connection at the end of the day and what we're trying to do is come up with an engagement point that that that technologists can work on but that totally describes you know the problem that that uh you know the the civil society folks are having like a good example that comes to mind we did one of these when i was at the state department in romania with the roma population there and one of the problems that came out is how can we improve the image of the roma in the the media including new media right that's because they're seen as thieves and gypsies and all that a technologist can look at that and say oh yeah digital storytelling that'll work for that or we can do a blog-based approach to start gathering videos or i mean there's like probably five or six or seven different options that a technologist could pose for them and so the idea on the second day is a selection of problems to work on in the first day the technologists are leading all the sessions and the NGOs figure out where they want to go the second day the NGOs take the lead and the technologists go to the problems that they think they're best able to solve so to me it really is you've got to connect these two communities and that's the best way you're going to sort of come up with those innovations and i do think it's a key question um that is posed and that is um you know are the is a peace builder are the peace builders actually framing it the right way um but it really goes back to what rachel said earlier which is you've got to you know you've got to design and um a program based on very specific uh goals and objectives and and you've got to start with what the goal is and you've got to start with the audience and who the and beneficiary is and you've got to start with the people aspect and not the technology aspect so i think that's you know that that's really what we all need to keep in mind because you can get a little bit carried away with the the next you know whiz bang technology i do have a jump in yep and then we're going to move on to the next topic real quick john when he mentioned the sort of the over emphasis on metrics and evaluation technology is something that that really is about rapid experimentation you can't stick your metrics in place for a project you're going to do eight months from now and expect i mean you're you're going to need to learn to fail to modify to change it's it's rapid experimentation that works with technology it's not let me plant it ahead of time and that's the process i'm going to stay with because gosh that's what my funder told me to do i mean that just doesn't work in that field it's a it's a healthy tension that we we have to live with every day isn't it yeah um well i was going to connect these two actual interventions because in some ways it's about how the civil society connects to the technology and then the specifics of peace building connecting to technology and i work on this problem every day in fact i would say it's the thing i spend most of my time on is how do these uh communities connect and it's because obviously as we i think have been hearing the civil society group or the peace building community own the problems in the context but they also are laboring routinely under a misapprehension of magical thinking and that's because it's not as michelle said that they don't know xml or whatever it's that there isn't the basis of a literacy to allow i'll give you an example um this morning dealing with some of my partners in country who didn't understand when we said we were building the phone app that that didn't mean we were building an application to run on all phones for all time well you know as technologists you normally wouldn't even ask that you know that wouldn't even occur to you well i never promised all phones all time because that's just it's just not it's never was a going concern that it'd be all phones um of course i'm building it for those phones but um so a question i have and this actually even goes also to my sort of lack of my more pessimistic feeling is that it's a very long conversation in my experience it's a very long conversation start now to have it because your community however you define your community needs to be talking to the technical community because those connections take time and it's a skill to actually for the peace builders the people that want to use the technology for their real world purpose to get them to ask the right questions right they're not used you have to be very goal into exactly what do you need to do and it's not you know and we'll help you get there but but teaching them to frame the question right so it can be translated into a technology solution is really an art so i have um i've left us not very long um for art is that two minutes overall do i get to do our last topic okay why don't we do one question um around the topic of big data how's that for an ending and uh you know this is obviously an enormous topic um but i'm gonna i think this next slide actually kind of gets to um the core of what we really wanted to talk about which is who are we actually producing data for um we have in our our team just giving superman data for superman exactly um data for superman the big decision makers here you know the washington the policy community are we producing data for them or we actually as peace builders trying to produce and get data in the hands of the people on the ground who really are as we heard earlier doing the work day to day on conflict resolution so i guess with that why don't we throw out one question and give everybody a chance kind of a go around and then we'll end in in three minutes but are you you know who how do we actually are we in a position to be able to really take all of this information that's feeding into us and turn it into very local actionable uh data and if that's the case do you have examples of of how you're doing that i think we heard a little bit from rachel and ccdmi you know in in the election context but um do we have any thoughts any closing thoughts on how we're tackling that monster challenge no why don't we start with you yeah i'm gonna start off by saying no and the reason is we aren't the ones that can do it right it's it's back to to this picture here and so you know so often we do data collection in a conflict zone we go back and do the analysis and sometimes we'll give them the answer right but the really the answer is for those guys because somehow obama is going to be able to fix this bokeh haram thing if he just has the right information well no it's really about people in the conflict zone changing changing their expectations of the future and how does that happen in really what you're getting to is they need to own their own data profile it to be able to make big data granular enough for folks on the ground in a conflict zone to actually make sense out of it they almost have to be able to do their own analysis right they in how do you get you know people in dc don't really do data driven stuff very well so asking you know peace builders in conflict zones to do it that's not an easy challenge in it's it's right where michael said i mean our approach we're looking to do we have this peace tech lab idea but the idea is to find one of the best technologists maybe two from the conflict zone we can't afford the best technologists here but maybe we can in cartoom bring them to the lab as a peace tech fellow teach them about data analysis data visualization and how to use it but just as critically teach them about conflict analysis in mediation in facilitation skills and then have them go back to the conflict zone with a physical space if possible where they can bring in peace builders to a you know what we refer to as an open situation room where there's physical maps on the wall that they can list data at in over a period of months they're able to come back and start eventually starting to ask the questions that that is going to help them and the answer probably isn't going to be a fancy 3d moving diagram it's going to be an alert to a cell phone right but you're not going to get big data granular enough until the peace builders themselves are asking the questions that they need answers to and you gear at that and that's a long-term prospect and that's a prospect where you need to have the capability in the country from a trusted source that they can actually start learning how to think from a data driven mindset i would actually say yes and no and and this goes to the other point that i made that you really have to know what question you're answering before you can decide what data you're going to use and what tools and if you're answering a question on early warning and response and you're in country and you have this finite event that you're dealing with big data is probably not the solution that you're going for there unless you call you know getting sms's from 65 000 people big data which some may some may not i i'd say that's very little i'd say 50 tweets a second is a small amount of data um so it's all relative to what you know on the other hand peace building is not just about the people on the ground there's big data is quite relevant when you're looking at post-conflict analysis and you have the ability to take the time with the data and to find the right tools um and there still is uh uh quite a bit of um you know maybe brock obama is not going to solve it but there are lots of people working on the same problem in the government um and they are spending vast amounts of time and money on infrastructure to handle big data and um and on that so we need it everywhere on the other hand the techniques aren't great and you really you know it's not about you can't ask the same questions with big data that you can with other data and and data science is not just about how do we handle big data it's about knowing when to use one versus the other and and how to handle that it's not just about scalability it's if you actually need the scalability okay um michael last one and then i'm we're getting the proverbial well the the answer to the question as it's posed in the slide is clearly no it's a it's a holy grail kind of um question and it's clearly the answer is no because as i understand the way you form the question Nancy is how are those kids in that school pictured on the slide gaining and empowered and being empowered by the data and the analysis of the data as opposed to just being the source of the data if that's the question our answer is we don't we don't have an answer yet um i would take disease in there right i mean health i mean there's a lot of data that goes into keeping populations educated and healthy and and a lot of daily data analysis maybe for for development reasons and health reasons but on the on the specific peace building yeah i don't we don't but and it's definitely a holy grail question um it's a question that we are endeavoring to answer in our modest ways by new analytics that then push out to say the kids in that school but it's absolutely not a question that's been answered yeah rachel take us home and then we're okay i'm gonna say i actually think that fundamentally the order that we have to think about these things in is often a bit off because i think if there's an assumption that we should monitor and monitoring is good i'm not sure that that's a right assumption i think that what we have to go in and say what change do we want to make and if one of the steps that needs to be taken in order to create that change or influence a situation a certain way is that we need certain information that we can act on or that someone can act on then we need to back up and design a monitoring platform that gets information and data to the people that need to take action or that need that information for a response so i think who are we producing data for varies in one case it might be we want to influence international decision makers in another case it might be that we want to help local responders have better capacity to respond faster but what we need to know first is what is our goal and then back up and say is monitoring even necessary how much do we want to monitor using big data medium data small data micro data right but i think it's really the order that we ask these questions it has to shift it has to be goal first and then do we monitor and then how do we monitor okay well i want to thank you all for attending and i want to thank the panelists very much so for um for their excellent input thank you very much thanks to the panelists before i formally close the the session for today i want to ask you all for a favor if i might since you're captive in our auditorium i'd like to ask you to help us be a better resource for you here at usip i think many of you know we do training and education courses online here in the fields we develop research but papers books toolkits we throw a p-stack lab we're developing technological tools so on and so forth we're doing a lot of this with the intent to help all of you be more effective and all our partners in the in the field be more effective if we can do things that will help you please let us know if we're doing things that aren't that useful please let us know that too or how they could be more useful because that's why we're here uh please good way to do that is by getting engaged so if you're not following us uh if you're not getting our regular emails if you're not following us on twitter the usip feed but also you know all of our scholars you know nulls on twitter i'm on twitter share even sharin's now on twitter you know others say it's a good way for us to keep up with each other um so please uh that that's my uh favor i'd like to ask all of you but in closing the session today let me just say i think it's been a terrific day i hope you agree we're very thankful to the alliance for peace building and everybody there for their partnership here i'm supposed to give you um some requests from afp and usip one is to please fill out your survey because we'd like to do this even better next year we can't do that if you don't have your feedback we want data it doesn't have to be big data it's just enough data for us to make the conference better um so please fill out your survey um it's low tech it's very low tech so even the techaphobes can can deal with it there's a cocktail reception at the beacon hotel at six that's very important news uh but the last thing i need to do is thank some specific people who really helped make today possible melanie greenberg and the whole team at afp i already mentioned of course our panelists right here and then my own team which i definitely don't want to forget to mention because i have to keep working with them um and these are only a small subset so i'll thank everybody else collectively uh limwood hand sharen morris selenia canota jamie schilling bill vaughn matt lolich steven blake brian hammond the porters adriana sanchez and anna velasquez at seasons i want to thank all of our team for the work they do every day but especially thanks to all of you who are helping us to promote our shared vision of preventing mitigating and resolving violent conflicts around the world thanks very much