 Obviously I'm honored to be here, so thank you to Julia and Dave for organizing this and Lionel for having invited me. It is two projects that I'm working on and because they're our current graduate students in the audience, I want to put them in context just perhaps to show you at least in my experience how sometimes ideas for papers or texts come about. And indeed they're in very different stages. I have circulated. There's a paper copy of the current iteration of the cloud piece, so please feel free to take one. There are incomplete bits in it, incomplete footnotes and that sort of thing. I'm still revising it, but it's there if you wish to comment. Please do so. I'm always glad to take your comments. So that particular piece, the cloud piece is in a more final stage. In fact, I hope to have it done very soon for the purposes of submitting it to a journal, in part because the topic is rather hot at the moment. The other project is a more long-term project that actually came from my default thesis as it's evolving. And so two very different kinds of work, two very different kinds of projects. One of them is very much a long-term project, the virtue piece, part of the book project. The cloud project is a side project that came up as a result of teaching a course last September in the law faculty at McGill. So very different origins and perhaps illustrative of how things evolve as you're teaching or conducting a career as an academic. So the first piece is the cloud. I guess as I had said, I was teaching a course in September about the time that Lionel Bentley visited us. In fact, it was towards the end of that week that Lionel arrived. It was called Analog Copyright, and it was a one-week intensive course taught with a fellow who produced music. A lot of music that I like including Black Sabbath in the Clash and Blue Oyster Cult. Maybe you wouldn't put Blue Oyster Cult in the same category as the Clash. I wouldn't either, but in any event, we had a number of exchanges on the cloud with our students. What is the cloud? Well, first we need to look at what the internet is. This is a picture that I took off of a website on the internet. The idea I want to convey to you is that the internet, as we all know, is a linkage of computers and servers. It virtually and physically, I suppose, circumnavigates the globe. All these servers, computers, link together either with wires or wirelessly. And there are a number of metaphors that I think are useful in understanding how this original internet worked. Now these are ideal types, so a few caveats. These are ideal types with respect to understanding the internet metaphorically. Second, we are still, for the most part, here. We're still in this internet age, if you will. I think we're moving towards an age that we can call the cloud, or the clouds, with what I will argue are a very different set of metaphors. But to some extent we have aspects of both, as they're ideal types or pointers, we're pointing more towards the original internet, as I speak. But it's changing rapidly. The metaphors that I want to deliver to you, first of all, based on that original idea of an internet being linked around the world, is the first metaphor, first of all, is horizontality. In a sense, all of the various links are on par with each other as they go around the world. If you were to send information across the internet, if we were exchanging a file, as I did this morning with Gaynor, that file would take a number of different paths in order to go through from my laptop, through the connection here in Cambridge, back to my server at McGill, and then back here. The file would be broken down into packets, and the various packets would take different paths to go back and forth across the ocean, and get here, and then would be reassembled into the document as I had sent it, for the purposes of downloading and copying. The idea of horizontality is, then in part, the idea that there isn't one path for the information to follow, very much multiple paths, and therefore very difficult to control any one path. Governments that try to set up barriers, for example, can do so to some extent for some period of time, but there will always be a way to work around it, if people really work at it, in terms of trying to get around barriers. The second metaphor is that the internet was meant to be open. Lawrence Lessig points out in his book, Code, that when the original programmers of HTTP and HTML had to decide on whether to use an open code or a closed code in working out these protocols, they opted for open code, which meant that the internet could expand rapidly, anybody could add to it. They built it to be open in terms of its architecture, such that anybody could add, and people did add to it. Brett Frishman, in his very recent book, Infrastructure, calls this end-to-end, the end-to-end architecture of the internet. So all the way from the hardware of the internet, the physical infrastructure, down through standards, the very standards that make it operate, to programs that we use, to the content that we put up, and then to social networking sites, was meant to have this open aspect to it, in terms of its very conception. That too is important, in the sense that we're trying, again, along the lines with horizontality, we built something that was meant to be open for, it was like a building that you could add on to, at any given point. Third metaphor is anarchy, this is obviously a bit overstated, but in the early days of the internet, people like John Perry Barlow were arguing that this was a lawless zone, John Perry Barlow was a lyricist for the Grateful Dead, and the first president of the Electronic Frontier Foundation. So anarchy is a strong word here. I think what I wanted to represent is the idea that there's a decentralization of power here, with respect to, again, the horizontality and the openness. So the internet wasn't meant to be controllable, in a sense, in a way that perhaps other aspects of governance that we deal with as lawyers, as legal academics, or think about as lawyers, as legal academics, would be controllable. Partly true because of the technology, it moves so rapidly that oftentimes the normative response is a step behind, or two steps behind, with respect to the WIPO copyright treaties permanently behind. So again, this idea that there is some flexibility here in terms of the way the internet operates and the normative responses to how it operates. Fourth point, robust interaction with the internet. We, and this is still true now, or up until now, have tended to interact with the internet with relatively powerful computing devices, desktops, laptop computers. And for people that have the capability or the knowledge, they can do wonderful things programming on their own. They can, to a large extent, set the parameters for their own interaction with the internet. What Cori Doctorow or Jonathan Zittrain would call general purpose computing, and this power to interact, what Jonathan Zittrain would say would be that the power to create or to be generative on the internet, to have a generative internet. Finally, fourth, fifth metaphor, and this is, I think, the most controversial one. Again, it's an ideal type, and I don't pretend that this covers the picture completely. The metaphor of sharing, and I use it widely to mean, not just file sharing, but I use it widely to mean going on the internet, taking an object, downloading it, paying for it, or not paying for it as the case may be, depending on the business model or one inclination towards file sharing, as well as the legal norms that govern it. Making one's own copy on one's own hard drive, and then using it, perhaps, making it available to other people. So that's part of it, this idea of downloading, keeping a physical copy, sharing it. Also the idea of the software being shared and movements towards open source software. So taking those initial inclinations of the original programmers, and taking it one step further by creating kinds of software operating systems, as well as applications that could be used, modified, taken by other people, etc. So an important ethos. I'm not crazy enough to claim that it's the only ethos. It's a controversial claim to say that it's the dominant ethos. I'd be willing to defend that, but I also see, quite realistically, the weaknesses of that argument with respect to that fifth point. That's the internet I think that we've come to know up until now. Now, I think it's changing. It's changing with this system that we call the cloud. Basically the cloud is the idea that we delink the functions or that we would like our computer to perform for us, such as storage or computing capacity. We delink it from our own computer and we allow some other computer somewhere out there on the internet to perform that function for us. And we have no idea where that computer is. We have no idea if it's even one computer. It could be many. So that's the cloud. That's the operations that take place somewhere in the non-physical world of our own computer. It's obviously stored in a physical world somewhere. We just don't know where it is. It's probably in a Northern climate. It's probably in the UK or Norway or Iceland. But it certainly isn't somewhere that we can identify. The cloud came about because various internet players performing various functions, so sales, Amazon, research and information services like Google, realized that they had a lot of excess capacity and therefore decided to make this excess capacity available to people in order to make some money. Fairly well understood impulse in the market economies that we live in. And so they provided storage service and computing service and this began to move forward. Through an idea known as mirroring and virtualization, your computer gets tricked. Your laptop or whatever computer you're working on gets tricked into thinking it's dealing with computer X somewhere uniquely. But that isn't necessarily the case. It could be computers X, Y and Z. And you have no idea where they are. There's an exceptional amount of potential in this model. I'm going to make some comments in a moment that are skeptical about the cloud. And my goal here is to make you worry about them. But that's not to say that I'm not a great believer in the cloud. I would like to see this work. I think the capacity here, the potential here is limitless. I just think we need to open our eyes a little bit and be careful to, I think, give our movement to the cloud some direction to prevent what I think might be an era of increasing control over users. So far so good. There's a bit more technical descriptions in the paper if you need them, as well as footnotes to other texts that also describe the technicalities of the cloud. For my purposes, I think we're okay with just this basic idea of computing capacity out there somewhere that we have access to through the usual internet connections that we already have. I'm going to put you that the metaphors are changing. We're moving in this direction. We're not here yet, but I think already we can see a number of different trends. First of all, as opposed to horizontality, we're seeing a more vertical model, a more hierarchical model. In a sense, you have to move up to the cloud in order to have access to those services. And you move up into a cloud, it could be Amazon's cloud, or about half a dozen. It could be Microsoft's cloud, it could be Google's cloud, it could be HP's cloud. It's four of the six, there are a couple of others that escaped me right now. But you are moving up into that area, and you need to have access to get into that area. It's not something like the internet where you could find a different way to get in. You could find a back door, a side door, or your packets would move everywhere. Second, the architecture, the protocols of the cloud are not open in the way that the base internet protocols were open. The phenomenon here is one that's been called the wall of garden. Unfortunately, the real groundbreaking has been Apple, right? Developing, and I'm guilty of owning one, developing systems that are closed, that don't inter-operate with other systems. Meaning you have to, you know, if you want to buy into, you want to use Apple, you've got to buy into Apple. You have an Apple computer or Mac, you have the iPad or the iPhone or whatever, right? They work well amongst each other, and they present a wonderful degree of simplicity for the user, as between the various objects. And then you go on to their software applications like iTunes and that sort of thing. But you do have to get into their system in order to use it. And the same is true of their cloud. If you want to get into Apple's cloud, you have to use Apple. And it doesn't, it is important that that is important, the content is important from cloud to cloud, and indeed the hardware is inter-operable. The cloud makes it easier to do that, in a sense, right? You want to use Amazon's cloud and buy into Amazon's service. You're going to use Amazon's service, don't even try to move to Apple. So there's a great deal of potential here for closed systems to become the dominant model, in a way that they haven't been yet with the internet. It's true that Apple has always been there, but for the most part there are other systems that have been much more open, including open source systems, which I think work quite well, but even other systems, the whole PC system was a whole lot more open than the Apple system in terms of its interoperability. That creates the possibility for the third metaphor of control. You can have a degree of control at these nodal points as you move upwards into the cloud, that I think just wasn't possible on the internet. Yes, yes, you could control aspects of the old internet governments tried, others tried, but for the most part those were exceptions and not the rule. I think with the cloud control will be the rule and not the exceptions. Fourth metaphor, you'll recall I mentioned robust interaction with the internet, with the kinds of computing devices that we were using. That's already changing. We're more likely to interact with the internet with something like this, or an iPad increasingly than we are with a laptop. This is something that Jonathan Zittrain called the thin client. With an iPhone you become, I think more of a content taker than a content generator, but you also become beholden to the software that's used on that iPhone. It's very difficult. Apple has conceived it as a whole to make it easier for the user interface as between various products and software, but it means it's hard to change. It's very hard for you to get beyond the barrier in order to program things on your own. That's true with all of these various small devices. It limits what you can do. You really have to be good to get around it. You have to arguably put more of an effort into getting around it to do any programming on it, or to program it to make it do what you want it to do, then you could with a laptop. Again, Dr. Rowe and others point out that a laptop is a very powerful device for allowing you to generate what you want to generate. These other things are still powerful. Don't get me wrong, an iPad is still pretty powerful. You are conditioned to use it, you're controlled to use it in a certain way, and you have to be so much more adept at the technology in order to not use it in the way that Apple wants you to use it. This means that we become less democratic in terms of the internet. The ability to control from above, even as you have the device, and I think of the famous example with the Kindle eBooks in which George Orwell's 1984, quite appropriately, was pulled off everybody's eBooks because they hadn't quite worked out the copyright. So Amazon pulled it, even though people had already paid for it, so it just disappeared from your Kindle. They had to put it back. There was an app drawer, and Amazon worked out pretty quickly with the copyright holders what they needed to work out in terms of licensing agreement. But that goes to show you that these devices can be controlled from above in a way that was much more difficult to do when one was dealing with desktops and laptops. Finally, the idea that we are now streaming more than share. Again, this is meant to be an ideal type, but the picture that I want to convey to you is that stuff is coming down and we're taking it. We're no longer saving. You listen to music on Spotify or Rhapsody, depending on what part of the world you're in, and you can access all the music that you want. You pay for the service, but you don't ever make a copy. You can't make a copy. You move very much into a model of buying services, and there are technical terms, software as a service, infrastructure as a service, etc., a program as a service. They're explained in the text. But the idea is here, you become a contractual purchaser of services on the internet, as opposed to the old way of doing things which was to make your own physical copy onto your hard drive, and then you could do whatever you want with it. Totally, I think this may be a generational thing. I speak to my kids and nieces and nephews, and this is fine. Why would you want to have a copy? I think I'm still part of a generation that still like to have a copy on my own in case I can't get hooked up to the internet. You can always get hooked up to the internet. What if you can't? These are actual discussions that I've had with people 20 years younger than me who don't seem to see this as a problem. Maybe it's just my own bias towards having some kind of object, not necessarily physical. It can be a file on my computer, but I still like to have it. The sum total of this is, I think, we may be moving towards a kind of enclosure. We owe the vocabulary of enclosure to Carl Polany on the left, the Hungarian, Austro-Hungarian or Hungarian-Canadian political philosopher. A lot of people don't know that he actually lived in Canada while he was teaching at Columbia. So all the years that he spent in North America, he lived in Canada because his wife was a card-carrying communist and therefore couldn't get into the United States during the post-war period and McCarthy period and all the rest of it. So he taught at Columbia commuting from Toronto and had a capital influence on a large group of Canadian political economists at the University of Toronto. He famously talked about the enclosure movement here in England as the great transformation when lands were taken from a fairly complex normative system of commons and enclosed to allow for the industrial revolution to take place in England. James Boyle, who's on the right, used the metaphor as a second enclosure movement to describe attempts by copyright holders in particular on the internet through various kinds of rules that we see in the Infosoc Directive and the Visual Money Copyright Act to try to sort of add to their power, expand the power of copyright if you will into what's been called paracopierite or copyright plus controlling access on the internet, et cetera. My argument is we're making a quantum leap even from that in this regard because here with the architecture I believe is so much more controllable and the potential for control is greater, more effective than that second enclosure movement that Boyle was describing with respect to copyright holders. Many of those measures just won't work or can't work to some extent, maybe they have in some places, but certainly the idea that I want to convey to you is that if we are moving towards changing all these metaphors the potential for enclosure becomes much greater. So that's the fear that I want to strike into your hearts. Yes, there are privacy issues and a number of our intellectual property colleagues, privacy colleagues in both in Europe and in Canada, so people like Daniel Gervais now in the United States or Pierre Lagouche here in Europe have spoken about the privacy issues and written about the privacy issues. I share that. I think we do have a challenge there with respect to privacy, particularly in an age where we seem to care as lawyers about privacy and our 15-year-old kids don't seem to care when they're putting stuff up on Facebook. So there's a disconnect between what users feel they need to have in terms of privacy and what we feel we need to have in terms of it's because we're older or more mature because we're lawyers, but I share those various concerns, but my concern goes further. I'm concerned with the architecture and if we don't work at keeping this architecture open we could fail to realize the potential of the cloud as has been described by a variety of different people. This is obviously a very political question. It's not necessarily about IP as such. The original model of the Electronic Frontier Foundation was architecture's politics. It's now architecture's policy. I agree with Lawrence Lessig that I think the first model was better. So we need to think about the architecture of the cloud in another way. I propose that we think about it as a commons. Along the lines of a knowledge commons as Eleanor Oostrom and Charlotte Hess have proposed, I can flip through this quickly because we do have the other paper to talk about. I think we should think about the architecture of the cloud as a commons, so as something that we share and therefore we need to keep open. Such that private activity can go on in the cloud. I have no qualms with people making profit on the internet with a variety of different kinds of business models. That's fine by me. But I do think we need to keep the infrastructure open for others who want to use it in other ways and in particular to keep channels of communication open as between people. It's a flow resource as again Oostrom's work develops. Flow resources have a value that is a public value and needs to be shared and therefore we need to maintain I suppose open tabs in order to keep the flow of this kind of resource moving. It's a common good. This is something that's going on in Italy right now this whole discussion. Brett Frishman proposes infrastructure as a commons and I would tend to agree with him on this that we need to think about the net infrastructure as a kind of mixed semi commons. And this he's done in a number of articles but also in his book which just appeared about three weeks ago or four weeks ago. If we don't, and this is Frishman, private infrastructure owners have a number of reasons for choosing to reject a commons management strategy such as opportunities to price discriminate, vertically integrate and operate for a subset of downstream markets control future progress. That's exactly the fear that I've tried to describe in those metaphors that we have that we've just gone through. What might be solutions? I don't want to be completely negative. I think we need to try to make the cloud as, keep the cloud as accessible as possible. Beware terminology. When someone talks about the public cloud they mean Google, Amazon, Apple. Cloud that the general public can access even though they are privately held. They're owned by private entities. So I call that the privately delivered public cloud. So we need to try to keep that open as much as possible. I also think we need to try to work on a publicly delivered public cloud that will be there as a safety valve, if you will, to keep the privately delivered public cloud accountable. What do I mean by that? Well, with respect to the private sector and the privately delivered public cloud we have the option of competition law and incentives to try to keep moral persuasion, to try to keep technology interoperable, to try to make sure that there's no price gouging or collusion as between the five or six major entities that are operating cloud services. We could use other kinds of incentives to keep it open. So the usual things that we use with respect to private actors we've done it already with Microsoft, we've done it with others, sometimes with success, sometimes with less success. But in general we need to do that to keep the privately delivered public cloud as open as we can. I also think we need to have a publicly delivered cloud that at the very least provides services to people who might not be able to afford the services of these other private entities. So it would require at least some countries and here I think we have to push countries in Northern Climates in particular. One of the main challenges of providing cloud services is cooling, literally, and hence there's a great deal of cloud infrastructure that is moving towards places like the UK, like Norway, hope to think like Canada, like Iceland. And therefore I would hope that some of these countries would, using perhaps their universities, providing resources to the universities, computer science departments or whatever, to develop the structure and then maintain it. Now obviously this is a financial burden but my sense is it doesn't have to be all that big. It just has to be there. It has to be there to provide the service for people who otherwise can't afford it and to keep the private entities honest, if you will, make them accountable because if there is this publicly delivered option, chances are they will react by being more open themselves. So my view it doesn't have to be a massive cloud. It just has to be big enough to provide access to anyone who cannot or doesn't wish to use a privately delivered cloud service. I'll conclude with Lon Fuller, as I like to do. And you'll recall that in the morality of law, the first part of the morality of law, Fuller talked about law's aspirational morality or its morality of aspiration. But in the second part of the book, when he was trying to articulate the substance of natural law, he couldn't. The best that he could do was focus on what were effectively procedural principles. And the argument was, well, if we work on the procedure, hopefully the substance will work itself out in the long run. And that's true. I learned law in an environment where we thought Fuller won one of the heart-fuller debates and I appreciate that that would be different here. My own view is that heart won all the little battles, but Fuller won the larger point because at the end heart couldn't escape the idea that while there was no necessary connection between law and morality, there were so many contingent connections that in practice it was the case. Now, I presented this a couple of weeks ago and Kevin Gray was at a conference and pointed out to me that there was even a stronger argument in the second part of the morality of law with respect to what I was arguing. So I'm going to point out to you something that Kevin has pointed out to me. When asked to name one central, indisputable principle of substantive natural law, natural law of capital letters, what was Fuller's response? To open up, maintain and preserve the integrity of channels of communication by which men convey to one another what they perceive, feel and desire. If we were forced to select the principle that supports and infuses all human aspiration, we would find it in maintaining communication with our fellows. Communication is something more than a means of staying alive, it is a way of being alive. Even Fuller was pressured in that respect to have identified communication as being so central to the way we perceive, feel and desire. And I can't really say it any better. If we stress with the cloud communication and the ability to communicate, I think we can make the cloud live up to its potential and not allow it to slide into something that becomes another round of enclosure. So that's the first one.