Featured Fellow: Paul Hoffert
This is one of a series of posts on Berkman's fellows (Derek Bambauer, Mike Best, Diane Cabell, Phil Malone). The Berkman Center is home to approximately thirty fellows, all of whom focus their time and energy on issues concerning the Internet, including Internet governance, privacy concerns, intellectual property rights, competition policy and antitrust issues, electronic commerce, the role of new media and journalism proper, and digital media, among many others.
Paul Hoffert is Chair of the Bell Broadcast and New Media Fund, Chair of the Guild of Canadian Film and Television Composers, faculty fellow at Harvard University, Fine Arts Professor at York University, and a Board Director of the Glenn Gould Foundation, and the SOCAN Foundation.
Below is a Q&A with Paul Hoffert about his work.
Question: You are currently one of the principal investigators on the Digital Media Exchange (DMX) project – a Peer-to-Peer on-line service. What is the point of DMX, how does it work, and what distinguishes it from the Peer-to-Peer services currently on the market?
Paul Hoffert: The entertainment business model is broken. Content owners are receiving less revenue than they believe they deserve and many consumers feel that the cost of legal content is too high. The easy availability of content through illegal exchanges, for free, has undermined the traditional business model, and attempts to fix the problem with legislation and policing have so far failed. DMX presents an alternative model for the entertainment business as suggested by Prof. William (Terry) Fisher in his book "Promises To Keep". To accomplish this, DMX must deliver more money to copyright holders while charging content consumers much less than the current business model does. A tough problem that DMX nonetheless solves.
One part of the solution is to attract hundreds of millions, perhaps billions, of paying customers The second part is to remove some of the unnecessary profit from the distribution chain by making the DMX service non-profit. And the third part is to remove the hundreds of millions of advertising dollars spent on attracting consumers to legal Internet content distributors, by collecting content payments from end-user aggregators, such as universities, ISPs, and mobile phone companies, which deliver 100% of their subscribers or students to DMX.
The DMX P2P service differs from illegal ones because it is compliant with copyright laws. The DMX network only transfers content that has been licensed for such use, plus content in the public domain. DMX ensures that content files are excellent quality, virus free, and do not contain ads or other bogus messages—problems associated with illegal file sharing systems. An objective of DMX is to convert the tens of millions of illegal online file-exchanges into legal ones, so it replicates almost all of the user-desired features of illegal P2Ps, such as unlimited downloads, streaming, and copying. A small difference is that DMX only allows a portion of its catalog to be used in derivative works, those for which permission has been granted. Illegal services have no such restriction.
Q: It seems that DMX is a uniquely cooperative venture that seeks to address the interests of both copyright holders and online consumers. What are some of the challenges that you’ve faced in taking this approach?
PH: One challenge is to gain credibility for the idea that a non-profit coop can replace the existing distribution channels of the entertainment business, which have been highly entrepreneurial and profit-driven.....
The DMX model suggests that companies that have been making money by distributing CDs and DVDs need to alter their thinking so that they can make most of their money from subscriber fees.
So far, it's been an easier sell than expected, because the income from CD and DVD sales is declining and highly eroded through piracy, while DMX can provide new and significant income to content owners very quickly. DMX's non-profit status and association with Harvard Law School are big pluses for content owners, who like having direct access to the usage and revenue information that DMX provides.
Another challenge to DMX was to invent an accounting system that tracks plays, downloads, and copies (on computers, CDs, cell phones, and PDAs) without encryption or technical protections.
A third and ongoing challenge is to attract wholesalers (ISPs, universities, and mobile phone companies), willing to pay monthly fees on behalf of their subscribers. Again, the buy-in has been much quicker and with greater penetration than expected, because DMX provides an excellent incentive for wholesalers to protect their organizations from litigation and to attract many new customers with the DMX services.
Q: To what extent do you anticipate aggregators of subscribers (e.g. universities and ISPs) to register for DMX and pay the monthly fees?
PH: Our first targets are subscriber aggregators in countries with high illegal content use and large populations. These represent the most lost revenue to content suppliers and the heaviest international pressure to clean up their illegal content activities. Since ISPs, mobile phone companies, and universities in these countries are usually tied closely to their governments, there is a big incentive for them to participate.
The impact of DMX in these countries will be immense, both for local and foreign content owners, who are currently earning almost no revenue from their copyrights in CD, DVD, and online media.
Initial response in China, the first DMX market, has been excellent. We've had three university meetings so far, signed up one immediately and the other two are in positive discussions. We've met with the ISPs that serve almost 100% of the broadband community in China (100 million broadband subscribers today, projected to have 200 million by the end of 2007) and they have both agreed to participate. We hope this buy-in will be replicated elsewhere.
To what extent are individuals going to be signing up for DMX themselves?
In the China example, all broadband subscribers will be DMX customers through their ISPs, so there’s no need to sign up individuals. Intranets (ISPs, mobile services, and universities) provide more efficient and less intrusive service to users than does the internet. For example, DMX needn’t require users to enter usernames and passwords, since these are already registered with individual user’s service suppliers. Internet service to individuals is DMX’s plan B, if the wholesale arrangements don't work.
Q: In exchange for posting their copyrighted property, content providers on DMX are paid according to a multi-variable algorithm. Could you talk a little about the algorithm and the considerations that were involved in formulating it.
PH: DMX royalty payments are based on a formula that incorporates several royalty distribution schemes, such as number of downloads; number of plays (experiences); a novelty bonus for recently added works; number of copies made to fixed and removable media and devices; availability of works for derivations; and royalties linked directly to individual subscriber payments and consumption. The weighting of each of these terms will be adjusted from time to time by the DMX governing council.
Q: Who’s on the governing council?
PH: Four reps from content owners, four reps from ends-users, four reps from Berkman, and up to four appointed reps who have particular technical, business, or other expertise.
Q: DMX makes a point of not requiring restrictive technical means of digital rights management (DRM). However, without technical DRM, how can DMX ensure that users will not reproduce and retransmit (possibly on other P2P services) content in violation of copyright law?
PH: DMX removes the incentive for illegal trading. DMX users can access and trade DMX content legally, safely, without degradation, and with many value-added services that are unavailable on illegal file-shares. And they can do it for free or almost free (depending on their service supplier). Some users may still use illegal services for content that is unavailable on the DMX system. That will create a natural pressure for non-participating content owners to license through DMX. As for non DMX subscribers, they can already get any content they want through illegal services, so they have no reason to get the same content illegally from DMX users.
Most content owners agree with this reasoning and are making their content available without DRM, which users hate. A few content owners still believe that the best way to protect and charge for their content is through DRM and encryption. In some territories, DMX will provide a separate tier of DRM-protected content, likely available as an additional pay-per-download service, for these content owners. The marketplace will determine whether a popular title on the DMX basic tier, paid for at a low rate by a very large subscriber base, will earn more than a popular title on a DRM-protected tier, paid for by a fraction of DMX users willing to pay more and deal with the inconvenience of DRM. If the non-DRM tier delivers more money to copyright holders, the DRM tier will wither.
Q: I was interested to see that a key element of DMX is a dispute resolution mechanism. Why did you originally see the mechanism as necessary and how does it work?
PH: DMX relies on information provided by content registrants about copyright ownership, authorship, and other metadata. But content rights are frequently bought and sold, and frequently for just some territories. There can be many authors, publishers, record companies, and film companies involved in a single work, and it's not uncommon that more than a single entity (person or company) claims authorship or ownership (two members of a band each saying they wrote a song, for example). In such cases of disputed authorships and copyright ownerships, DMX policy is to put the payments for disputed works into a suspense account and not pay either claimant until a) proof of authorship and/or ownership is received by DMX; b) both parties resolve the dispute and notify DMX; or 3) DMX's internal dispute arbitration service is able to facilitate a satisfactory conclusion. DMX will place suspended funds in an interest bearing account until the dispute is resolved.
Q: Your extensive background in both music and software engineering must make it particularly stimulating to work at the intersection of the two fields. How did you originally become involved in this area?
PH: From 1986-1989 I was Vice President of Research for a company that developed music compression algorithms that were the forerunners of mp3 coding. I became convinced that digital networks would be fundamentally disruptive to the distribution of entertainment content and the business of intellectual property rights management. My interest in the field was in part self-interest. Since my teens, I had supported myself and my family with royalty income as a songwriter, recording artist, film score composer, and later as a software and book author.
I had an appointment on the York University faculty as a Professor of Film and Music, and was able to found CulTech Collaborative Research Center at York in 1990, to investigate issues relating to digital content distribution, including technologies, impacts on users, and copyright law. When the World Wide Web was invented three years later, CulTech was well positioned to undertake content trials. I was able to raise $100 million from industrial and government sources to build a broadband connected community (1994-2000) and test consumer reactions to online distribution of music, video, and other content.
In 2004, I met Professor Terry Fisher, Director of Berkman Center, at a copyright law conference at which we were both speakers. I was immediately taken by his clarity of thought, application of research data, and analysis of the problems and potential solutions for the entertainment business. I was familiar with some of the excellent work done at Berkman Center and, when I read Professor Fisher's book, "Promises to Keep", became convinced that working at Berkman would be a wonderful opportunity to move forward with my research passions. Happily, I was asked to become a Berkman Fellow and now work on the DMX project, which is a concrete expression of my varied creative interests.
Q: What other projects are you involved in at the moment?
PH: I chair the Guild of Canadian Film Composers; I chair the Bell Broadcast and New Media Fund; I'm designing an online course for Berklee College of Music in Boston on composing soundtracks for videogames, Web, PowerPoint presentations, and mobile phones; I just finished a textbook on the same topic; I maintain my faculty position at York University and recently taught a course on culture and the arts; I'm on the boards of several foundations and a United Nations project on digital content; and I'm finishing a new DVD/CD with my band Lighthouse.
Q: Still performing music? Composing?
PH: I perform with my band Lighthouse once or twice a month and will be doing a tour with them and a symphony orchestra. Not much time for composing music these days.
Q: What’s the best album you’ve heard in the last couple of years?
PH: I've been returning to my jazz roots. There's a new 5 CD release of classic jam sessions at Carnegie Hall by an all-star band made up, illogically, of bandleaders such as Buddy Rich on drums, Oscar Peterson on piano, Charlie Parker on alto sax, Dizzy Gillespie on trumpet, Ben Webster on tenor sax, and so on. Never before released because the jams take place at the end of each “Jazz at the Philharmonic” concert (circa 1954), without rehearsal, sometimes just a series of chord progressions with no song at all. Everybody improvises his guts out for about 20 minutes and then they stop or sometimes just peter out. Awesome. Downloaded from iTunes. Put it on my iPod. Great driving music. Improvisation is what jazz is about.