Sunil Abraham (Director - Policy) is a Bangalore based social entrepreneur and Free Software advocate. He founded Mahiti in 1998, which aims to reduce the cost and complexity of Information and Communication Technology for the Voluntary Sector by using Free Software. Today, Mahiti employs more than 50 engineers and Sunil continues to serve on the board. He was elected an Ashoka fellow in 1999 to 'explore the democratic potential of the Internet'. He was granted a Sarai FLOSS fellowship in 2003. Sunil is also a sub-board member of Open Society Institute's Information Programme. Between June 2004 and June 2007, Sunil also managed the International Open Source Network a project of United Nations Development Programme's Asia-Pacific Development Information Programme serving 42 countries in the Asia-Pacific region. Between September 2007 and June 2008, he managed ENRAP an electronic network of International Fund for Agricultural Development projects in the Asia-Pacific facilitated and co-funded by International Development Research Centre, Canada.
22 November 2008, Bangalore
Some people one knows by circling around them and finding your self to be in the same environment again and again. This is how I have known Sunil Abraham for many years until one day in Delhi we ran into each other on a formal occasion and we ended up discussing our work of organizing events and being responsible for the well being of all who attend. This role of being the initiator, producer, orchestrator and director all at once is a highly complex one and the dilemma's one faces can be confusing. Being the catalyst in social political events where many people have a lot at stake, demands a strong presence and conviction of one self but of course one has doubts and so it was great to have a chance to share some of our concerns. Sunil has been a catalyst in social technological developments in Asia over the last decade. With good humour, a lot of friendliness and a sharp mind, he has been building bridges, switching structures and created several initiatives that last till today. Being both a techie as well as a social entrepreneur, he has easily shifted between these domains and many of us have benefitted from his capacities. We meet on a Saturday afternoon in his brand new office behind Wockhardt Hospital in Bangalore. We just arrived from Amsterdam and he is about to go off to the UK.
Sunil Abraham argues that an individual exists online because of transactions online, not because of having a webpage. You have to do transactions all the time to prove your existence in time and space. Records of direct transactions should be available for the public. Trust emerges online in the social web of institutions and people. From this perspective credibility appears to be a transferable property.
There are two types of transactions. There are machine records of transactions and there are mediated witnessed interactions (editing of pages, public postings, making links, 'I show I know you'). Digital witnessing contributes to the establishment of reputation and authenticity, through hyperlinks and records of transactions between humans and machines as well as between humans mediated through machines.
Presence and trust raise the issue of authenticity. Online there is hegemony of text. Existence on the Internet only occurs in the infrastructure of knowledge. This subverts existing authenticities in societies. There are over a hundred versions of the story of Rama as they have been told over the centuries. Today on the Internet there are only two or three. Normally people online can judge real from fake, especially in a series of transactions. However people with a lot of smartness and who take a lot of effort can 'con' a community.
When tuning one's presence when one is about to become each other's witness, this tuning is actually a testing of the boundaries of the social contract. Like the handshake is a tuning, as are two modems connecting. There are no universal rules for handshakes in an online community. The community will correct each participant, through comments and remarks of other people, till everybody is equally uncomfortable. Complex dynamics characterize such tuning moments. Who is the boss? What is the right tone? How to fit in? And these are also influenced by expectations, which are the result of previous experiences in real and online spaces.
When discussing the different tones in mailinglists it can be concluded that the same system is different in every context and is defined by many variables: level of English, use of the Google translator, political context, culture in certain 'scenes', whether people know each other IRL, how people know each other online and/or IRL.
When tuning one's presence in an online community one can read the history of the community, lurk for a while and decide how to pitch one's presence. Young people experience 'giving data away' as establishing authenticity online. That is how one exists online, by opening one self up. Older generations do not experience this as such yet, younger generations do. In the offline world one can live with different identities in different worlds. In the online world one has to find one's own lowest common denominator between the varieties in identities one has, to be able to keep as many people happy. The impact of having a life online causes that one knows much more stories, direct and first hand of others people's vicarious experiences. This gives multiple perspectives, more relationships, more understanding. One becomes aware there is not 1 single truth; more empathy and understanding emerge for personal issues and a deeper understanding of political issues as well.
Sunil Abraham is deeply involved with the open source community. He defines open source as follows: to ensure that intangible goods will have the same properties as tangible goods when acquired for free or for a fee: to use, to study, to modify, and to share. Open Source is a license, a specific way of social organisation and a specific way of publicising. Witnessing of each other is a profound mechanism in the Open Source community. Open Source communities are hierarchies under constant negotiation. The hierarchy is more dynamic than IRL as well as more contextual. The hierarchy can change dramatically depending upon the subject at stake. Trust in the Open Source community functions on different levels:
The building of software is all publicised. Therefore in an Open Source community one works for an audience. Failed performance towards the deadline of a release date damages someone's reputation and generates a loss of credibility. Who has previously contributed successfully to the documentation, the bug report, the patches or to the CBS archive, will be more easily believed and accepted. One can conclude that the contributions to an Open Source community are acts of performance within the legal framework that defines access and ownership. The hierarchy of trust and authenticity is based on current and earlier performance in the community. Emotions evolve in relation to the performance: to be recognized, to fill the space etc. Benevolent dictators of the meritocracy orchestrate and mediate these performances, emotions and relations between the participants in the community.
Abraham argues that Autonomous Systems consolidate human behaviour and change human behaviour. They can only emerge out of a practice of reflexivity. For example Google gives credibility to a page because real people give value to the page. The more links to a page, the more credible it is rated. And people can debunk it. A good autonomous system mediates human behaviour with its algorithms. Because the machine mediates human behaviour, it is becoming implicated in the business of authenticity and trust.
Authentic machines can organize their own destruction or defeat. Only in such a case an autonomous system acquires authentic presence. Like human being's presence, systems should be capable of steering towards well-being and survival or not. An autonomous system should be capable of gaining or loosing authenticity. This is dependent on who has access and on who controls the system. In this respect there are two kinds of systems people use: systems under their own ownership and control, and systems not under their own ownership and control. When a system comes into being, branding and social context create its credibility. When a system is established, its moments of crises and celebration will move its credibility in one way or other direction.
When downloading an Open Source patch, trust is mediated through open source software, open communication and trusted encrypted download protocol that has not been hacked so far. One only needs a limited amount of time to download and this happens between independent actors on the Internet. When all is connected and the transaction is happening, the network behaves in an autonomous way. This is network reality. In other contexts, not using open source software, trust is established in different ways.
Human beings give trust to systems and base their trust on systems as well. This can be false. A lot of trust is based on ignorance and perception instead of on reality and analysis. Trust is a hugely important question, no simple answers like more databases etc. There are a lot of irrational components to the design of trust.
And when already we left the building Sunil added: the best way to establish trust is by smiling and making people laugh...
The following is an edited transcription of the conversation. Film fragments of the conversation are included to illustrate parts of the transcribed text.
CN: Thank you very much for wanting to speak to us, just before you go traveling. I would like to think with you the next hour about how it is that we change, how we relate to each other. For a long time people and societies communities and villages, neighborhoods, organizations, were based on I would see you, you would see me and because we would see each other we would enforce a certain social conduct on each other. My action becomes a deed. And the moment I see you doing, your action becomes a deed. I can testify about it and I can also do intervention. And these things are changing very fast because we work a lot of time on-line. You have been working on-line a long time your self. And I was wondering, can you describe how it is you recognize people online?
SA: Well, I recognize people on-line, I am able to trust people online because I look at the web of social relationships with other people that I might know, or other institutions I might know, recognize. So it is as if credibility is a transferable property and by being associated with the credible individuals or institutions somebody earn or gets the property of being more and more credible. But to go back to the other thought I had, when you were kind of professing your question with some background, I want to tell you a story that happened to me when I was organizing a large event. As you know NGO's are very particular about systems and procedures, and they have very evolved management systems that are often implemented in these donor agencies. As part of the event we had to get a cultural troop to come and perform at the camp, so that the international audience could benefit from our culture. So the cultural troop came from some village, and when the local organizer, our local partner, an Indonesian NGO went to the village and spoke to this cultural troop and ask them for a quotation, they had never generated a quotation in their life before. It's called a performer invoice. It's not the real invoice it's an indicatory invoice. So this time the Indonesian NGO created the right quotation. It was just a fabricating paperwork, in order to fit the system. In future they will realize that we cannot only fabricate the document, but we can also fabricate the numbers. So in many ways you could have a system, which is designed to create trust in a remote context, but the end result of the system is direct opposite. The system will actually teach NGO's from developing countries to be more corrupt. So just having a remote system or just having electronically mediated system does not automatically guarantee that we will have more trust, or even the same trust that we had in a system which was socially mediated through physically witnesses. Here, in most parts of India, where people buy groceries, almost everybody who works at the grocery shop gives items, products to the customer and also puts their fingers in the cashbox and deals with the cash. But if you go to MacDonald's or Barrister in India the whole systems has completely changed, because one person is responsible for cash, other people responsible to serve you. And there is a sign that says right outside: 'You have right to receive a bill, if you don't get a bill, please cal this number and complain'. What the sign is actually saying is that the person who represents our company, Mac Donald's or Barrister, is potentially corrupt and we don't trust him or her. And you shouldn't trust this person either, so please ask for this documentation. What you realize is people are quite clever. And they will able to do all that paperwork and still do something, which will completely betray your trust. So this was the kind of thought I had when you were bringing up your interview background.
CN: This resonates very much with my next question. As a designer of technology, which you have been for many years, you have created possibilities for people to meet each other. What is such a format? So like the stories you just tell, can you go deeper on what such a format is?
SA: What the design of such software would all be about?
CN: No, the nature of the format. So when we meet we have a biological reality as a format. If I hit you, you get blue, all those things. But in technology we create formats. Like all the social networks, you have a profile and you said very clearly that you check/trace reputation of people or reputation you could see. If I am not on any website you won't believe I am here any more, even though you knew me. Yes she has vanished, where is she? Even when I walk in, you'll think what happened to her, because she used to be in all those things and now she is not there any more. But with those formats we are actually doing something to the human presence. And certain things we find important, certain things we do not find important. So as a designer of technology, in which people have to collaborate, what are important features for you?
SA: from the question of this initial establishment of trust?
CN: Establishment of trust and of being able to witness each other, to know who you are dealing with, to recognize other people
SA: I don't know if I can answer that question, I haven't spent enough time thinking about it. But I think, on the Internet, and it is interesting that you use this term witnessing. You exist, not because there is a page which has information about you, it is not because there is a record in a database that establishes your presence or your existence etcetera, that's not how you exist on the internet. You exist on the Internet because of transactions. Either some other entity, either a human being or a machine has noted activity from you. You edit the page and therefore the machine noted it down; it's history that you decided to edit the Wikipedia page or somebody else wrote you a mail on a public mailing list and therefore you exist. So you basically exist in space and time in a sense. So if all the transactions between you and the machine, and all the transactions between you and the people were only existing 5 years ago, then people won't believe that you exist today because all these transactions seem to be from a limited time period. In order to exist and continue existing, you have to constantly have transactions with machines and individuals and the record of those transactions should in a sense be scrutable to the public if you are trying to establish credibility amongst the public. Or it should be scrutable to the user group in which you want to establish credibility. Then the other really strange thing that happens is, that we have this unbelievable phenomena that you can have online presence when the online person itself doesn't exist any more. So I have a friend and a friend also of Patrice, who was a very respected hacker, philosopher from Africa, and unfortunately died of heart at the age of thirty-five. But he still has a face book account and nobody knows the password and nobody can delete that account. And his friends continued to put all sorts of messages. And every birthday they write on his face book wall. So in a sense you can have presence today, even after the original object no longer exists in it's original form and now you have presence in the sense of an oral memory that lingers on through these interactions between machine and entities. Then you come back to the question of these interactions. There are two types of interactions. There are interactions, which are actually a communication and all the machine is doing is recording the content of the communication that went between two entities, with the machines and the human beings or human beings amongst themselves. And they're also these other transactions, which are witnessing transactions. So when I see an interesting page on the Internet, I bookmark it. When I make a link on Facebook to that page, I am witnessing to its presence. Or, I think, when I archive it, when I make a copy and put it on my website, when I bookmark it, or when I connect it through a social networking site, by this act of digital witnessing, I am actually contributing to the reputation and authenticity of that digital artifact or digital entity in some sense. So when we design an online system which will try and do this, this is the most important feature I guess, the feature of a hyperlink, the feature of a record, of keeping a record, not so much of the person and it's details, but the transaction between the machine and the human being, or the human being the human being. So unfortunately I haven't thought enough about these questions so I think most of what I have said must have been quite silly, or basic, and covered already in your research.
CN: No it's not, so just let's continue. Otherwise I'll think this question is also stupid. It's just a real strange field of research because it's very mundane because everybody does it and it is hardly formulated. So if you say the hyperlink is actually..
SA: Can I say another thing? So in a sense the establishment of presence and trust is also to some degree dealing with the question of what authenticity. And what is happening, and coming back to the whole NGO story, is the hegemony of text. So unfortunately the first way to digitalization is completely based on text. Not based on audio, not so much based on video, etc.
CN: Or touch
SA: Or touch. So when we look at a knowledge source like Wikipedia, Wikipedia considers something to be the truth if there is a textual reference to it. And in many societies like in India, large numbers of the population is illiterate and a large quantum of our knowledge is old. These systems completely subvert existing authenticity in society. So to give you the most important example is the story of Rama. Rama is a warrior God in Hindu Mythology and we have very clearly more than hundred different stories of lord Rama from south India to North India. But since the textual references are only two or three on Wikipedia, we have only one story, all the other ninety-three stories of Rama are inauthentic, false and they simply disappear. So when we say people exist on the Internet, I don't think they exist outside the infrastructure of knowledge. In many ways we establish our authenticity through our relationship with the entrusted knowledge. So if the knowledge itself is considered inauthentic and fraudulent, then the identity of these people will also be considered inauthentic and fraudulent. We won't worry about their political vibes and their cultural rights, because we don't buy their authenticity.
CN: This is exactly what I want to address, this gap between the real life situations. Because we have known each other for several years, if I disappear from the Internet and I walk into your office, you will still think, 'hey there is Caro', you will recognize me, because you have known me long enough and in several places. So there I can survive the not being part of the social networks or social knowledge production. But as a professional I am discredited if I am not present in those networks, till today. I am at the moment supervising a master student who is interviewing people in social networks. It's very shocking news everybody believes that all profiles are true, but this is a hundred percent! No doubts. Because it is the hyperlinks, as you say, it's the hyperlink that grades the trust. So if you are false the hyperlink will reveal you, it's only a matter of time, so yes you can trust social networks completely.
SA: No. I can give you examples where it has broken down
CN: So tell me about breaking down. When does it break down?
SA: There is a conference which is coming up, which is called Internet Governance Forum, you might be taking part in it, and there is a very important mailing list from civil society perspective called the Civil Society Corpus, which is working with Internet Governance Forum trying to amplify civil society opinion. And on this mailing list there is person called Jeffrey A. Williams. And nobody is sure whether he is one person or two people or what he is. But he has been rather clever in building his history online through a variety of transactions. He is very present on the mailing list and he is quite informed about the technical issues that challenge the Internet today. So if you were to read one or two of his postings you would think that he is a credible guy. The second thing that he does is, every time that he signs, on his signature, he says I represent the INE group. And the group, when it was three years ago, he said it had sixty four thousand members, and now it has come up to something like two hundred and sixty four thousand members. So it's a very carefully designed false number in his signature. But that signature occurs in every single post that he has made. And when credible people engage with him, because he is intellectually credible, their names are mixed up with his post tag so that establishes his credibility as well. But what happened is, he went too far. He tried to falsify even events. And then people tried to even attend those events, and then they couldn't attend those events and then they realized he was a fraudster. And so now there is a whole collection of mailing list posts, which establish that he is not true, and they try to call him and stuff like that. But then he hit back by classical defense. "And see, if you really need to know about me, why don't you file and write a 'freedom of information request' to the US government, then you will know about me. And nobody seems to have done that, so nobody knows even today. And because the mailing list and the space are trying to deal with questions of privacy, anonymity, all those kind of things, they don't even block his contributions.
CN: So he is a good 'conman', he knows the system and he knows how to do it. I also know some of those stories. But in social structures between people, this is not an issue because people truly use it. Or do you think it is?
SA: I think people may be able to judge between what is real and what is false on the Internet. Because to give another example, my wife, now and then when she is upset that I keep befriending all sorts of random woman on Facebook, without making sure that I really know them in real life etcetera, she might divorce me on Facebook. And then people write questions and say, how come, what happened etc., etc., etc. And the moments it happened, two or three times, now people make comments, which reveal that they're completely aware of what is happening. So they no longer believe that remarrying her on Facebook is actually a re-marriage. So they are able to see these kinds of things. But if it comes to some other question like my educational qualification or something, then maybe I could get away with a lie for a longer period of time. I don't know, I am quite surprised by this result you have uncovered that people trust data completely.
CN: Lets focus on the tuning of one's presence online. For example when we meet in the hallway just now, I tune my presence to you and you tune your presence to me. That should be two-lane. So there is also tuning happening online. How do you tune your presence to people online that you collaborate with? And especially people you don't know?
SA: I think this question of 'tuning'; I like to use a kind of different metaphor for it. I would call it the handshake. So it's when modems of connecting, there is the initial transaction between the modems which establish a lot of things. So what you are trying to establish, in an online community is often very complex decisions. For example, who is the boss in this community? Who will take decisions and who will be the people that will follow in those decisions. Then you also are able, as part of this handshake process, to decide what is an appropriate tone that can be used in the communication. So in many ways, I think what people do in this process of handshaking, is the testing of the boundaries of the social contract. So some people try and will be unnecessarily aggressive, just to see whether the community will take that kind of aggression, and if they get away with it, they continue on that line of aggression. And some people will be completely the reverse, be unduly polite and apologetic and see if other people say: 'actually you don't have to apologize for those things. I understand what you mean and this is a very partial answer I can provide, which is, people seem to have some expectation of the social contract, based upon the previous interactions in similar either real spaces, or online spaces and then they attempt to go through the handshake, by pushing the boundaries. By seeing how far they can get away with various types of behavior in the group.
CN: For example is that in the Nettime list you have to find the right tone? How do you know? You use standard general polite forms? Or some people don't do it at all?
SA: No, I don't think there is a norm that everybody follows. It is as if each member on the group will attempt a particular form. And then through interactions with other members, when people say, no that's rude, or that's an invalid statement etcetera, they get corrected. And then they all settle down at a level where everybody is equally uncomfortable in a sense.
CN: Yes. What for example happened to us: Vibodh introduced me to you, I didn't know he was going to introduce me to you, so this is embarrassing, like I don't know who you are. You decided not to feel offended, which is really kind. You said 'Oh, but it would be really nice if you remember me". If this would happen in real life, maybe you would be much more offended. So there are also certain things that you don't know and then you have to take some extra time. I find for example, when working online with people, especially people that I don't see a lot, like you, once in a few years, I have to really sit and tune my mind to know you and then I write to you. So, that takes a lot of effort.
SA: I think what you could say broadly, the argument that you are trying to build, is to say that people are extra cautious online.
CN: Well, maybe I, I try to find whether there is a special preparation to be online. For example I interviewed a woman who works with programmers online who she never met, for many years already and they make great software. She could in detail tell me how she does this. What words she uses, when she, because they have to manage a deadline, someone falls ill, etc. She had all those lines, she could just tell them to me. What you say is that it is different with everyone.
SA: No, I think she might have reduced it to a system, because she was dealing with the same number of people. I would like to say that it is actually quite different in every context. So suppose I am dealing with discussing Kashmir in India. I find that overall, the whole conversation thread is very, very offensive, every single comment can be construed as insult, that is the tone that they are keeping, and they are happy with that and they continue discussing at that very violent tone. Then you find a very different example, which is the free software group people, and there are two kind of extremist there; the really out an out violently rude people, and the other extreme which is the really polite people who try and build the community. So you have on a daily Linux user group, now, a list of people who are saying, please I am subscribing. And then you have all sorts of other comments saying that you have to pay us or buy for us. And then you have to do this very complicated ritual and a long discussion, which is basically insulting the people who want to be subscribed. So that kind of tone is acceptable on that Linux user group, but it is not at the same level of saying that you're fascist, which is acceptable on the Sarai mailing list, which is now discussing Kashmir. And then you have still other groups, who may have never ever met in real life. So now I am working with a Copy South group, which is an international research group looking on copyright and we have never met in real life. So it is very, very polite, and very, very tentative in tone. And then you see similar differences, even between the countries. So, if you look at mailinglists in a place like Thailand, I find it very polite and in a place like Indonesia not so polite. So I don't think there are kind of universal norms for how people behave online. I think there are various variables, which work in different ways. For example even competence in English. Suppose everybody is clear, it's a European mailing list, and everybody is clear that they are good in English, and then you will find that the language becomes a lot more sophisticated, there will be people hinting at things, and people will pick up those comments etcetera. If it is a mailing list where it is clear that it is a different group and people are half Spanish, half English then deliberately people will remove that kind of language because it is going through the Google translator. So it is still an online mediated discussion. But just because it is going through the Google translator, the discussion will dumb down and will become a lot more simpler because they don't want the Google translator making the joke offensive, etcetera.
CN: How did you learn this? And how did this change your language? How did you learn to recognize all these tones and languages?
SA: Personally, before I join a community, I usually go through the history of the community, I read what discussions have been like before, or I lurk. I think that's another way, lurking in many ways is a handshake exercise. Sometime you wait to see what the other people say, and then you decide at what level you want to pitch your interaction...
CN: your presence...
SA: your presence
CN: Did your life online, did it change the concepts in your head? Have you changed language and concepts?
SA: Around questions of trust and?
CN: Yes, but also like how you look at things, how you perceive things. So do being involved in this practice, in this elaborate cross cultural, from simple language to elaborate language, being present in those environments, what is the impact it had on you?
SA: I don't know if there is a specific change
CN: or maybe if you had been a professor in school here in India, and not doing the international life that you have been doing, what would be different from who you are now? I am trying to find how spending so much time in mediated presence, how it changes us.
SA: Ok, I think the difference between people who don't have such a large online presence like me, could be the kind of vicarious experience. So when I talk about an issue, because I often know stories, which my friends have gone through and told to me as a narrated narration, not made abstract or not theorized etcetera, I am also being able to enrich my life and my thinking and my theorization through these vicarious experiences which other people have had. So I think we are more full of stories, compared to someone who is just from one context. So it makes it much more complicated for me to have a simple opinion on any issue. Because I hear all sorts of stories that make the picture much more nuanced. So suppose if someone is just in India and looking into the question of piracy then there is a wonderful story to say: Indians have so much more access to global knowledge, thanks to piracy. There is a rise in expressing culture; there is a rise in cultural production, etcetera. But because I have also gone to Tajikistan or if I have a friend in Tajikistan who will tell me the story, I see how Russian piracy and English piracy wiped out the Tajikistan culture. So piracy is not only a positive value, positive notion. I have to have a much more complicated thought around piracy thanks to these vicarious experiences. But I don't think I am answering your question at all.
CN: Well, you say, the witnessing is actually the hyperlink, and how it changes you is that you can hear and recognize more stories then before.
CN: So you become a better being because you have more understanding of more things?
SA: Yes, I think that's a kind of trusted theory that other people might also accept. If through your relationships with other people you develop empathy. Then you are much more likely to be somebody who is much more accepting than somebody with a kind of fixed agenda of their own.
CN: So I like to go then now and take this as a bridge to the open source as social engineering. Open source, maybe you will say in two lines what it means, but could you comment on how open sources and also communities of open sources create a shared morality? So could you comment on what is open source and how such a shared morality comes into existence?
SA: Ok, I think the best would be to define open source is to compare it to your coat, which is a tangible object. And there are many freedoms that you take for granted when you buy tangible products; the freedom to use it for any purpose, the freedom to study it, the freedom to modify it and the freedom to share it, either for free or for fee. And open source is an inundation based on copyright law. It tries to guarantee the very same freedoms that you get with tangible products. You get the same freedoms when you buy intangible products. This is the old design and simple definition of what open source is all about. There are two ways in which open source creates trust in the community; the first is the license itself. So what the license is able to do, whether it's a copy left license like the GNOPGL (36.30), or a copy center license like the BSD (?) it's able to establish to a large number of people the terms for access control and ownership of technology. And because it clarifies those terms, a large number of people who have never worked with each other directly or never met each other directly, are able to establish trust and work together.
CN: So it's done by the rules of ...?
SA: So that it's the legal component of the game. But that is not sufficient because in many places we have seen open source communities, which have started of with great good intentions, suddenly collapsing. So that usually happens when the social fabric of that community is attacked in some very serious manner. And often those arguments etc. can be totally unconnected to technology and totally unconnected to the legal aspect. That happens because in the world of open source there is this other kind of commonly accepted theory and structure of meritocracy and the phenomena of benevolently dictator. So what benevolent dictators do in a sense in the community is that they are repositories of trust and authenticity. So for some reason, historical reason, the community is linked to deposit, or placed their trust and credibility with a particular person. And what that person often does during the transactions, is make space for other peoples ego in the community. And that's a negotiation, because in the community there is a limited space, virtually as well, so you can't let somebody dominate it and that prevents new people from coming in etc. So there is in every open source community a kind of social contract that is advertised. So there are phenomena like this where they say BSD hackers are very rude to new comers. In a BSD community it is very difficult if you are a newcomer. But there are other communities and they are very proud and every new comer is treated very specially, there is a lot of leeway given to newcomers when they make mistakes etc. So it is that kind of a social contract, that is a result of the benevolent leadership of that community. Because it is always through, what one will call, arbitration. All disputes are not settled outside using law. And people don't bring up the license agreements; to solve the dispute is not about copyright. The dispute is about 'how come he managed to get away with that behavior, but I don't manage to get away with that behavior, or something like that. So the dispute is usually within the group, and often publicly, and the person that takes the decision is usually this benevolent leader or benevolent dictator. So I see this as the kind of way in which the free software community establishes global trust.
CN: So there always has to be this benevolent leader or dictator?
SA: Not just one. It's a hierarchy. So when you enter a open source community, from day one, the moment you become aware there are people who are really fully aware of their participation in the community, they don't understand where this fit overall, and they are still going through that process of negotiation. But once you have decided, then it is almost a clear hierarchy. But there are two interesting features of this hierarchy. First thing is that the hierarchy is quite dynamic, and while in other classic social hierarchies, for somebody very junior to clime up to the very top it can be quite difficult. In the free software world it is possible for you to change your position in the hierarchy quite rapidly. And the second thing that is different is that it is also context. In different contexts the hierarchy behaves differently. So you could have a Linux user group that come in, meets in an office, and on a technical issue you will notice that these are decision makers. But when it comes to advocacy then suddenly the hierarchy completely inverts and the people who not give up on the technical issue, will become lambs and follow us and will follow somebody else who already has some capability in advocacy.
CN: So what is the role of witnessing in establishing the hierarchy or in the dynamic of the hierarchy?
SA: So witnessing in establishing the dynamic of the hierarchy is very, very heavy. For example recently there was this incident in Cochi. There was a national free software conference and some free software advocates who were opposing novels relations with Microsoft. They set up a peaceful protest. And then the organizers called the police and had these people thrown out of the conference. So there is several questions; was the protest peaceful, were the police brutal in their behavior, did the organizers behave in a proper manor, etc. And the way that this whole story is being constructed and it is on slash, it's on the mailing list etc., is that different people are coming up with stories of what happened. So students, who are part of the college, they are putting up entries, the activists themselves are putting up blog entries, frequently asked questions about the incident, etc. And then they have photographs and mobile video clips. And it was very interesting that the police tried to delete all the mobile footage, so they tried to interfere in the process of witnessing. Then they came back to Bangalore and they attached the mobile phones to the computer and recovered the files and all this story is being said and put online. But let's look at a totally different witnessing that happened suddenly. Richard Stallman wrote to this activist and said: "congratulations on being the first person to be physically abused for upholding freedom, software freedom". Now what this blogger has done is he has taken this message and stuck it on his blog entry. So you see how complicated the witnessing is. The witnessing was already a mail from Richard Stallman to the blogger, private mail, he took the private mail and put it on top, because he knows Richards has a huge credibility and through Richard Stallman's credibility, his story will be validated, because nobody else's story has been validated by the community. So then the other people attacking him are people from this Marxists list. And there is this whole conspiracy theory that the Marxist Party is trying to court the free software movement. So they will establish their credibility by saying 'I was an editor for a Marxist journal" or "I was a Party member" etc. .So the whole witnessing of this incident is happening in very, very complicated ways.
CN: Can we go back to just working on a piece of software collaboratively in an open source. How does witnessing function there?
SA: When you are building software, now it's a very kind of classical technical question, there are different ways in which you exist as part of the project. So if you write documentation, then your name is against the documentation, if you file a bug report, your name is against the bug report, if you actually contribute a patch then your name is on the CVS. If you discuss the issue then your name is in the mailing list archive.
CN: So how does it work, because my name is in several places, how does it work that I see your name or you see my name? What happens? You think 'o that's not a very good patch" or o that's great she sends a bug report, or does it influence my reputation?
SA: Yes it does
CN: And is there any public discussion about it?
SA: yes, and not only public discussion. When you go to the CVS, you have different types of rights. What they call commit permission is given to people who have managed to establish their authenticity. I was talking about this hierarchy and this meritocracy that exists in a very tangible way. People who have contributed obviously have greater authenticity, people who have already successfully submitted patches in the past; their patches will be taken more seriously. So suppose a person who is reviewing patches, I will first go through the list and clear the people who have already submitted successful patches in the past. Only then you go through the second list of completely new patch submitters.
CN: And why is it dynamic? How does it change?
SA: Because, what happens in the software is we are trying to build software towards a deadline. And they have a concept of release date and on that day a particular version will be released and that version will usually have certain features. A list of new features or a list of what exists. And this feature list and this bugs list is usually publicized in advance to motivate everybody and to keep the community going. What will happen suddenly is somebody might be stuck with either an academic commitment or a professional commitment and they are not able to meet that deadline. And at that point you can see people getting very rude and saying how come you are not responding and then you loose whatever credibility you had. You can quickly loose your credibility and at that point people will say 'ok, we are not depending on this guy anymore'. So then on the list of features, if you go to an open source project, on the list of features and on the list of bugs fixes the names are there, who are the people responsible for it. So your name will get removed from that list etc. So in the next version if you propose a feature they will be less likely to take you because they don't want to be embarrassed. In many ways you can think of open source production as a performance act. Actually at one level the software is not so important, it's the audience that's important. If the open source people do not have a large audience consuming their software, then they will be less motivated to produce that software.
CN: So you actually say that the community thrives on emotions?
SA: and audience!
CN: and audience. Drama is also emotion, performance is drama, it's drama, it's performance, it's emotions, me you and all that.
SA: yes, yes, yes..
CN: So that's what makes it going. Actually you are explaining to me that it's all human emotions. So, and tracing human factors, how trust comes into existence.
CN: Lets talk about autonomous systems. It's the last ten minutes of our interview. Can you describe when do you trust a system? When there are no humans behind it?
SA: So I trust the system only when it gives weight to human behavior. So why do I trust Google? That's the biggest question. I trust Google because of the way Google gives credibility to specific pages. It is based on the number of real people and real other pages that have linked to this page. Google has sophisticated ways in which it prevents people from stuffing the search engine and Google bombing etc. so in some sense it is trying to deal with the abuse of such hyper linking system. But the fundamental core of the design is if my website is linked to many others, if many other websites link to my site then people trust the content on my website. So since the autonomous system is only consolidating human behavior, it is not adding machine intelligence in any way.
CN: But there is a problem in scale. And speed. Google works with speed and scale I do not perceive.
SA: But I am willing to accept it. So if for example, in online gambling, it's basically a transaction between players. So there is a fee, which the house charges and the system, all it does is it transfers money between the players. Finally someone wins the game, they take all the money and go. So at that point everybody knows that that person got to win. It is not like a slot machine. In a slot machine you are playing against the house, right? So, what I am saying is that a system which is just a mediation system, and - as with Google I can go to every single website in India and check those links and then I can go and see a new website which is not on that list and say 'how come all these links are not reflected in Google - there is a way for somebody to debunk the system at least a small sample, people may trust it. So in many ways I think people are geared towards just trusting such a system where at least at the smallest level they are able to validate that the system is only consolidating human behavior and not introducing new behavior.
CN: So what about the sketch that we have human actors and systems in a community. And for example a system like Google, it changes our language, we all Google, and also how we prepare for meetings, or how we read text, I mean Google actually has a huge behavioral impact on many of us. So it is in that sense definitely part of the community. Could you describe such an autonomous system? As an actor in it's own right? Because you say it only consolidates human behavior.
SA: No but then it's the whole business of reflexivity right? So you then contribute back into the system. I am saying people trust it. On the face it looks like it's consolidating human behavior. But of course it will modify the behavior itself. So now when we set up a Center of Internet and Society, one of the jobs was giving notice from Google. So we took effort to do that. And that was artificial. We actually did things which may not be considered absolutely fare in that sense. So I believe that finally the consolidated human behavior through the machine is not the same as pure consolidated human behavior. It's much more then that, in a sense.
CN: there is something else eh,
SA: Yes there is something else.
CN: Do you know what it is?
SA: It is the machine becoming implicated in the business of authenticity and trust. So that is what it is.
CN: Please elaborate!
SA: It's like Facebook. It's like Microsoft spaces in China. So on Facebook, because people know that the founders are associated with CIA, there is a certain degree of holding back, that happens on Facebook, by at least some of it's users, because they know people are connected to CIA
CN: Facebook owners…
SA: yes, or they have worked for CIA in the past or something like that. So what happens in many ways, people will create groups on the social network system that is anti-Facebook itself. So you have so many anti-Facebook groups on Facebook. So there is the human nodes in the network try to undermine the credibility of the machine as well through processes like this. Sometimes it's very hard for the machine to establish the credibility in the first place.
CN: Do I understand correctly that you say that 'because you have the anti-Facebook on Facebook so a true authentic machine has it's pro's and con's. It can organize it's own defeat.
CN: So then it wins authenticity. Like human beings can move towards well-being and survival and can move to destruction and death. So the moment the machine can have that as well, it actually has authentic presence.
SA: Yes you have said it in a much more beautiful way, but I agree with what you say.
CN: It is very interesting because in my dissertation I take a stand, a position that presence is actually the sense that helps you move towards well being and survival. You get sensorial input, cognitive input, input from your body, so all level input. And sense of presence is that you can move towards well-being and survival, which also in evolution it's been very important. And now with this sort of face book you actually say 'the moment a machine has a choice to live or die it get's authentic presence. So a car is very good, because a car can be ruined, a car can be in a crash, so a car can be trusted as an autonomous system.
SA: yes. I am trying to link this back to Google. How can humans contribute to the undermining of Google's credibility? We are back to the question of trust.
CN: You said an autonomous system should be able to establish authenticity and for that reason it has also be able to loose it. What is the authenticity of an autonomous system?
SA: I think in many ways it is the access controlled ownership of the system, to what degree you can access what information, who controls it and who owns it etc. For example if you go to Yahoo, and if you take the case of the famous Chinese blogger who was arrested, so there was a system and in many ways you were able to work the system of sending mails without any human intervention. So it was an autonomous system in that sense. But then the Chinese government made a request for information and without following any legal process, Yahoo gave the information of the blogger to the Chinese Government and they arrested the blogger and put him in jail for many, many years. So that moment of crises where it becomes clear that it is not really Yahoo that controls the database. It is actually the Government that controls the database that changes the trust the average Chinese person will have in email etc.
CN: How do I know who owns a system?
SA: There are no straightforward answers to this. So when a system comes into being, often it's authenticity and credibility depends upon the other users of that system. So I will use Gmail and I will use Facebook, because my friends use it. So that is the kind of first position with which people enter into a system. But once a system has been built, it's moments of crises and moments of celebration in a sense then will move the credibility in one direction or the other. So people won't start with an absolute baseline, or a very accurate baseline. It will start with a kind of authenticity or credibility that is associated with brand or the community, but then moments of crises can radically change what people feel about it. So if more and more people become clear that Google is tracking the behaviour of other websites, especially the websites that runs the add services and linking it back to their behaviour owned Google websites itself. That moment can change things quite dramatically. People are able to prove that Google has been… At the moment they say the machine reads your email and automatically serve you advertisements. But if people now notice that their behaviour on other independent sites are also now being monitored by Google, then there will be fresh outcry
CN: So there is one big distinction when we have a community of actors, of human beings and autonomous systems, it's only human beings who can judge and who can give trust and not the system?
SA: No, but often human beings themselves based their decisions of trust on the system consolidation. So it can be a false impression of authenticity and the whole thing is linked into an incestuous cycle.
CN: Yes, so can you as a software designer, are there queues?
SA: Queues in what sense?
CN: Well for example, as a completely naive user, I imagine it would be really nice if I use a system I just get a little flag this is owned by this and this person and I want legal procedures, not in terms of agreement but in two lines. This is your danger, if you use this system. Or for example I fill in data somewhere, I would like to say, "these data can stay half a year in the data base and then they will vaporize". And they will be empty fields. I don't have that kind of control as a human being. Not on my data identity, not on the tools I use.
SA: Yes, I think again here you have to distinguish between two kinds of systems that people use. Systems that are not under their ownership or control and systems that are under their ownership and control. So for example, If I am piloting software, then I am downloading open source tool onto my machine, I know that I can trust that tool because it is open source and then I am using public asymmetry key based inscription to connect to other people on the Internet that are sharing the same file and therefore because the technology for this inscription has not been broken easily globally so far, I trust the inscription system, and I only need to trust those people to know they have the same file I am interested in. For that period of time I want to download it. So here you can see an example of trust that is mediated completely through open source software and open protocol of communication between independent actors on the Internet. So in a sense, once the link is connected and all these different machines are connected and the transferring is happening, that network behaves in almost an autonomous way because it's not controlled by any of the users. It just happens based on network reality, right? But if you look in another context like Facebook, here I don't have access to the server or I don't have access to the software, I can't scrutinize the software etc. So it's a different type of trust that I will have in that system.
CN: But then you could say, if trust is negotiated on every level the witnessing and the public auditing are very crucial. Or not?
SA: In some cases there is no public auditing. The last mile in my house, when some one comes to set up my wireless network, often there is no third party or validating that transaction. So in many ways you could say a lot of the trust on the Internet today is founded in ignorance. Because I don't know any better because I don't understand the technology from that fundamental level. I trust it and even with Google by just repeating the line repeatedly, Google is not evil, Google is not evil, many people just believe that at face value. So I think it comes back to the question of the research you are talking about and people trusting other people on a social network, in many cases it is ignorance, trust is actually founded on ignorance and perception rather than reality and analysis etc. right?
CN: Yes and at the same time because people like to trust, more trust happens. So that is also a very strange thing; the more people trust each other the more trust will be its result. So it is a very strange dynamic, trust. It is not one plus one is two. If we both trust we go immediately to four or eight. If you trust and I don't we immediately go down, we go below zero, immediately. So the witnessed presence and the trust that comes with it, is an exponential dynamic. It's what you say like in crises and celebration it goes, or it goes down. And it is really hard to chase that. But in negotiation on all those levels of which people are more and more aware, I think, and judicially my agent, I am responsible for what my agent does. Only two or three nightmare's have to happen and people don't want to have agents anymore. They don't want to be liable for a piece of software travelling around the world, making shit happen in various places, it's like your dog. You want to have an insurance for your dog potentially hitting a car, causing an accident. So the whole liability thing is also very complex.
SA: But sometimes you wonder because the whole MySpace photo Gallery, all the private photo galleries were linked and on Pirate way and there was this huge 7 GB 10 GB file which you could download and watch everybody's MySpace photo's. But somehow there didn't seem to be a serious enough crisis for people to loose faith in MySpace. People accepted it and moved on. So it was somehow not important.
CN: That person has a problem that he wants to do this.
SA: Yes. So it is very difficult… But I think this question of giving away you data online, comes back to the question of design of authenticity. So because these people are so aware that they exist only if they give data away, that they are used to giving data away and they see that as a process of establishing authenticity online.
CN: And off line as well.
SA: and offline as well. So that's how one exists on the Internet, like opening oneself up. Of course it's difficult for the older generation to handle that. And I don't for example do anything personal on Facebook. For me it's a business-networking tool. But I see some off the young people that I know
CN: Well your wife divorces you and marries you all the time... (laughing) So is there anything you would like to add?