Select 'Print' in your browser menu to print this document.

Feb. 28
Andersen Auditorium, Haas School of Business

8:45-10:30
DRM as an enabler of business models
Carl Shapiro, Haas School of Business, UC Berkeley (moderator)
David Reed, Cable Labs
Allan Adler, Association of American Publishers
Bob Blakley, IBM Corp.
Donald M. Whiteside, Intel Corp.
Cary Sherman, Recording Industry Association of America
Lon Sobel, Entertainment Law Reporter (paper on ISPs as digital retailers)
Sarah Deutsch, Verizon Communication

Carl Shapiro:
Welcome. Good morning. Pleased to have you all here, and I'm delighted to be able to moderate the panel. We have a variety of different perspectives on DRM. The title is "DRM as an Enabler of Business Models." I'm an economist here at the Haas School, so to me, ultimately - I've been listening to the technology, I've been listening to the legal side of it yesterday - but ultimately I would say, showing my economist stripes, that it's going to be about money and control of digital content. I think we'll see that from different perspectives from the panelists this morning. Again, from my perspective, when I think about the business models, I'm really thinking about the content owners and how are they going to generate revenue streams that are protected in the presence of lots of danger of unauthorized copying or access. And it's rather exciting to the economist to think about the slicing and dicing of content and renting it in micro-payments and so forth. And of course we have to deal with the distribution side as well as the content creation side and there is a lot of fear out there for those content creators whether they will get returns.

So there is one view maybe the returns will actually be higher because they can get greater control and a greater ability to price discriminate. Of course, the other view is they are going to lose control and have reduced supply of creative materials. I want to start with the first speaker, who is Lon Sobel from the Entertainment Law Reporter. I hope you have access…I know you have access to his paper "DRM as an Enabler of Business Models: ISPs and Digital Retailers." So I'm going to turn it over to Lon who is going to describe his paper. The session is not entirely structured around his paper. He, however, wrote a paper, and it's very interesting from my perspective, and other panelists will be free to comment or not on the paper but as always, will also be able to express their own views. But he's going to start with one business model. Rather than just say we need a new business model, he's got one for us to consider and let's start with that. Lon…

Lon Sobel:
Thank you. I'm going to begin this way: forgive me Jack Valenti and Cary Sherman for I have sinned. I've proposed the creation of yet another statutory copyright license, a copyright license in which copyright owners would be required to tolerate the unregulated distribution and redistribution over the internet of their works. Unregulated, but not uncompensated. This is one of the reasons that I'm hoping for forgiveness. An essential element of the suggestion that I've made and the model that I've proposed is that copyright owners would, in fact, be compensated for the redistribution of their materials over the internet and most importantly from my point of view, they would be compensated at royalty rates they themselves set, not royalty rates that were set on their behalf by law or by government panels. Let me first explain how this ties into digital rights management. The plan that I've proposed is one that takes advantage of digital rights management, and at least in our law, the part of digital rights management is referred to as copyright management information, in order to determine precisely what works are be distributed and redistributed over the internet, and who the ultimate customers - I'm going to call them customers - who the ultimate customers of those digital works in fact are.

The kinds of information would include not only the name of the copyright owner and, with respect to those works that have multiple copyright owners, like musical recordings, what interests each own, but most importantly what the royalty rate is for the distribution of that work at the moment that it is being redistributed. The royalty rate can be adjusted in my contemplation by copyright owners as time passes, almost whenever they want. Let me describe to you the essence of the plan and then explain how it fits into other uses of digital rights management for the creation of business models and then to tell you what I think the justifications for the plan are, and where even I acknowledge there are some problems that have to be dealt with. The essence of the plan is that internet service providers, those who have the last connection with retail customers, with us - this would include, for example, universities as ISPs as well as Verizon and AOL as ISPs - they would pay copyright owners royalty rates for the transmission through their system from the ISP server to the ultimate subscriber, if you will, at the royalty rates that are set by the copyright owners on a work-for-work basis. With respect to digital versions of works that are created by copyright owners after the plan goes into effect, copyright owners would know to include watermarks within the digital versions that they themselves create, if they're not creating watermarks already. And these watermarks would include all the essential pieces of information: who's to be paid and how much.

With respect to those works that are in existence already and don't have these kinds of watermarks, or with respect to works that when digitized had watermarks but were stripped of their watermarks because somebody captured the analog version just as it reached the loud speaker, just as it reached the computer monitor, and then re-digitized it for posting up on the internet, it's my belief, my understanding of the technology, that a fingerprinting system could be created in order to enable the ISP servers to identify those un-watermarked works as they flowed - and to the extent the utility of my plan depends upon your grasping that technology perfectly, I'll respond to that in questions - but I have been assured by companies that are in the fingerprinting business that if they are given even an analog version of a work, they can create from the analog version a unique digital identifier which I call a fingerprint, and that that fingerprint can then be used by ISPs to identify non-watermarked files the same way that an ISP server could be used to identify watermarked servers. So what would happen now is, people would be permitted even to rip MP3 files from CD's, would be permitted to post them up to their web sites where they could be downloaded, would be permitted to freely exchange them over p-to-p networks, all without any liability whatsoever because as the file reaches its ultimate recipient, its ultimate downloader, that downloader's receipt of it would be detected by the ISP, just the way our telephone companies now are able to detect when we make long-distance telephone calls.

And the ISP would log back that we have received this file, would note what the fee was just the way our long-distance telephone companies now note how long we talk and how far we're talking, and would compute the charges, and would monthly bill us along with what our monthly subscription fee is for all the copyrighted files that we have downloaded. Because we need to know how much we're going to be spending, we need to be informed in advance of the fact that there is some copyright-protected material coming to us over the website that we're visiting, or the emails that we're receiving. And I envision the little pop-up window that would give us that information the way I, at least, am now informed whenever I am about to download something that might be a virus-bearing file; that the thing that I'm about to download might have a virus in it and do I in fact want to download it, and, if so, where on my hard drive do I want to put it. So, that's the way in which the system would work. Let me explain roughly where this fits in and what some of the advantages of it are as I see it. I'm motivated by a couple of objectives. The first and foremost for me, given the nature of my practice, and where my former students are now working, is to get copyright owners paid for what they create at amounts that they decide among themselves.

I do have another objective, however, and it's an objective that I share with computer manufacturers and technology companies, and that objective is to keep the government out of the business of legislating the manner in which technology is designed. It's part of a broader view that I have that it's not a good idea for the government to be involved in legislating design features. It's not anything specific to technology. This broader view that Congress should stay out of the business of legislating design features is an attitude that makes me, for example, very disappointed and unhappy with what the House of Representatives did yesterday when it voted to legislatively ban all kinds of cloning research. So it's part of this overall view that I have that it's a bad idea - maybe it's because I'm old enough to remember the introduction of Post-it Notes, and know that Post-it Notes were an accident, and believe that a lot of wonderful technologies result not from a plan, but from serendipity, and that view that I have that Congress cannot see ahead years to know exactly what legislation should be.

So, this system would allow people to develop technologies - would allow computer manufacturers to manufacture computers however they like, consumer electronics companies to manufacture consumer electronics devices however they like without looking over at their shoulder at law, would allow p-to-p networks and those who provide the software to create p-to-p networks to do what they do without concern about legislation, would allow all of your fourteen-year-old children to have websites from which music and movies are downloadable without your concern about their being liable for doing it - and they wouldn't be liable for doing it because anybody can download it from your fourteen-year-old's website would be paying if they chose to do the download when they got their monthly ISP bill. Now, let me explain where this fits in. Right now you know that in the session of Congress that ended last Christmas there was something called the Hollings Bill which, had it been enacted, would have in fact required computer manufacturers and others to build into their devices circuitry that recognized the existence of certain DRM that told the device what, if anything, the recipient of that material could do. That bill wasn't even voted upon, let alone enacted. But in the current session of Congress, or as we meet today, the government communications division has before it a proceeding on broadcast flag proposals. And broadcast flag proposals, while not photocopied from the Hollings Bill, nevertheless, if described in a single sentence, would be described in a single sentence as doing the same thing the Hollings Bill did - would be described as requiring those who manufacture television receiver tuners to include within their tuners circuitry that would recognize the existence of broadcast flags, and would allow those who have the television sets and the video cassette or digital recorder circuitry to do only that with the television program what the owner of the copyright of the program is permitting to be done.

And the SEC is also put on the question of whether computer manufacturers and consumer electronics manufacturers should be required to include this same kind of broadcast flag recognition circuitry as has been suggested for TV set tuner manufacturers. That's an example of the use of DRM to enable a business model. But the business model that's been enabled is a business model that restricts what can be done by people with the equipment they have, and restricts the design of the equipment. I am suggesting, as I just did, that I have some concerns about the utility of that sort of thing - indeed, some serious concerns about the technical utility of it. And we saw a presentation yesterday by Microsoft that shows that Microsoft now has some design features in mind that might make these sorts of systems workable. On the other hand, just a couple of months ago, some researchers who were commissioned by Microsoft put out a report called "the Dark Net" that suggested that anything that resides on the end-user's equipment is capable of being circumvented: either quickly if it's in software implementation or more slowly if it's in hardware implementation. And if the hardware is perfect it means that the hardware will become dated more quickly than consumers expected, given the amount of money that they paid for it. So my sense of things is that, again, this shouldn't reside on the end-user's computer. And so instead what I have suggested is this system by which people will be paid - that is to say, content owners will be paid - and the internet service providers will be the retailers. Let me explain how I envision the internet service providers doing this, and why I am hopeful that it will be attractive to them. I am suggesting that internet service providers pay what I'll call wholesale: the royalties suggested - demanded, not suggested - the royalties demanded by copyright owners at whatever levels the copyright owners insist upon. But I am not suggesting that these be costs that be absorbed by the internet service providers. What I am imagining are that just the way retail stores today pay wholesale and then mark up what they've purchased for sale to their own customers, so ISPs will be privileged to mark up the copyrighted works that they have purchased for resale to their own customers. My plan does not purport tell ISPs how much of a markup they have to take, or indeed that they have to take one at all. Just the way in the world of retailing we have an enormous variety of business models, from the small boutique storefront in expensive neighborhoods where the markup might be a hundred percent or even more over wholesale before we get to retail on the one hand, to Cosco on the other hand and other big-box, warehouse-type retailers where the markup might be as little as twenty percent or even less, so I imagine that ISPs will develop their own best business model by which they mark things up as much as a hundred percent or more if they choose to. That wouldn't be surprising in the world of entertainment (I see some smiles in the audience). Bookstores mark up books a hundred percent. Record stores mark up records a hundred percent.

Videos are marked up approximately a hundred percent. By which I mean to say when we buy any of these things at retail, half of what we spend at retail is actually retained by the retailer, and only the other half goes to the publisher or the distributor. So, rule of thumb, what I expect is that ISPs will be marking things up a hundred percent: fifty percent of what they collect they would keep, the other fifty percent would go to the copyright owner, but that would be up to the ISPs. And I can in fact imagine a situation in which an ISP would choose to have very small markup in order to attract subscribers to its basic service to pay the monthly fee. Part of the benefit of all this is that this would give all of us that don't have broadband connections now through our ISPs an incentive to have a broadband connection even at fifty or sixty dollars a month because there would be a lot of valuable and rich content that we would now be able to download. As my time runs out, let me acknowledge - lest there are those of you in the audience that are thinking, "Oh, he hasn't thought of this, that, or the other" - a number of problems. One of the problems that I have anticipated is that if we think we have problems with spam toady, wait until people can expect to be paid for sending us copyrighted spam.

And I've acknowledged the existence of that problem, and there are both technological and non-technological solutions to the problem that I have described in my paper and would be happy to talk to you about during question and answer. The other problem that I anticipate, particularly among the privacy advocates, is that this would enable, indeed require ISPs to know everything that we've looked at or downloaded for viewing, and that that would be an intolerable invasion of privacy. My response to that is simply that, as measured from a baseline of complete privacy, it would be an enormous invasion of our privacy. But we live in the year 2003 when our credit card companies know where we buy things, what we paid for them, in what part of the world we were when we bought those things, and for any of you who, like me, have had the experience of traveling, from a distance from your home, going to London for the summer, you know, for example, that your telephone at home is ringing by somebody - live human being - from the credit card company that is saying, "Your credit card is being used in London. Is it you, or is it a thief?" So people know already, given the credit cards that we can't do business without, where we are and what we're doing. I'm suggesting that such invasion of privacy is likely required in order to implement the system I've described is a small, incremental loss of privacy. It is not as huge a loss of privacy as it might have been if we were starting from a position of absolute privacy. Thank you.

Carl Shapiro:
Thank you Lon. And I encourage all of you out there, if you have questions about Lon's proposal, to hold them for later. The panelists will address them as they see fit as they have their turn. The next speaker is Donald Whiteside who is Vice President, Legal and Governmental Affairs and Director of Strategic Programs Office at Intel. Don.

Don Whiteside:
Good morning everyone. First I wanted to thank Pam and the organizers of this conference for inviting Intel. First I want to take a minute to thank Pam and the organizers of this conference for inviting Intel to participate. As many of you know, Intel, as an innovative company, develops many of the technologies that deliver new and exciting digital media experiences. And our focus in this endeavor is certainly delivering these products that create new choice and new empowerment for consumers, and not to limit what consumers can do. The growth of our business depends on exploring new business models for digital media, and creating exciting new digital media experiences for consumers. As Carl highlighted, I work for Intel. I manage our government affairs organization focused on broadband and content, and in this effort I have responsibility for directing what we do from a technology perspective, a legal, a legislative, and a public policy perspective. I'm not going to go into great detail on specific things, such as broadcast flag, during the few minutes I have here, but I certainly would be more than welcome to answering questions during the Q & A session or later today or tomorrow. What I want to do is explain a little bit about the environment we're in. You know, this moniker: the internet changes everything. And I think we need to conceptualize that a little bit, and recognize that the internet is just one of a multitude of disruptive technologies that have come together over the recent past - whether it's exciting new encryption technology, file format, compression technology, CD burners, ubiquitous internet connection, powerful processors - all of these technologies are disruptive. And they're disruptive in many ways.

First of all, this panel is focused on how technology has disrupted business models. How it has created challenges for the existent business models and at the same time created opportunities for new business models. But I think we shouldn't lose sight that it's not just a business models discussion. Technology has really redefined how consumers use media. It changed their expectations. And a lot of this discussion yesterday centered around fair use. I think if you ask a consumer, "What is fair use," most of them don't know. What they know is what their use expectations are. They have a customary use expectation, and that expectation continues to evolve based on technology. "I can, therefore I should be able to" is a pervasive view in the market place. At the same time, as usage expectations evolve, so do some of the societal ethics and morality. What is right? What should I be able to do, irrespective of what the law says? So I think we need to recognize that this isn't just about a challenge to existing business models and how we move to a new business model. This is an even greater challenge for the media industry, the technology industry, and law makers. But this panel is talking about DRM as a business model enabler, so I'd like you to think about the fork in the road that Cary's customers, or Cary's representatives - the motion picture studios - are faced with. And that is that we're at a strategic inflection point created by these constructive technologies. And the common element at a strategic inflection point is that fundamental decision about which path do I take. If I'm the incumbent, I have an existing financial responsibility to my shareholders, I have an existing relationship with my distribution partners, and I have an existing relationship with my customers. And that's a business that I'm in.

It creates value for my company, it creates jobs for my employees, and it creates return for the shareholders. And we need to recognize that existing, incumbent media businesses have a responsibility to continue to milk the existing business model for as long as they can, and hopefully, have an opportunity then to invest in new business models. So I have great regard for the existing record labels and movie studios as they're challenged by technology, and I respect the fact that they have to preserve their existing business models as much as possible. But on the flip side, there are new business model opportunities. Technology creates a wealth of new business model choices. And I think you've heard one scenario of how a new business model could be created, and I suspect other panelists here will talk about others. The reality is that it's very difficult for the incumbents to embrace those new business models. It's the small, nimble companies that are going to be first to bridge the gap and explore the new business models. Yet it is the incumbents who should take that first step to preserve over the long haul their relationship with customers. And I think what we've seen over the past six months with the record labels is that they've realized that technology has changed everything, and there is a new business model or business models that need to be explored.

And I encourage the more broad licensing of music and the expansive growth of broad catalogue services such as Press Play, and AOL's recent introduction of Music Net. So while we're sitting here talking about potentially a world in the future where content flows freely and everyone is happy and new business models emerge, the reality is that it's not going to happen overnight. The incumbents need to move quicker, or lest they lose their relationship with consumers. I think in the music industry we're seeing some acceleration of embracement of new business models. And I think the movie industry is soon to face the same perilous decisions on whether to stick with the old or embrace the new. You know, this is a panel to talk about the role of DRM, and there's really two ways to look at DRM. One is DRM simply as a protection mechanism. And in business model environment, DRM can protect the existing business model. How do I make sure that no one has an opportunity, through content protection and DRM, to explore something new? I don't think the predominance of technology companies investing in content protection or DRM technology view the technology investment they're making in that way. Intel certainly doesn't. Our focus is on protection technologies in DRM that enable new choices, new users, and new control for consumers. The creation of a protected environment should enable an opportunity for consumers to have liberal copying rights, liberal, flexible use of content, and the technology should really focus on the fringes of where usage crosses over into copyright infringement - in copyright infringement primarily from a redistribution standpoint. As many of you know, Intel has been involved in the DVD CCA activities around content scramble system. We're involved in the 5C development of ATCP, as well as the 4C Entity. We also participate in the broadcast flag development, as well as the new group focused on analog reconversion.

The common element in all of these is that those forums are not government led forums; they're not content industry led forums, and they're not simply technology group forums. They're cross-industry efforts looking at industry challenges and developing solutions in the private sector to create new opportunities for all constituents. And Intel continues to be focused on developing solutions that create value for consumers and new opportunities for content owners, product vendors, and distribution parties. And doing that without the intervention of the government. So, I think I'm at the end of my time. I'd be more than happy to answer questions either later on this panel or later today or tomorrow morning.

Carl Shapiro:
Thank you, Don. Our next speaker is Cary Sherman, who is the President and General Counsel of the Recording Industry Association of America. I'm sure many of you realize the RIAA has been right at the center of a lot of these issues within the music industry, and I'm really looking forward to hearing how they have a forward-looking view with the opportunities here, rather than really threats.

Audience:
Laughter.

Cary Sherman:
Thank you, Pam.

Audience:
Laughter.

Cary Sherman:
When I spoke to Carl the other day about what the subject was going to be, it became clear that he wanted me to address more than just DRM as business enablers, which I think we really feel very strongly is the fact, but rather to give you an overview of what's going on. Specifically he asked: what are we facing (you know, what are the constraints we're facing), what are the record company attitudes towards these copying issues, and what are the companies doing? So I have ten minutes to summarize everything that's happened, and I'll try and do it as quickly as I can, but I'm going to speak fast.

Carl Shapiro:
You actually answered my questions. I grant you two extra minutes.

Audience:
Laughter.

Cary Sherman:
Okay, I'm going to take them then. First, what are we facing? I'm trying to give you the business and market place realities of what's going on as candidly as I can. One thing we're facing is a need to continue selling CDs. The fact is, there are somewhere between eight-hundred million and a billion CD players in the market place. We can't just stop selling them. We can't suddenly encrypt them in order to protect the content, in order to protect against internet piracy. All those CD players would become obsolete; our market would disappear. So we basically have to continue selling these CDs while we transition to new models. Which means that if there are new formats on the horizon, we have to realize that the introduction is going to take a long time. It will be years of transition until a newly-installed base of players is in the market place. And that's true whether it's physical product or online delivery. In the physical area we have a couple of new formats: DVD Audio and SACD. Unfortunately, we have a format war between those two. RIAA certainly can't get into the resolution of that for anti-trust reasons.

We have competitive forces at work, so that the companies have different views on the different technologies, and there are different deals being struck between hardware manufacturers and record companies. So, even there, you have trouble agreeing on a new format that might help migrate consumers back into record stores. Online? Let's face it - this is a very tiny market place. It's a huge market place for free, but it is a tiny market place for buying. And there's a generation of consumers right now who think music ought to be free. That's what we're facing. Which means it's going to be difficult to migrate them to legitimate services no matter how appealing the content, no matter how appealing the delivery mechanism, no matter how appealing the price. So, some constraints. Second: uncontrolled file sharing on p-to-p and potential. I don't think there's any need to belabor that point or cite statistics. I think everybody's aware of what's going on. Winning cases clearly doesn't solve the problem. In fact, even litigating the cases takes what in internet time is an eternity. Then we face uncontrolled CD burning. The number of people giving away and selling burned discs is skyrocketing. And this has actually had a collateral effect which hasn't gotten a lot of attention, which has been an enormous increase in street piracy, because anybody can become a…you know, piracy used to be CD playings, underground CD playings. You could use investigators and close them down. Now piracy can become a cottage industry where you buy a tower and in your kitchen you burn discs and then you sell them on the street the next day. So we have an enormous problem with street piracy - not just in the United States, but around the world. And all this is only going to get worse. Computers routinely include burners now, blank CD-R sales are multiplying, and the prices of the discs are declining.

Then we're faced with a combination of illegal downloading and burning. And that's when we really started noticing the impact on sales - was when people didn't just download this stuff and leave it on their hard drive, where there was only limited utility because people want portability, but when they began to burn them to discs and that became a real substitute for purchases. We're facing double-digit sales declines over the last two years, and that has an impact on the ability of record companies to continue to invest not only in music and talent, but in all the new infrastructure and delivery mechanisms. And we're also faced with a complicated rights situation. Everybody thinks that the labels control the ability to go on line, and so on, but the fact is we're only one part of a web of rights holders. There are song writers and publishers who have independent rights. There are artists who have independent rights. Very often they will license those rights to the record companies, but the more powerful the artist, the less likely that is to be - the more likely they are to have the right to control electronic distribution. And therefore they all need to be bought into the new business models, the new pricing strategies, and so on. That makes it complicated. Okay, what are record company attitudes towards copying issues? There is no one attitude. Different companies have different views, and I promise you that different people within the same company have different views. Having said that, if I had to categorize the most widely held views, it would be this: that consumers should be able to make personal use of music - they should be able to burn extra copies, they should be able to transfer to portable devices, and so on - but they should not be able to distribute it on the internet, and they shouldn't be able to sell or give away burned copies to others.

This is not a matter of fair use. This is not a legal doctrine or something. And I think that fair use is a much misunderstood and much abused concept, personally. I'm happy to talk about that later. It's rather an issue of commercial expectations - it is what Don was just referring to. It really is what the market place demands. Record companies understand that their consumers need to have a good consumer experience if they're going to remain customers. And it's as simple as that. They want them to have that kind of experience, and that means that they have to have those kinds of personal use capabilities. All right, what do the companies do? Once again, many different perspectives among different companies and different emphases at different companies about which strategy to pursue. Everybody knows there are no silver bullet strategies here. Rather, the companies are pursuing multiple concurrent approaches. Probably the most widely-held view is to get online as fast as possible with as many different delivery models as possible. Everyone believes that the only solution to online piracy is legitimate alternatives that consumers prefer. That's where we have to get. It's finally happening. There are now an extraordinary number of services that are out there with hundreds of thousands of tracks, incredible licensing activity going on, lots of new delivery platforms and delivery choices in terms of price points and what you get for it, and when burning's included and when it's not, and streaming, and downloading, and intended downloading, and they're getting better all the time. There are à la carte downloads as well. Not just subscription services, but you can buy that one song on the CD that you wanted - sometimes ninety-nine cents, sometimes forty-nine cents. But it's really happening.

There are still obstacles. Don't misunderstand. You may remember the Gateway deal with PressPlay where you could pre-load two thousand songs on a new Gateway computer. Couldn't get access to it unless you subscribed to PressPlay, but it was a way of promoting PressPlay. Well the music publishers have blocked that, and as a result that is no longer happening. The Beatles still aren't licensing their music online. So we still have more progress to make, but it is improving steadily. And while this is happening, the industry worldwide - the music publishers as well as record companies, worldwide, are building the electronic infrastructure for identifiers and messaging, and so on and so forth, to facilitate e-commerce. Second thing: companies are obviously looking at copy protecting CDs in order to prevent distribution on the internet and uncontrolled burning. But companies understand that consumers need to be able to play their music on computers, especially in the United States, and that they need to be able to make personal use of music. And the technology just isn't there yet. It's just not there. The technology that prevents ripping sometimes also prevents playback in some devices, like DVD players and car stereos and the like. And the vendors of these technologies keep improving them, but the record companies haven't yet found them good enough to use them on commercial releases. That's why we only have to date four or five albums in the United States.

There is a second section on these CDs, and that's what hosts the DRM. And those are pre-compressed files so that they can be played on computers. But to avoid the very problem that this was intended to solve, namely internet piracy or uncontrolled burning, you need the DRM to continue to protect the content when it gets transferred to the portable device. That means the device has to be an appliance that will accept protected content. And more and more devices are doing so, but you have a lot of different DRMs out there, so there's no one standard that everybody can turn to make everything work. That complicates the situation. And once again, for anti-trust reasons the industry can't simply agree on a single standard or two. You also want to enable some level of burning so that the consumer can make compilations, can make a number of personal use copies. But again, the technology isn't there to allow some burning but then prevent the burns of burns, which would eliminate the problem. Do I get the two more minutes? Thank you.

Audience:
Laughter.

Cary Sherman:
Third thing is new physical formats. There is progress being made in DVD Audio and SACD, and we will see how far that goes and how much it's promoted. There's new business strategies on CDs. The record companies are trying a lot of new things to add value to the existing CD: bonus tracks, free DVD if you're one of the first million customers to buy a CD, you get concert tickets when you buy the CD, you get access to merchandise. There are all sorts of experimental pricing strategies. So a lot of that stuff happening in the market place. Another strategy is to make p-to-p systems less attractive. That's where spoofing comes in so that you make it a little less easy to violate the law and encourage… another reason for people to go to the legitimate services. Another thing that companies want to do is start forging business relationships and give incentives to companies to help prevent illegal downloads. And that would be people like the ISPs. There is a deal between Rhapsody, which is a subscription service by Listen.com, and an ISP where the ISP is now selling subscriptions, and therefore has on its website: "Don't become an unwitting server. Do you realize that you're offering uploads when you are on KaZaa? This is how you should disable it." It's an example of where business relationships begin to incentivize ISPs to help with the piracy problem.

And we'd certainly like to see more of that. Then there's education and enforcement. Certainly, you've seen a lot more in terms about public education. There are artist ads and radio spots where they're talking about illegal downloading, we have outreached to universities and corporations and asked for them to take proactive steps to address this issue, and it's just generally stepped-up enforcement efforts so that people understand that there could be consequences at some point. And the last point is really to say that at the same time as the industry is transitioning its business relationships with its customers, it also has to concurrently be transitioning its internal business relationships with the artists and producers, the songwriters and publishers, and so on and so forth, because if you are going to sell less product, you need to find new ways to define your economic relationships to make the thing work. There have already been thousands upon thousands of layoffs. Hundreds of record stores have closed. Hundreds of artists have been cut from offers. The idea is to come up with business relationships that work in the new environment as you transition to the new business models that the future requires. A lot of moving parts here. That's what we're facing. That's what the industry is trying to do. There you have it.

Moderator:
Thank you very much Cary. Our next speaker is Sarah Deutsch who is the Vice President and Associate General Counsel for Verizon Communications. Her portfolio seems rather broad. She handles all global Internet policy issues at Verizon, Sarah, please.
Sarah Deutsch:
Well, we're very happy, you know, to hear that the RIAA members are finally looking at new...business models, but unfortunately as we've learned from our experience, the great associations are still pursuing the old business strategies. Part of that strategy is to continue to shift the burdens and liabilities for solving their new uh...business problems onto third parties including the IT companies, service providers and and continue to blame these parties rather than threat them as content partners and we're looking to be content partners. We're looking for some of those incentives that Terry's talking about that I think in a - in a friendlier manner. And, just to cut through all the rhetoric if you look, for example, at Verizon's DSL website you'll see that we want to grow our broadband services and we're seeking legitimate content to do so. We offer mp3.com and Rhapsody and a wild tangent in online gaming service. We actually make a profit when our users sign up for these services and we do not profit from key-to-key file sharing. Service providers were one of the first industries back in 1998 to agree in the DMCA to respect digital rights management and we did that in Section 512(i), which said that the service providers will accommodate and will not interfere with standard technical measures for DRM standards. The deal was that we would engage in voluntary and open standards talks with the content community to respect DRM technologies and come up with solutions.
The only caveats were that the standards be, you know, open to anyone to develop. Once they were developed, they couldn't impose substantial costs or burdens on our network or any kind of substantial cost. But, interestingly, since 1998, the content community has never once begun these talks with the service providers and uhm...unfortunately they've launched an unending series of the tax in the form of a three-legged strategy of legislation, litigation, and regulation, and a lot of this I think is because of the growth in P to P. P-to-P did not exist at the time the GMCA was negotiated and they're now trying to desperately twist an old law to fit any business problem. Last year we battled the Hollings bill which tried to mandate DRM solutions on industry uh...totally eradicating this particular section of the DMCA and we have talked about the voluntary solutions. We had the Berman bill that permitted self-help which I use loosely, and something called the Biden bill that was mentioned yesterday in uh...a presentation and that also created liabilities for service providers.
Uhm...most recently we've had the lawsuit that RIAA brought against Verizon, and many of you yesterday asked if I could just spend a few minutes talking about the implications of the case and what this means for enabling new business models. We have uh...appealed the case to the D.C. Court of Appeals. The lower case ruled against Verizon, so let me just spend just two minutes on that and - and then turn back to some newer things. As many of you already know, RIAA sued Verizon demanding that we turn over the name of one of our customer's name, address, and phone number of a person they allege has infringing materials on the customer's hard drive. They asked us not only to turn over this customer's identity but to disable access to the allegedly infringing materials via our system. Verizon in this case serves only as the conduit or pipe between the user and the Internet and none of the allegedly infringing files are stored on our network and, most importantly, we have no idea what's on the user's hard drive, nor should we be interfering with private communications. The only way for us to disable access to the material in this case is to actually terminate the user's Internet access account. The district court's decision, we believe, creates an extraordinary new roving subpoena power. It authorizes the clerk of the court without the judicial supervision of a judge to rubberstamp a form subpoena and have the service provider turn over the identity of the user. The form subpoena is based only on an assertion that someone somewhere is engaged in copyright infringement. The one-page form essentially allows anyone to obtain your identity based on an e-mail you've sent to a private party, a website that you may have surfed, or even an IM message that you sent. Every time you visit a website or you uh...post a message to a listserve where you sent an e-mail your Internet protocol or IP address is visible to someone. The service provider holds the key to your identity but the IP address unlocks the door. So, our plaintiff with that IP address and the mere assertion of a good faith belief that someone is infringing your copyright uh...this will allow service providers to turn over the identity to anyone and there could be stalkers, identity thieves and worse people out there who get access to your identity. So, we think the case has significant privacy implications. We've already seen this issue being abused.
The copyright owners use a copyright search box which speedily looks through the net for file names that match the names of their copyrighted works and you may have heard about the example where Warner Brothers wrote to UUNet asking them to terminate the account of a child who had downloaded the Harry Potter book report apparently or that was available on their hard drive. We also think that this private search warrant power is unprecedented in U.S. law and is unconstitutional. Article III of the Constitution requires that a party can not use the power of a court, in this case a subpoena, without a case or controversy before the court and that case or controversy language is being raised on appeal. There's no requirement in this case that the copyright owner file a lawsuit with the court and if RIAA wins the case is going to have a huge chilling effect on user privacy and RIAA says they don't want to sue anyone. They won't have a remedy other than having the service provider turn over the names but the fact remains that they always had a remedy and that's called a John Doe lawsuit, and Verizon and other companies have used the John Doe lawsuit successfully with a real judge and there's due process and privacy protections built into that process. The remedy may not be as convenient for RIAA as the roving subpoena power but we think it's the right test.
They would rather be able to use the ISP to send hundreds of thousands of cease and desist letters to their customers or uhm...you know, send these files to a private investigator who can then make sure that when they do bring a lawsuit they sue the motorcycle dude with some body piercings and jail time rather than a 12-year-old child. So, but we don't think that that's the right use of the law. And to quote the ACLU: "While technology is constantly changing the principles of the Constitution remain constant." So, we think there are some important lessons to learn for why the content community's scorched earth tactics to shut down P to P rather than compete with it are doomed to failure. If you consider - consider this law, this passage from a real law, no person shall manufacture, sell, barter, transport, import, export, deliver, furnish, or possess, fill in the blank, except as authorized by law. Is it illegal DRM tools or anti-circumvention devices? No, the missing piece here is that it was illegal to manufacture, sell, barter, et cetera, intoxicating liquor. This was the 18th Amendment to our Constitution. I think we can learn an important lesson from prohibition. There is a startling analogy between prohibition and the doomed efforts by the copyright owners to shut down P to P, threaten users, and force mandated DRM solutions. When prohibition went into effect, millions of people immediately began violating the law with a vengeance. Police records showed drunkenness among teenagers and children had increased tenfold, and there was a proliferation of organized crime.
There was apparently enough beer and illegal gin in Chicago to float a battleship. At a 1926 Senate hearing on prohibition it sounds eerily like this weeks P to P hearings, Senator Meade of Missouri asked a Yale student how easy it was for students to get alcohol; the answer, why, it is obtainable sir, the greater the attempts at enforcement the stronger the sentiment against it. The dark net paper that the Microsoft employees wrote concludes that shutting down P-to-P is doomed to failure. It basically says that it's only going to force file sharing to go underground, creating anatomizing routers, overseas routers, object fragmentation where pieces of the files do not exist on any one server and other means to complicate the efforts to find these copyrighted bits. These authors in short say that if you're competing with dark net you must compete with it on its own terms and they say that that means convenience and low cost rather than additional security.
I may be out of time to react to Lon Sobel's paper but I'll just hopefully get to that in the question period but I'll just say very briefly that his attempt to propose this compulsory license I think and turn ISPs into digital retailers should be commended. He's thinking about this in a creative context. We've been saying all along that we want to be thought of as content partners. Verizon even raised the issue of a compulsory license last year and it would be the most severe form of understatement to say the content community did not appreciate that idea but I'd be glad to go through some of the concerns I have with the details of this paper but I do commend him for starting to think along these more creative routes. Thank you.
Cary Sherman:
This is hard to do because we start from very different starting points. From the very beginning, the notion that we're attempting to switch liability to ISP is exactly the opposite. The idea was not to hold the ISPs liable but to go to where the infringement is occurring. That was the whole idea of the DMCA. The DMAC was you should have no obligation to monitor your networks but when infringement is brought to your attention you have an obligation to do something about it and to let us know who the infringer is so that we can enforce our own rights. That's all that we're trying to do is get the information so that we can enforce our own rights. It is really preferable that we have to sue in order to do that as to be able - instead of being able to issue a warning to the people who are actually engaging in infringement activity. I don't see the public policy benefit of doing that and I don't see that there's any difference in the burden on Verizon in terms of providing that information and that's what the court found. If you haven't read the court opinion, you should because it makes it very clear that the law is very clear. The public policy objectives are very clear and that Verizon's reading of the statute was completely strained and unsupportable. In terms of the form that we file, this is not something that can be automated. You have to go down to court with a sworn declaration that this will be used only for this purpose and there would be penalties if you use it for any purpose other than copyright enforcement, and that there has to be a good faith belief on the nature of the infringement, so on and so forth. So we can go on and I can refer to the alcohol and all that as examples, but I understand that we've got a limited amount of time.
Moderator: Okay, thank you Cary. I think it must be time to turn from lawyers to computer scientists, so our next speaker is Bob Blakely who is the chief scientist for security and privacy at IBM Tivoli software.
Bob Blakely: I want to start out by asking a question of the audience. Raise your hand if you think the Internet security architecture is a failure? Come on, don't be shy. Come on, you all wake up. Okay. All of you who have your hands in the air are wrong. Okay. The Internet -
[Laughter]
Bob Blakely: No, it's demonstrably true, okay. But most people don't know what the goal of the Internet security architecture was. The goal of the Internet security architecture was to make sure that packets got from one place to another in case part of the network was destroyed by a nuclear explosion. Okay, that was what the Arcanet design goal was. That security goal was admirably met and, in fact, it sort of gives rise to the problem that we have today. The Internet facilitates very, very efficient distribution of any kind of digital material so naturally, of course, as soon as people noticed this they began to think, okay, well, you know, we don't need to use this just for like military command and control. There's other sorts of things that we could digitize and we could distribute them, and, you know, realizing, you know, pretty immediately uh...that it was going to be impossible uh...to stop content from going pretty much everywhere because that's the whole design criterion of the Internet, we in the technology industry immediately encouraged you to digitize everything, right, and you fell for it and so, you know, this is sort of, you know, when Barbara says yesterday that the "Sum of all Fears" is the redistribution of digital content this is what she's really talking about. The sum of all hopes originally was that we would have cheap, easy distribution of digital content, and what we got was, you know, that which was granted and we also got cheap, easy redistribution of digital content. Uhm...so the first question for all of you is, you know, do we really want to do this again without thinking about it with a new technology, DRM, or, you know, would we prefer to actually put some thought in and look at the possible failure modes. And, of course, as a security guy that's what I do. I look at failure modes and I, you know, sort of - I love failure modes. They're lots of fun. So, for the next like about eight minutes or something like that I'm going to ask you to imagine the following alternate reality. Okay, the DRM fairy has arrived. She's waving her magic wand and you get what you ask for. Now, by the way, of course you all know what happens in fairy stories when you get what you ask for.
[Laughter]
Bob Blakely: Okay, what you ask for in this case is every time somebody tries to invoke an operation on an item of digital content, it succeeds if and only if the content owner has granted some appropriate enabling right to the person attempting the operation and, not only that, but writes in value. Okay, so it's not, this isn't just some sort of fantasy that uh...you know that people are willing to pay for rights. That's actually true. Okay, so you have to pretend that that's the case, even if you disagree with it, okay, and not only that uhm...but you have to pretend that you are now a content owner and you don't really care about what happens to consumers or people's privacy or anything, and we're just going to talk about what happens to you as a result of the - of the DRM fairy doing this.
Okay, so if you could see the screen right now, what you would see is you would see a picture of two different cameras. One of them is a Leica M6 and the other one is a Hexar, a Konica Hexar RF, okay. These are both great cameras. They're really, really excellent. The the Leica M6 new at B&H Photo, which is, you know, an online retailer, costs about $2,600. Used on Ebay it costs about $1,700. The Konica Hexar RF is actually a slightly better camera. In fact it's better - oh, there. It's starting to appear. Uhm...it's even a little bit better than the new Leica M7 because it has automatic wind and what have you and new on B&H, this better camera is $999 and used on Ebay it's about $700. Now, this isn't an accident, okay. One of the reasons that Leica can sell the M6 for $2,700 new is because the user can sell it used for $1,700. Resale value supports a high initial price. Any of you who have, you know, bought either a nice car or, you know, a Hugo understand the principle.
[Laughter]
Bob Blakely: So, you know, you want to make sure that your DRM allows people to maintain resale value or else they're not going to be able to support a high initial price. Another thing you can do with DRM is you can segment things geographically. Okay, you can decide that your CDs are only going to play in England of Japan or Greece or what have you and that's great, you know, because it, you know, it does things to your market model. But the other thing that it does is it raises your inventory costs because if you run out of CDs, you know, in England, you can no longer just call up somebody and have him send a bunch over from France on the Eurostar because they're not the right kind of CDs. So, your distribution network and inventory costs have to be set up to handle whatever you're doing with DRM. There - by the way, there are tricks that you can pull and we'll talk about some of those in a minute, oh, actually right now. So, one of the things that you can do is you can put the DRM on at the last minute, right. You know, there in the store you've got these generic CDS and you sort of, you sort of stamp them with their specific uh...DRM nature at the last second. Uhm...well, there's a couple things about that. The first is you've got to have the right flexibility to give the right set of rights.
One of the bad things that the security industry observes is we keep trying to sell single sign-on products, right, which means you want to be able to log onto every different kind of device that you use on a daily basis, and one of the bad things about trying to do this is we always, you know, start out with the theory that if we cover 80 percent of the most popular log-on targets in the world we can sell the - the stuff but that really isn't true, of course, because each individual customer wants 100 percent of the stuff that he personally uses. So, you have to have the flexibility to get the right set of rights. The other thing is, of course, if it's the guy behind the counter at Power Records who is burning the DRM onto your disk, you know, he might care about your revenue, but he might also just load up all the rights, you know, for - if you slip him a $20 across the counter. So, you create opportunities for fraud downstream if you have customization with DRM stuff. Another problem that you've got potentially is you got to think about exchanges, okay. You sell somebody the British disk in Greece or the Greek disk in Britain, you're not any longer when a customer comes back in and says I got the wrong thing, you know, it's not like I got a scratched CD, give me another one of the identical thing. It's I got the wrong thing. I live in England. You need to go back in the back and get the England thing instead of the Greece thing, so it's a more complicated exchange process. Uh...another thing, of course, is any time you add a sort of security light system by which I mean something which is designed to make the device fail under certain circumstances, the device will sometimes fail under those circumstances, or maybe even under other circumstances, and when that happens the user is going to call you up and he's going to say the sucker is busted, right, and you are then going to have somebody on the other end of the line who says no, it's not busted. You're trying to do something evil. Or, yes it is busted.
You bought the wrong device or whatever and that, of course, means a help desk and those are expensive. Uh...another thing that you might want to think about is not necessarily whether you're actually invading people's privacy because, you know, of course as the, you know, as the provider you don't really care about that. Well, maybe you do. But whether you're actually incurring a privacy liability so you have all these, you know, these records of what people are watching, you know. The Green Party Manifesto, NRA instructor's guide, you know, all of the things that subject you, you know, to, you know, European human rights provisions and data protection directive and what have you. You now have all that information just because you, because you implemented DRM and now you've got liability for protecting and using it properly. Uh...of course it's also possible to discourage sales by collecting information about the wrong thing and uh...you want to pay attention to whether you're discouraging sales. You all can read that, right, so you get the joke? Somebody raise their hand, in fact, if you can read that. Thank you. Uhm...another thing that you should worry about is, you know, if you uh...are using DRM to restrict uses of a product, uh... you know people may not just value the content, right. People trade ring tones on cell phones. Now, ring tones are not, you know, like great art, okay. The content actually isn't the thing that people are interested in. What they're more interested in is sort of the social interaction of trading and collecting ring tones and if you don't let them do that with your ring tones, then they'll find something else to do it with. So, you ought to think about whether the actual rights, rather than just the content, are things that people value and whether you're going to create niches for competitors to compete against you on the basis of the rights instead of the content uh...which leads to the next point. If you believe that rights have value uh...to the content owner, then the way to get value to the content owner is by having value to the customer so you sell them to the customer.
If they have value to customers, then your customers as soon as the DRM fairy waves their wand are going to start uh...comparison shopping between you and your competitors on the basis of the rights that you offer. If they comparison shop, of course, this is going to put price pressure on the rights owners and for a while you're going to be able to differentiate by offering more rights, but only until everybody grants all the rights that can possibly be granted, and then at that point you're going to have to start lowering the price. So, you should think about whether, you know, there is some sort of revenue curve and you get value out of rights for a while but then after that you get price competed away and you have to give away all your rights for very little value and you've implemented this big, expensive DRM system which is sitting around generating help desk calls but not a lot of revenue.
[Laughter]
Bob Blakely: And, finally, of course you have to worry about what happens uh...if the DRM fairy has a bad day. Okay, once your - your new business model is in place and everybody is competing very efficiently because everybody's paying for every piece of content, you know for a while that lowers your unit cost, right, because more people are paying for your units. You're selling more of them uh...and you get the price competed down and then all of a sudden, you know, some brilliant photographer breaks DRM. Now your costs go back up but what do you do about the price? That's really all I had to say, which is an excellent thing because I had zero minutes left. Thanks very much.
[Applause]
Carl Shapiro: The tooth fairy is a tough act to follow but David Reed will be able to do it I'm sure, David Reed going up there from CableLabs. I think we'll, as we heard from Sarah at Verizon, we have the perspective of cable companies kind of caught in the middle as distributors, David.
David Reed: Yes, I thought I'd speak from up here so Cary doesn't have a more direct shot at the table there. It came to my attention that a bio on your web page said that I wrote the book, "The Hip and the Hype" and that I was the Internet mortgage lending expert making $42 million per year. I hate to disappoint, but the truth - the truth is that the book I wrote was called, "The Engineering and Economic Analysis of Residential Fiber Optic Networks." If I had $42 million at my disposal, I'm not sure I'd be standing here right now.
[Laughter]
David Reed: So, that being said, I'm going to give an overview of what has been going on in the cable industry with regard to DRM, a current events of sorts, and the important points I want to emphasize uh...in my brief address is that one, cable is caught between competing interests. This is an interlocking problem, tectonic place. A dozen industries are colliding here, and that the cable strategy is trying to balance the fair use concerns uh...of the content providers with the business concerns that were necessary at that time to enable me to use this model. That's one point. Number two is that if they were building the infrastructure now to head - to lead the transition to do a little television in a mass uh...market uh...fashion, and frankly today we're deciding on the interfaces, the digital interfaces that would be supports and set top boxes and televisions, and the attributes of the copy mechanism - copy mechanisms on those interfaces so that's happening so that's coming along. And number three, that's relevant and leads to the focus here of the panel is that cable has proposed a strategy for covering the existing business models uh...that are provided to cable services today but the problem of future business models uh...is a tough one and ultimately likely to be deferred to regulatory to the FCC in order for them to solve so it is a very tough problem. So, let's cover those issues quickly uh...in a bit more depth. Number one, the need for Internet industry cooperation here and, today, the FCC requires the set top boxes uh...will be available uh...through the retail channels by 2005.
At CableLabs, we've been designing the engineering specs that will support that requirement in a project called Open Cable. The development of the rules and the engineering specs that are needed to make this requirement happen are complicated by the digital format of the programming that's flowing through these devices. The spec objective that we had is to enable the customers to watch what they want and when they want it as well as to have the features of their TV work. In order to do this, it requires cooperation of cable, of the consumer electronics vendors who build this equipment of the studios or the content newsgroups. So, this is a battleground that has been drawn in this area and trying to reach consensus on these specifications. IT often focuses for the most part has focused in the past year on the technology licenses associated with the interfaces that are being defined in these ports and to the television. And, it's been a subject in Washington. They have had roundtables where industry has gotten together and they're discussing in these public forums in order to say, okay, content provider what's your position on this? Cable operator what's your position? CE vendor what's your position? And so, out of this discussion the cable industry and the consumer electronics industry in December announced a major agreement that uh...on how they will proceed to deploy and support what we call one-way devices. So, these are devices that will receive programming in a downstream direction. That is from the head-in, the cable head-in to the consumer.
That's the one-way direction. So, that was a major agreement where, you know, had been a significant hurdle in terms of the the consumer electronics companies and the cable industry being able to agree on a direction to proceed. Uh...we have achieved that agreement and within that agreement we get with the FCC with the uh...consideration emphasis to how those rulings, you know, will be likely to create regulations out of this agreement. Now, within the agreement there are encoding rules and on these interfaces are what we call encoding rules that have an impact on the business models. That will be supported uh...by the various copy spectrum mechanisms that will be supported on the cable network and in the consumer electronics prices. So, it's an interlocking world that we live in in terms of trying to support the technical solutions that support these requirements. Point number - point number two on the course of the set top boxes on TVs what are they? Well, there was 1394, IEEE 1394, and it's a physical plug that would be in the television or set top box that supports for example digital devices. This uses the digital transmission and content protection of what's called DTCP. Another port that is part of the agreement is called DVI, that's digital video interface and that permits uncompressed digital signals to the TV display. The bandwidth there is so high because it's uncompressed that it does not support recording devices. The copy protection there is called HBCP, which is high bandwidth digital copy protection. Basically everything is copy never and display only because of its uncompressed nature. And finally, there's the part called the high host interface and that's what the government has required the cable industry to implement, and the specs, then cable has developed the specs that separate the set top box, which we call a host device and the descrambling function into what we call a point of deployment module for a pod module and that plugs into the host. It - it descrambles or un-encrypts the - the conditional access system that you have on the cable network and, of course, when you do that you now have an interface that's in the clear and so the encryption technique that that is used across that interface is called dynamic feedback arrangement descrambling technique or feedback uh...and it's important since that has been a - a very hot topic in Washington over the past year a part of the negotiations uh...in order to support that requirement. So, feedback with the encryption is used across that pod host interface for copy protection. So, what you have here in the big picture, what you'll see emerging out of this uh...agreement is the cable network coming in to a uh...set top box, for example a high definition a high definition set top box. You'll have this pod host interface that will decrypt the information that's coming over cable and the output ports will have 1394 that can connect to a BBR. You'll have a DVI interface that connect to a TV, digital TV and you'll have component analog outputs then that would support older TVs. It's kind of a simple way that you might anticipate it to emerge. Now there are encoding rules that are part of the agreement as I said. For the most part over 1394 they mimic for those of you who are familiar with that interface that's known as 5C.
Those of you who are unfamiliar, basically that means that free over the air broadcasts may be copied freely, any digital program sold by a monthly subscription so what's called SVOD, for subscription video on demand may not be marked more strictly than copy ones. The programs sold as video on demand may be restricted as copy never uh...but may be shared for at least 90 minutes. So, these are the type of areas where the encoding rules focus and in terms of that particular set of agreements emerged from consumer electronics company and the studio - two of the studies agreeing with that set of encoding rules. And the rules apply to these existing business models so that the sticky and gray area here is whether or not you can support new business models. So, what - what the agreement basically says about new business models, which is impossible to get an agreement over what you don't know will be, it's likely you'll have some type of a FCC type of a hearing where somebody will submit a new business model and the FCC would make a determination on that business model based upon the public access. So, the challenges going forward here then are for the cable industry, we're going to be working with the CE industry to for agreement on two ways. We've got the one-way and we'll do a little bit more in trying to establish how a two-way device, a two-way interactive devices and the rules of the interfaces now have implications obviously for the business models that will be supporting them. So, in summary, cable is in a distributor role here between the content providers and the folks who - who build the boxes and we're adopting protection measures that permit reasonable consumer copying but a balancing role for new business models as well as trying to get the most attractive content that's available. Point two is we're strongly based upon strong encouragement by Washington we're moving forward in order to support and solve these problems. And three is that the business models, new business models, are likely to be solved in a regulatory environment and that...uh...we have obvious concerns for innovation here that any regulatory model that you put forward can then support innovation, of course that's a concern of all that are involved.
Carl Shapiro: Thank you, David.
[Applause]
Carl Shapiro: Our final panelist Allan Adler from the Association of American Publishers. Allan is the Vice President for Legal and Governmental Affairs there. Uh...he's going to uh...not surprisingly talk about e-books. I think that it's - it's easy for many of us to, when we think about DRM, to think about music and then go on and think about movies or of course textual material, artistic works. This is very relevant for us I think it's…Allan take it from there. We'd like to hear something rather complementary to what we're hearing all day.
Allan Adler: Thank you, Carl. Thank you for the invitation to participate today. Carl has already sounded my opening theme which is that hopefully my contribution today to this discussion, if it accomplishes nothing else, uh...might get people at least persuade people when we're speaking of DRM and other copyright related issues to stop homogenizing the very diverse interest and viewpoints of the various content industries to shorthand references to Hollywood or the entertainment industry. Uh...book publishers aren't movie studios. They're not record labels, their products, their markets, their business models, or even their treatment under copyright law are quite different than those industries. Uh...I'm not going to elaborate further by raising the cause of needlepoint pattern makers but you should know that there are a great deal of diversity of interest out there uh...with respect to copyright and more so than you mostly tend to hear about. Uhm...and if copyright was ever intended to have a one-size-fits-all model with respect to uh...the uh...introduction of creative works, original works, for instance creative works of original expression uh...as products in the marketplace, it's pretty clear that that one-size-fits-all model doesn't work with respect to the digital environment. So, let's talk about e-books for a minute and maybe I can give you a different take on the issue of uh...some of the recent calls for government mandates that people say stifle innovation. Why do we even have e-books? I mean the simple fact of the matter is that uh...copyright law doesn't require the availability of literary works in digital formats and, as far as I can recall, the advent of digital technology didn't raise a groundswell of demand uh...to see books published as electronic uh...books. Uh...but the simple fact of the matter is the book publishing industry recognized that this new technology was allowing it to present literary works in an entrepreneurial fashion to the reader community with added functionality not possible with traditional book technologies to enhance the reading experience and enjoyment of literary works. But the same digital capabilities underlying that product enhancement, of course, as we all know pose risks, serious risks of market failure and publishers' ability to control reproduction and distribution of those works. I'm not telling you anything new here.
Traditional hard copy print publishing technology inherently imposed physical limitations on ease reading and of reproduction and distribution, the binding on the book, the spine on the book, the fact that you have to turn the pages of a book in order to be able to photocopy it, even the degradation of the quality of the copy that's made when you photocopy with each successive generation also to protect those market interest against uncontrolled reproduction and distribution. With e-books on the Internet environment such limitations are wholly eliminated and the technology instead inherently facilitates instantaneous perfect reproduction and immediate global distribution with just a few keystrokes on anyone's PC. With such potential for market harm, it's clear that e-books wouldn't even exist as a mass-market product but for the security afforded by DRM. So, have we managed to control DRM for e-books and to properly calibrate the way DRM works with e-books? Well, no, we haven't. In fact, I would say that the first generation focus on content security succeeded probably too well in the sense that it put the successive e-books at risk by frustrating consumer preferences regarding the very new and valued uses and functionality that justified the production of literary works and digital formats in the first place. There are added problems to this too. Uh...book publishers don't make DRM. We have to rely on hardware and software industries for that, and frankly DRM that was initially proposed with respect to e-books ran into some conflicts with respect to the vendors' own focuses on technical feasibility rather than consumer preferences, the struggle of these vendors for their own proprietary dominance in the marketplace, and their own mission of convergence which, for their own purposes, uh...would have made it more attractive to have them protect all content in the same fashion.
Well, as I said before, one size doesn't fit all, not with copyright law nor with DRM. We've relearned a very hard lesson in the book publishing industry in the DRM context. That is that business models derive from customer needs and preferences, yet the user consume of e-books was frankly the missing player in the DRM considerations in the early stages. It wasn't because the publishers didn't want to talk to the consumer. It's simply because of the business model we don't deal directly with the consumers. They buy the books from retail outlets and from other middlemen distributors. But through recent efforts we tried to catch up, especially through collaborations with libraries. Publishers are learning more and more about consumer preferences regarding e-book functionality and are not generally convinced the currently available DRM systems fall short of consumer expectations in very specific respects. I don't have to go through the litany of those various respects with you. You can all go look at Joe Kraus's website, and in fact, I wouldn't even say there's one list but each of you probably have you own list of the types of functionalities you would like to see enabled with respect to e-books but which may not be enabled today because of current DRM systems.
Well, the fact of the matter is, DRM is now becoming a competitive factor among publishers in their competition to introduce e-books as successful products in the marketplace and working with technology vendors and libraries and in standardization projects, for example like those conducted by the Open E-book Forum, we're beginning to see substantial improvements in DRM systems to facilitate both a standardized set of DRM features and functionalities in a package that users can come to routinely expect and understand in e-book products. Are publishers and users really on the same page? In fact, are we ever going to really get on the same page? I'm not so sure and the reason for that is because we have two major problems. First is that the strongest consumer preferences with respect to e-book functionality curiously enough are the capabilities to do with e-books just what they've normally done with traditional print books. In fact, in some cases they seem to want uh...e-books essentially to just be an electronic environment of the print book. But that's not where they start because the truth of the matter is once you introduce them to the added enhanced functionalities possible with digital technologies they're really not satisfied with the traditional limited uses of print e-books. They want the flexibility. They want the convenience. They want the enrichment of the reading experience that comes with the capability of reading literary works in digital formats. But there's a paradox there too. The paradox is, is that e-book functionality as a subset of DRM issues have produced calls for unnecessary regulatory interference with competitive practices and industry standardization efforts in the very still nascent e-book marketplace.
And remember, when I say nascent, we're talking about a product that didn't even exist five years ago, and the fact of the matter is, is that they're calling for these issues to be addressed through government technology mandates, the very thing that some of those same voices condemn when they talk about some of the efforts of people among the content industries to invoke government mandates for purposes of protecting their interest in the marketplace. While consumer expectations are otherwise ready to embrace the enhanced functionalities of e-books, they're apparently unwilling to do so if it means they can not apply all of the familiar copyright use rules they have become accustomed to applying in the use of traditional print books. But why should this be so? Why should consumers expect to enjoy the additional new added value uses of the literary work beyond what they could have facilitated through traditional print book technology without having to accept the idea that some change in the application of those traditional copyright use rules to accommodate the risk attributes of the very same digital technology that facilitates the new and valued uses must be made.
Well, frankly, the product is - the problem is that the e-book market albeit still in its fledgling stage is evolving through competition among publishers, through competition among technology vendors in the software and hardware industries, as well as through coordinated standardization efforts by the industry organizations like the Open E-Book Forum to address platform shifting and all of the other various functionalities and other such "consumer expectations" as affirmative rights of competitive choice rather than as anti-competitive government mandates that are so-called rights. And, the fact of the matter boils down to this from the perspective of the publishing industry. If the ability of users to access copyrighted works on multiple devices or technological platforms or to do all of the other things that are characterized as consumer expectations is really important in today's marketplace, that these are issues that should be met through evolving business models of competing e-book publishers and technology vendors in an environment that's governed by market forces and private sector cooperative initiatives rather than by intrusive government fiats that distort the first sale and fair use doctrine. Thank you.
[Applause]
Carl Shapiro: Rather than ask those who want to defend intrusive government fiats, I think I'd like to turn a little bit back to Lon Sobel's proposal because I know there are some responses to that and then it's open to the floor for questions in a few minutes. Sarah, I know you had some more to say about Lon's proposal, please go ahead briefly.
Sarah Deutsch: Sure, thanks. Very briefly, Lon says his proposal amounts to a compulsory license, but my question is, is it really one, you know, would you have Congress actually saying that the materials system is considered legal and not infringing? Then, I had some fundamental questions. Do you want to turn the ISP into the tax collector, unlike Kmart that at least knows that these Martha Stewart goods are coming into and store you can charge - decide to raise the price or proportionately to lower the price with the blue light special for Martha's goods? We're not going to know what's coming in the door and it may be very difficult to get people to buy into this system and get them to agree to pay. So, why not think about a flat fee system if it doesn't compete with P to P or at least feel like free wouldn't people just continue to seek out illegal content? I'm also a little wary of some of the technical issues here, for one a marketing digital fingerprint and the ease by which they could be circumvented because people with a little creativity are always one step ahead of the technology, which is always two steps ahead of the law. Uhm...perhaps, also fundamentally given the billions of bits of information traveling across networks, the idea of this complete and individualized end user billing system is extremely ambitious. I mean we're really good at billing but I don't know that we're that good at billing. If you think of the bits traveling across the network now as, you know, a six-lane superhighway unimpeded, we're talking about a proposal where you're going to stick a toll booth there and then rather than little plopping in your 25 cents and waving to that glazed person in the booth, that glazed person is going to be charging you individual amounts for every song that you play on your radio, for example, that each song is individually priced and you have multiple copyright owners, and everyone can be a copyright owner, so you've essentially created the world's most complicated billing system. I also just wanted to raise the privacy issues that Lon mentioned are very difficult. The ISPs could be turned into Big Brother with copyright owners wanting to mine the watermarking and fingerprinting detections the basis for enforcement purposes and there's a slippery slope argument where other third parties uh...with not so wholesome purposes could be asking to attach their fingerprinting into your databases and into service providers' networks.
[Applause]
Carl Shapiro: Lon you want to uh... just give it a minute or two?
[Laughter]
Lon Sobel: I think I was counting about a half dozen so let me see if I can remember them. First of all, do I intend for it to be a compulsory statutory in the sense of an amendment to the Copyright Act, yes, so there really would be an exemption of liability for Internet service providers as well as for all of those that operate websites and P to P networks and the rest. Do I really want to turn ISPs into mandatory retailers unlike physical retailers that know what merchandise is coming in, yes I really do, and that is the trade off that I'm asking consumers through their ISPs to make in return for the privilege to be able to have ready access through a website, P to P networks, or whatever else might be invented for the copyrighting works of their choice. One thing that I'm not proposing or suggesting is that copyright owners, because remember every one of our e-mails is entitled to copyright. I'm not insisting that every copyright owner actually demand through watermarks or fingerprints compensation and, indeed, what I can imagine is the creation of copyrighted material suitable for display on websites for which the copyright owner would not in fact ask for any per use copyright fee, so that those who design websites, for example, would make decisions about what to put on the website and among the decisions, among the basis for the decisions would not only be the attractiveness of the material itself but the question of whether putting that material on the website would result in one of these pop-up notices that I have described giving the user the opportunity to decline. If I were designing a website I might decide that I wanted to forego anything for which a royalty was claimed so that when my page appeared, it didn't come accompanied by a bunch of pop-up notices as well. I'm trying to think now where that - you - oh, the billing. With respect to the billing, uh...this is all I can say. When I had this idea it occurred to me that I was asking a lot of computers, even big powerful servers that are operated by ISPs, but I talked to some people, actually at Pam Samuelson's suggestion, who were in the business, because though yesterday she was disclaiming knowledge of technology, she knows enough about technology to have had the same thought that you had, which is that the billing system that I have described is a billing system even more complicated than the billing system used by long distance telephone service providers. So, I did talk to those that are in the business, vendors in the jargon of the trade, those that are in the business of vending the kind of technologies, the kinds of software that can do what it is I want to do, and I was assured by those that I spoke to that their companies could provide, and indeed at that moment have available-
[Laughter]
Lon Sobel: -the kinds of technology that I thought my system required. I did mention that very telephone conversation to a colleague of mine who is in-house for intellectual property purposes at one of the major motion picture studios and, though this system I think would benefit her fine, her employer, she looked at me and she said with a smile much like your laugh, she said I think the key word there Lon was 'vendor.' In the meantime, here's my response. My response is though the technology may not work as I've describe it ought to work today, it is the case that there are things available today that do the sort of thing I want to have done, in other words what I propose does not require the invention of new technology though I acknowledge it requires the protection or may require the protection of existing technology.
And then, again, with respect to the billings, I'm not saying that a monthly bill has to include 1,000 entries for every digital image that was downloaded. I could imagine that the monthly fee that each one of us paid is to cover not only the cost of a connection as we do today but would also cover the cost of a certain quantity of copyrightable data that would be downloaded so that the task of undersigning to the ISPs is the task then of just keeping track of what was downloaded. Oh, I know, the last point that I think was by way of response was wouldn't it be preferable, this was sort of the implication of what you were saying is, wouldn't it be preferable to impose a blanket license fee on Internet users in the nature - in the nature of a blanket license, to use music industry terminology? Wouldn't it be preferable to impose a blanket license on Internet - on Internet service provider customers, Internet users, and then to have all of those royalties flow into a pot and to do some sampling of what it is those of us that use the Internet are downloading and to generalize sort of the samples with respect to our predictions of what was being used and to make an allocation based on that sampling much the way ASCAP and BMI today sample radio play in order to have an approximation of what songs were actually being broadcast? Administratively that system would be much, much simpler I acknowledge. It also is the case but I don't think that that would be attractive to copyright owners at all.
Among the particular problems that I foresee, for example, is the fact that not all digital data that's being downloaded today has the same value per megabyte as all other data. Adobe Photoshop is $600 through vested retail. The cost of a single musical recording from an authorized license source is as little as 49 cents, so there's this tremendous spread between the value of digital data per megabyte and to impose upon the copyright arbitration royalty tribunals or the copyright office the task of trying to assign the blanket license and then allocate it would be a task that uh...would not be accomplished during my professional career and even for my students who are in the audience I doubt that it's a task that would be able to be accomplished during their career and that's why I prefer to go to the copyright owner sets the fee. Even her work and even a changeable fee over time so that the work as it gets older the price might decline as a result of competitive market forces.
Carl Shapiro: Thank you, Lon. Bob, you wanted to make a quick comment on his point?
Bob Blakely: Yeah uh...just regarding uh...billing systems. I mean I think you ought to be very, very skeptical of the claims of vendors.
[Laughter]
Carl Shapiro: Do we have vendors in the audience?
Bob Blakely: Well, you know, so I'm - I'm a vendor although not necessarily exactly with this technology. In the computer industry we now have three generations of people who offer payment systems and micro payment systems. The first generation of these companies also seem to have been going out of business before the Internet bubble burst.
[Laughter]
Bob Blakely: The second generation were washed under by that and the remnants are trying again. Billing systems are very, very difficult and, of course, the failure mode of a billing system is really bad. It means a lot of real actual money disappears. Let me just describe what I would do in this system and we can think about it, okay? A piece of digital content with a little copyright watermark attached to it is another piece of digital content. I would uh...even if I wanted to get rich and - and had some place off the planet out of jurisdiction to move simply write a little worm that goes around the Internet and re-copyrights everything that's already gotten these little watermarks on it, add a - a one one-hundredth of a cent charge payable to me in addition to the charges that are payable to the legitimate copyright owner and I would just wait for the money to roll in.
[Laughter]
Bob Blakely: Now, by the way, if you don't like that one, you know, as long as you keep buying me drinks tonight I can think up more.
[Laughing]
Bob Blakely: I think, you know, it pays to think this through very, very, very carefully.
[Applause]