Eric Schmidt - Part I: Transcript of secret meeting between Julian Assange and Google CEO Eric Schmidt lyrics

Eric Schmidt: Well do you want us to start eating? Julian Assange: Well, we can do both. ES: Yeah, is that ok? JA: So this is... what's the date? Lisa Shields: June 23rd JA: ...June 23rd. This is a recording between Julian Assange, Eric Schmidt and...? LS: Lisa Shields JA: ...Lisa Shields. To be used in a book by Eric Schmidt, due to be published by Knopf in October 2012. I have been given a guarantee that I will see the transcript and will be able to adjust it for accuracy and clarity. ES: Can we start... I want to talk a little about Thor. Right. The sort of, the whole Navy network and... JA: Tor or Thor? ES: Yeah, actually I mean Tor. Uh... JA: And Odin as well. ES: That's right, sorry. Tor, uh, and the Navy network, and I don't actually understand how all of that worked. And the reason I'm mentioning this is I'm...I'm fundamentally interested in what happens with that technology as it evolves. Right. And so, the problem I would a**ert, is that if you're trying to receive data you need to have a guarantee of anonymity to the sender, you need to have a secure channel to the recipient, the recipient needs to be replicated, you know... What I'd like you to do is if you could just talk a bit about that architecture, what you did in WikiLeaks technically, you know, with the sort of the technical innovations that were needed and maybe also what happens. You know, how does it evolve? Technology always evolves. JA: Let me first frame this. I looked at something that I had seen going on with the world. Which is that I thought there were too many unjust acts. ES: OK JA: And I wanted there to be more just acts, and fewer unjust acts. And one can sort of say, well what are your philosophical axioms for this? And I say I do not need to consider them. This is simply my temperament. And it is an axiom because it is that way. And so that avoids, then, getting into further unhelpful discussions about why you want to do something. It is enough that I do. So in considering how unjust acts are caused and what tends to promote them and what promotes just acts I saw that human beings are basically invariant. That is that their inclinations and biological temperament haven't changed much over thousands of years and so therefore the only playing field left is: what do they have? And what do they know? And "have" is something that is fairly hard to influence, so that is what resources do they have at their disposal? And how much energy they can harness, and what are the supplies and so on. But what they know can be affected in a nonlinear way because when one person conveys information to another they can convey on to another and another and so on in a way that nonlinear and so you can affect a lot of people with a small amount of information. And therefore you can change the behaviour of many people with a small amount of information. So the question then arises as to what kinds of information will produce behaviour which is just? And disincentivise behaviour which is unjust? So all around the world there are people observing different parts of what is happening to them locally. And there are other people that are receiving information that they haven't observed first hand. And in the middle there are people who are involved in moving information from the observers to the people who will act on information. These are three separate problems that are all coupled together. I felt that there was a difficulty in taking observations and putting them in an efficient way into a distribution system which could then get this information to people who could act upon it. And so you can argue that companies like Google are involved, for example, in this "middle" business of taking... of moving information from people who have it to people who want it. The problem I saw was that this first step was crippled. And often the last step as well when it came to information that governments were inclined to censor. We can look at this whole process as the Fourth Estate. Or just as produced by the Fourth Estate. And so you have some kind of... pipeline... and... So I have this description which is... which is partly derived from my experiences in quantum mechanics about looking at the flow of particular types of information which will effect some change in the end. The bottleneck to me appeared to me to be primarily in the acquisition of information that would go on to produce changes that were just. In a Fourth Estate context the people who acquire information are sources. People who work information and distribute it are journalists and publishers. And people who act on it... is everyone. So that's a high level construct, but of course it then comes down to practically how do you engineer a system that solves that problem? And not just a technical system, but a total system. So WikiLeaks was and is an attempt - although still very young - at a total system. ES: For all three phases? JA: To deal with... not for all three phases but for the political component, the philosophical component and the engineering component in pushing out first component. Politically that means anonymizing and protecting... Sorry. Technically that means anonymizing and protecting sources in a wide variety of ways. Politically that also means protecting them politically, and incentivizing them in a political manner. Saying that their work is valuable, and encouraging people to take it up. And then there is also a legal aspect. What are the best laws that can be created in the best jurisdictions to operate this sort of stuff from? And practical everyday legal defense. On the technical front, our first prototype was engineered for a very adverse situation where publishing would be extremely difficult and our only effective defense in publishing would be anonymity. Where sourcing is difficult. As it still currently is for the national security sector. And where internally we had a very small and completely trusted team.