Eben Moglen's Keynote Address at the Open World Forum, October 2010

This is an English transcription of Mr. Eben Moglen's speech at the Open World Forum, October 2010. Mr. Moglen is the Chairman of the Software Freedom Law Center.


The transcription is made by Răzvan Sandu, directly from the English video available here.

Keynote Address

EbenMoglen, Chairman of Software Freedom Law Center 

Open World Forum, Paris, October 2010 

It's a great pleasure to be here. I want to thank the organizers particularly for allowing me to compete with your cocktails. I guess, given that situation at the end of a long day, the best I can do is to talk about something completely different - so I'm going to talk about Freedom. I didn't hear it mentioned much today, so perhaps it represents a useful change of subject.

The purpose of the software which everybody has been talking about, seen by the most of the people who made it, was to create Freedom. It was the consequence of the existence of the Free Software Movement – FOSSIL, apparently, if you have been listening to today's proceedings – but, actually, I think, a more or less living organism. The Free Software Movement is a conspiracy to save the 21th century from itself, generated in the last quarter of the 20th century by a series of people who worried about the nature of technology free of ethical an social imperatives for the securing of liberty. The purpose of the Free Software Movement can be defined fairly simply, as follows: ''Make technology that supports Freedom, through sharing. Put that software in everything. Turn Freedom on.". So far, we have completed stage one of that process: the software intended to support freedom now runs everywhere. When I was a young programmer, in the 1970s, software engineering had a single overarching goal, which was summed up in a short phrase: "Write once, run everywhere". Nobody did it. We did it. Freedom now runs on everything, from the smallest object in your pocket to the largest cluster of computers in the world, whichever one you're talking about. Indeed, neither the stuff in your pocket nor the largest computers in the world can run on anything else reliably and with the kind of portability of software that was the goal of software engineering when I was young. We achieved "Write once, run everywhere" because we were tightly focused on a single goal: "Make software that supports Freedom. Put that software in everything. Turn Freedom on".

Now, we have achieved, as I say, the goal of stage one: we have built software that sustains freedom. It sustains freedom in several important respects: first, it supports sharing. We used that support for sharing – predominantly, the GNU General Public License and the other arrangements that protected freedom to share – we used that legal and social technology to destroy a malevolent monopoly whose intention was precisely the opposite: "Put unfreedom everywhere. Charge royalties.". That monopoly is now failing and will soon be gone. That's happening because freedom runs everywhere and is available everywhere to anyone who wants it, at no cost, to copy, modify and improve. That process protects itself against appropriation, it protects itself against unfreedom manufacture out of free parts and it does what it can to prevent the rent seeking buried in the patent system by people who unwisely applied the patent system to software, in order to enable rent seeking. However, we have, in the process, created a network whose possibilities for unfreedom are as difficult and as threatening as the problems in the network we are seeking to prevent. We have created a platform which allows free culture: Wikipedia exists, Creative Commons exists, sharing knowledge is now possible all the way to the bottom of opportunity levels in every human society, because the mobile telephone is there and were it not for the rules of so-called Intellectual Property, we could deliver every piece of knowledge, every kind of art, every thing of human value to every human brain. In the next generation, we will make a determined attempt to allow every human brain on the planet to learn. For the first time in the history of the human race, we will make a determined attempt to prevent every single human brain from being starved of the knowledge that it seeks to acquire. If we succeed, we will change the destiny of the human race, as we always intended to do.

But we face a very significant problem, one we knew would develop over time, but which has developed faster and in a more threatening fashion that we expected and it is built of our parts. You now live in a network which surveils you more deeply, which knows more about you and which renders the knowledge about you more dangerously accessible to those who wish to modify your behavior than anyone of the 20th century political systems we refer to as totalitarian. Those systems could only surveil and observe your behavior and they could only do so to the extent that your behavior was accessible to other human beings who informed on you. The network now knows your thoughts, your plans, your question, your dreams, your fears, your uncertainties. It surveils every party you go to, every drink you have, every message you leave for a friend. Five hundred million human beings have decided to allow everything they do in their social lives, from day to day, to be collected in one big database. Run for profit, by a fool (applauses). If we were to allow that situation to continue for much longer, we would face a situation in which the very fundamental texture of human autonomy would begin to be adjusted by the network. What you search for, what you find, what you buy, when you decide to want something and what something you decide to want would begin to be thoughtfully shaped by machines operating at the directions of those who wish to take economic rents out of the modifiability of your behavior. Not to mention all the governments which can now, with a subpoena blank or an administrative order of some kind, seek from every third party who collects information about you, the information they require in order to interfere with, disrupt or forbid your activity. I don't need to explain to anyone who lives in France how power-hungry the establishment can become on those subjects.

Now we come (applauses), now we come to this unfortunate technical reality described by this cloudy name: The Cloud. The funny thing about the cloud is the people talk about it everywhere I go but, if you ask them to explain what it is they can't do it, because they don't know. So, I'm going to explain in two – I hope – fairly simple conceptions what the cloud is. First, the cloud is the server-client architecture of the Net, on steroids. And second, the cloud is the server-client architecture of the Net, on steroids, virtualized. The first part of it comes down to this: the Internet was constructed by researchers attempting to complete the task of architecting an entirely descentralised network consisting exclusively of peers. The goal in doing so was to allow arbitrary machines, in a network of arbitrary geometry, to survive arbitrary interferences by minor annoyances like nuclear war by routing all traffic in a completely non-hierarchical fashion without any regulation by masters over servants. This was reflected in the very form in which the network understood, at its beginning, the concept of a server and a client. Those of you who are custom to using only software produced by the dying monopoly may not understand this, but everyone who uses Free Software at least plain passively understands it, because your Free Software computer treats the server as the computer closest to you and the task of that server is to serve human interaction. That's how X-Window works, right ? The X server serves com processes in the network, all of which are assumed to be equal in hierarchical status, the opportunity to communicate with the human being - and so it serves the opportunity to get to your eyeball, your eardrum and your fingertips. In other words, the Net consisted of a non-hierarchical collection of processes, in which each human being had a server that served interaction with him. The basic technical achievement of the monopoly was to distort all of this completely. Windows was a degenerate version of the structure of the network originally architected by people who did not have the same neurological dysfunctions as the architects of the monopoly (laughs). The purpose of the server-client interaction structure in the Net became to create hierarchies among machines in which some were servers and most were clients. And human beings used clients to deal with servers which operated, with respect to that human beings, du haut en bas. We rebuilt the Net, over a period of almost 20 years, to prefigure this peculiarly hierarchical and statist conception of the nature of the Net, which was never any part of the original architecture, but was solely an artifact of the rent-seeking policies of the monopolist. Now, that hierarchical architecture of the Net is enabling the gathering of political power in the center, with respect to disempowered endpoints. And both manufacturers of proprietary software and the architects of proprietary networks absorb, as mother's milk, the business model opportunities created by a hierarchical Net, with power at the center and disempowered clients at the edge. Those are the preconditions for the cloud. The unfair, inappropriate, architecturally unnecessary, economically anticompetitive architecture given to the Net is the watervapour out of which the cloud condenses. All that is left to make the thing called The Cloud is to remove the relationship between iron and the server. This is accomplished by the uniform adoption of virtualization layers, many of which we wrote. The virtualization layers makes it possible for hierarchy to use the accumulations of power at the center evermore efficiently and evermore mysteriously with respect to the disempowered client at the edge. Taken to an extreme, that hierarchical network architecture allows the client at the edge to be completely disempowered. It receives what is known as Software As A Service - more appropriately, I think, labeled Software As A Disservice - that is to say you can't provide any services to the net, all you have is software given to you to execute by some master with a disembodied server somewhere. Moreover, and even more aggressively inappropriately, the cloud comes to be seen as “all your data, er, belong to us”. The virtualization of storage implies that the party at the edge of the net no longer exercises control over her own information, which is physically located at an arbitrary place chosen for her by the people who run The Cloud. Any legal obligations that may exist with respect to the control or management of her data has been lost neatly subverted. If there are, for example, rules concerning the nature of the regulation of the data in the cloud in a particular place, data will mysteriously tend to move, by means even legitimate or illegitimate, to locations where those regulations are weaker rather than stronger. The North American economy is slowly shifting thoughts to an economy based primarily on data mining, which might be thought of as exercise of control over the data representing the human lives of everyone on Earth. This is not a recipe for freedom. This is the biggest and most successful recipe for unfreedom that has ever been devised in history of human beings. We have come, at the moment we expected it to arrive, we have come to a faithful set of choices. The technology of digital computation, storage and control now offers us, at one and the same time, the greatest possible opportunity, namely the elimination of ignorance, forever, and the power to allow every human brain to learn and, on the other hand, the completion of a structure of social control, more sorrow, more complete, more technologically perfect than any that has ever existed in the history of human beings. Dearly beloved, I set before you a choice between life and death. Fortunately, the tools are still in our hands. We made this stuff and we can fix the problem.

As I have already suggested, the primary difficulties that we face arise from two bad technical decisions, both of them reversible. In the first place, we allowed the conception of the server as a form of centralized power divorced from the human being. It was not there in the original design of the Net, it was not there in the original design of Unix, it is an artifact of bad software, manufactured by people with ill will. We do not have to accept the idea that services are centralized. If we had an enormous number of cheap, reliable, low-power serving computers, running software designed to provide services to others on our behalf and storing our data in them, in safe locations, we could convert a network presently consisting of centralized services into a network consisting of federated services, that is services rendered by people to people, for people, in a non-hierarchical fashion. Second, we could reverse the virtualization epidemic, not for the purpose of eliminating virtualization or for preventing those who own heavy hardware for making better use of of it, but to prevent the idea that you ought not to have a computer of your own, you ought not to store your data on your own, you ought not to provide services to the world on your own, you ought to be a consumer of the services provided by the Net. So, do we have such a possibility for reversing the centralization of the networks architecture? Yes, we do.

Already, throughout the hardware environment of the world you will notice the beginnings of the distribution of the so-called "plug servers": tiny objects, requiring very little power, using comparatively low-power, low speed, but quite muscular little chips, mostly made by ARM. You can buy, at the moment, at prices indicative of the beginning of a hardware form-factors run, that is in the neighborhood of ninety to a hundred-fifty US dollars around the world, objects no larger than a cellphone charger, which contain a fully operational server with a wired NIC, wireless, some S-ATA ports, some USB, some serial console attachment possibilities and so on - in other words, a fully featured computer, capable of delivering services over the Net to anybody who wishes to interact with you. Those plug servers, which will soon be cheaper than cellphones – when they are made in appropriate quantities – those plug servers, supported by Free Software – that's us, folks, the people who believe in freedom! - supported by Free Software, will soon be providing a form of personal private cloud, which reverses the negative social consequences of power concentration in the Net. Your data, kept in your home or your office, securely backup-ed to the plug servers of your friends, capable of being read by nobody without a search warrant, protecting everything inside your apartment by intrusion-detection software which works, providing you free telephone calls over the Net through Asterisk, allowing your friends in China to climb over the Great Firewall using proxies servers several billion wide, many to many for the Chinese Communist Party to interfere with, affording you the opportunity to federate your social life rather than centralize it, moving you painlessly into a migration of Facebook, Flickr, Twitter and all the other bad things… You don't lose your friends, you all just leave, bit by bit, and, as you go, you cease to use the centralized servers, because you have a direct encrypted channel between yourself and each of your friends – in other words, we begin federating what the bad folks centralized. We do this in machinery which is so cheap, so ubiquitous, so low in its power demands that we no longer care about reliable electricity – one of those servers will run on 2 W bands. We will succeed – we WILL succeed – in providing freedom in a package so inexpensive, so reliable and so thoughtfully engineered by us, for us, that it will no longer be possible for people around the network to concentrate power, to concentrate data or to conduct meaningful surveillance of our individual lives. We will do this not soon, but now.

The hardware economics are not a problem for us – they're already solved. The software economics are not a problem for us – we solved those long ago. The technical problems are not a problem for us – we're good at that. That's how we got here! The political questions are not hard for us: we love freedom and we hate tyranny. We know exactly how to tell the difference. No matter what it requires a hoodie, or a spacesuit or a uniform. We know what we're doing. We gonna put freedom in new things. We gonna spread those things across the planet. We're gonna turn freedom on. So what I wanna tell you is: “It's OK! We're still here. We haven't gone away”. The Free Software Movement isn't dead, it hasn't become “open-”anything (applauses). We know exactly what we need to do and we gonna do it. We gonna make it and we gonna say: “Hey, here, we made this freedom. We think you may like it. Want some? Take it. It's free”. Wanna join up? Thank you very much!