Thursday, June 24, 2010

CIP: Google Book Search and User-Generated Content

Panel 1: Reuse, Recycle, Rethink: The Impact of Google Book Search

Panelists:

Peter Jaszi (Moderator), American U. Law: Scanning as benefit to libraries and incidentally to authors (for whom snippets might lead to purchases)—copyright owners however saw it differently. Note massive evolution of project from inception to settlement. One key question for discussion: is the GBSettlement good for culture?

David Balto, senior fellow, Center for American Progress: Antitrust issues are the teeny tail on the big dog. Ask who is complaining? You can tell whether something is good or bad for consumers by answering that question. The loudest complaints against the GBS come from Amazon, which wants to charge a fortune for access to the books Google wants to provide more cheaply. If a competitor is complaining, it’s good for consumers. Settlement is good, thinks it will be approved by the court.

Lateef Mtima, Howard Law: This is just one move in the ongoing readjustment of rights that characterizes the history of copyright. (Though I note that he spoke about what Congress has done to respond to technology in the public interest much more than what private actors have done; Jessica Litman’s writing on the ways in which private actors have simply decided to ignore the law as written is relevant here.) Among the things we need to take into account are subdivision of rights in new ways, combined with term extension—how do you find/aggregate the rights? Even if you know the publishing co. and the co. is still in business, it might not have the e-publishing rights.

When we don’t know what the lost author would want, what should we do? Should we presume the author wants wide dissemination? Or should we presume she wants to maximize control in the hopes of maximizing revenue? With individual authors as owners of rights, the former is a better presumption—approach orphan works from an opt-out perspective. Marybeth Peters says that turns copyright law on its head, but he thinks it respects authorship more. We don’t have the author in front of us, as the copyright law contemplates; we have to interpret and guess what the author would want. In keeping with the law’s objectives, we should presume the author would go for dissemination.

History of social justice issues around IP—African-American entrepreneurs unable to capture full value of their work because of refusals to deal by whites/lack of access to credit. Can’t ignore that in talking about rights.

Maria Pallante, Copyright Office: Doesn’t think the settlement will be approved. Expects a roadmap from Judge Chin for an acceptable settlement within the framework of class action law. The DoJ statement of interest is a little divided; starts with praising the settlement’s goals—a lot of support for a registry that would allow licensing and micropayments; pressure on authors/publishers to make themselves known (is this impermissible formality under Berne?)—people are a little impatient with the perspective of some rightsholders that they don’t need to do anything with a published work to make themselves known/findable. Settlement would help with access for the blind.

Settlement also creates a new procedure for orphan works. Right now the law is strict liability no matter how good your search is; on the table is a legislative model that would remove liability for individual diligent searches. But clearance on a large scale with multiple rightsholders is infeasible. She thinks that the registry system is something for Congress to look at, not the courts. To be fair, the argument is that Congress hasn’t done so, and Google has; the government doesn’t agree that this is proper. It would create a benefit just for Google—ability to use works with no liability; could anyone ever catch up? Just because Google invested in a fair use argument, does it deserve this?

Government position: scale it back; scope is too broad. Underlying litigation was a fair use debate: scanning books for a plausible fair use, snippets. Before the court is a broader proposal to sell books, sell subscriptions, etc. Google couldn’t have done this legally without consent. This is a class action question, not a copyright question. The settlement isn’t sufficiently tied/anchored to the underlying litigation. Scale it back, settle what was at issue; create new business models, but you can’t force people into them. Settlement rewrites authors’ contracts from 1923 on—the contracts don’t talk about digital rights; publishers don’t know if they own the rights; GBS just says by fiat how the split will work. Members of the class didn’t knowingly hire these parties to rewrite their contracts going forward.

Orphan works—adequacy of representation. Google scanning is ongoing, which seems like a compulsory license for one entity. If Google hasn’t scanned your book yet, do you have a ripe claim? Similar with foreign authors: if you’re affected by this but don’t know it, there’s an adequacy of representation issue.

Balto: Doesn’t trust government involvement with innovation. What Google’s done is create the landscape for access to wide varieties of information; other alternatives will arise. Even if that wasn’t true, he can think of few endeavors with as much risk and requiring as much investment as GBS; hard to see Google’s monopoly profits. Google’s being forced to protect authors’ financial interests, not its own monopoly profits. Google is also taking steps to get permission from authors.

Scope of the initial claims: frequently, in the course of litigation, settlements are prospective, settling the underlying conduct. You don’t want just to stop one instance of bad conduct, but you do want to resolve future conflicts.

Pallante: To clarify, she doesn’t know whether Google seeks permission. In the scope of the settlement, prior search would not be necessary (that is, if I understand her correctly, after the Registry tries to find the owner and fails, you don’t need to update the search ten years later; I think also she/the Copyright Office objects to the fact that scanning proceeds before the search is complete). The Authors Guild says that they think they’ll find most of the authors if there is a check waiting. There’s some experience with photocopying, trying to find people to whom royalties are due; some recipients freak out and think it’s a scam. It’s all about scale and who the beneficiaries of scale should be. We also have international treaty issues. Suppose we decide we love opt-out so much we want to extend it to universities/libraries (don’t other countries have these regimes?).

Mtima: The problem with much of the opposition is that 98% is too theoretical. (Like the possibility of a massive oil spill from offshore drilling? What could possibly go wrong with outsourcing regulation to the private sector? Okay, look, I am more on the side of GBS than not, but it’s far from crazy to worry about unmaterialized-as-yet risks.) Show him a perfect class action resolution. Also, remember that copyrights are not absolute. Real property rights are limited too. The author doesn’t have the exclusive right to determine who will read her work. Anyone who can get ahold of the book can read it. Reproduction and distribution, on the other hand, should be compensated.

Balto: there’s something disconcerting about our two alternatives. The chances of Congress effectively grappling with these issues is the chance I have of flying unaided to the moon. Then GBS is an unwieldy class action structured by a group of lawyers (unlike law in Congress, of course).

Pallante: Government agonized over this, but concluded that class action law doesn’t let you go this far beyond the underlying dispute. A properly defined and adequately represented class can settle a lawsuit over past conduct and also license a somewhat broader range of conduct. Nevertheless, class action settlements must be subject to reasonable limits to ensure they’re resolving actual controversies. Fullscale text sales weren’t within the scope of the controversy.

Why should Google be able to sell for profit books that were once in libraries due to taxpayer money, and libraries can’t? Why not have an opt-out system that would benefit everyone, and start with educational and library uses?

Mtima: He doesn’t think it has to be only Google.

Pallante: for others to do it, we need legislation, and legislation that broad is very unlikely (see: orphan works). We have international standards that are of concern.

Mtima: there are so many business models, and so many assertions that they were infringing, many of which fell apart when tested. We shouldn’t let claims that “you’re building a business through infringement” be dispositive. Many scholars wish Google would have stuck with its fair use claims.

Balto: He agrees this is a public good, but he’s suspicious of claims from proponents that we’ll get another opportunity (like an overeager salesperson). If it fails, Google will either have to litigate or lobby. Wouldn’t having more players at the table for fair use or legislative reform raise the likelihood that we’d get a decent settlement?

Pallante: Europe has accepted that mass digitization is a social good that should be explored with an acceptance of cost. Congress doesn’t yet grasp the goal of mass digitization. In Europe, she asked the head of the Copyright Union about orphan works—they want to do it before the Americans digitize their culture.

Panel 2: The Changing Landscape: Social Media, P2P Filesharing and Culture

Rebecca Tushnet (Moderator)

Questions: What is really happening on campus with filesharing and similar activities? What is this piracy people are always talking about (Adrian Johns’ recent book Piracy makes a fascinating case that IP is defined around, and after, “piracy”), and what should people in the educational community be thinking about, especially knowing that teaching and research are likely not what’s first on the minds of the people designing and implementing new technologies and business models? In particular, how should we think about the new legal requirements in the HEOA and the idea of “graduated response,” the PR-based renaming of “three strikes” proposals?

Greg DePriest is VP, Technology Policy for NBC Universal, where his main responsibilities lie in the area of digital content protection.

How big a problem is piracy? Every 16 seconds we detect & verify our content being shared/transferred on P2P network within the US. The US is only 10% of the global internet piracy issue, and we detect only about 10% of the infringement out there—we feel pretty confident about those numbers. Sysco, P2P is growing 16% through 2014; 75% more than today; roughly 39% of consumer internet traffic at the end of 2009. One click hosting is also growing (compare to P2P), which means cyberlockers and streaming sites like YouTube—there will be a transition from P2P to these other means.

On campus: Illinois State, April 2007. Mainstream: 42% of students in residence halls engaged in likely infringement, most likely music; 24,000 titles; each user averaged 18 titles; no legitimate use detected.

Huge problem: need to use tech to fix a problem that tech created. Three snapshots of our content on UGC sites like YouTube, DailyMotion, Veoh. In mid-2007, most of the content was found on US sites. Early 2008, UGC sites began to check before posting and filter out copyrighted content depending on wishes of copyright owner; NBC content fled to China as of mid-2009, with Megavideo an aberration—servers in US but headquartered offshore. Lesson: Tech works. (Is that the lesson?)

Recommendations: establish a tone that filesharing is unacceptable. Teach students to remove P2P software from PCs; students may not know it’s running. Use existing tech; don’t forget about your wireless networks when developing a plan to effectively combat filesharing—you can choose your own metric: recidivism; number of notices you are receiving. May wish to separate residential network from research network, allowing P2P on the latter. Block P2P and provide students with a day pass to do WoW upgrades or Linux or the like. Other schools use a box to detect transfer of copyrighted content and send students an email warning them, keeping track of who’s been dinged and escalating if there’s no change. Reduce bandwidth available for filesharing. Use DNS to warn students before permitting them to visit known pirate sites like Rapidshare, MegaUpload, OpenBittorrentTracker, etc. Michigan uses a system: be aware you’re uploading—sends an email to students who are detected using P2P systems.

Jim Burger, attorney at Dow Lohnes specializing in representing technology companies on IP, licensing, communications and government affairs, thus everything he does is relevant to this panel. These views are however his own.

Education on copyright is desirable. HEOA is misconceived and we are happy to tell you offline how to comply. But the problem is defining the threat. Threats are relative, shifting, and depend on context. What’s the impact? We have no proof that the business/economic problem is a big one, though there is no doubt that there is some loss. Multiple studies going multiple ways. GAO took a hard look and concluded that there’s a sizable problem but we don’t know the economywide impacts; difficult if not impossible to quantify. Huge numbers is not the same thing as huge impact. The copyright law is designed to spur creation; what is happening given this incredible amount of piracy? Book publishing surged, albums doubled, movies globally went up by 33%.

What happened to the record industry, then? There is a problem: a decline in music. It wasn’t until they killed Napster in 2001 that the industry started to lose sales. But also in that first period, DVD sales went through the roof. Then music singles sales went through the roof. People just started spending entertainment money differently. (This reminds me of the debate over newspapers. Piracy/copying isn’t the real problem; the problem is that people don’t want to pay for most of what newspapers want to sell.) Infringement online is wrong, and requires educational response, but also requires business models. P2P is passe; numbers are declining rapidly. Other things are taking its place—including legitimate cheap rentals and streaming. Netflix—friend or foe? If you can rent a blockbuster for $1, then we become a nation of renters; serious business impact. Netflix will be in 10% of TV households by 2010, and projection is that DVD shipments will fall off and streaming will replace it exponentially fast.

Time Warner says transition to Blu-Ray will be extremely good for us. But there are also concerns about trading analog dollars for digital pennies. At digital dimes now, but that leaves a lot of ground uncovered. Problem: sale is more profitable than rental. Electronic sellthrough in particular is super-profitable. (I paid $9.99 for Up for Grabs, a great movie about baseball and property law; that’s hard to see as anything but pure profit for the distributor.)

HEOA tells universities to pay to protect their property. If their numbers are right, why wouldn’t they pay a few million for you to install the tech to save them billions? He objects to cost-shifting to universities. We don’t even know how much money is lost. If the movie industry doesn’t execute correctly, it will do badly, but it’s not the fault of those meddling kids and their dog.

Jim Griffin, who operates the OneHouse consulting firm: self-describes as “the digital media guy.” His work focuses on the digital delivery of art, and before OneHouse his activities included running the tech department at Geffen Records.

We need a sustainable economy of ideas. We need to make it faster, easier, and simpler to pay for ideas, art, knowledge, and culture. We’ve made it extraordinarily difficult to pay. To the extent that business model depends on stopping copying, or getting paid when copies are made, we need a new business model. Few people now think of photocopying as requiring compensation, and yet copyright law does. But we don’t have controls on photocopying machines; tech advance has made that irrelevant/impossible.

It’s alarming that we can say that paying for music/movies/books is clearly voluntary. It’s not required. A civilized society can’t survive long with payment for art, innovation, culture purely voluntary. Yet it’s still more abhorrent to condition access on the size of someone’s wallet or their parents’ wallet. We shouldn’t use money to allocate the fruits of knowledge; then we are deprived of those seeds being planted. If we proposed libraries today, the believers would clearly be considered communist for taking tax money, buying things for people, and giving them away. A child should be able to read any book, watch any video, hear any song without worrying about paying; we are the ones who benefit.

Rising tide of digitization cuts a shorter path from source to destination, just as with a river; those who were at the bends are undoubtedly sad, as are those in the path of the new, more direct route. So long as that content is a product, we’ll think of digitization as theft. He hates the term “consumer.” A house is consumed by fire. Consuming decrements supply, as buying a physical record did from the store’s inventory. A downloader grows supply. We are transitioning from product to service; let go of the notion that there is less after we have taken some.

Part of the problem is gender. The content industries have largely been run by men. The notion of consummating serial relationships with consumers without even knowing their names is a guy’s idea of a good idea. Amazon is a woman; it remembers your size and eye color; it wants to start a relationship that never ends. We have to be in the business of making relationships; but we can’t let go of product just yet when we’re making billions off selling objects and the installed base of players for those products will last past his lifetime.

Creativity is moving from center to edge. Licensing was formerly something for the center of the network: the cable head-end, the TV network. You could take a compass and draw a line around distribution. Today the power of the press has moved to the edge. A woman who picks up a HD camera to watch her toddler prance around the carpet with a song playing in the background should not expect to know she’s violated a licensing regime; that’s ridiculous. Campus networks too are disappearing as increasingly devices are wireless and hooked into a GSM network. Controlling copying on campus is hardly the point when most students live off-campus. If you solved the problem of access on-campus, it would need to follow them on vacation, visiting parents, etc.

Rightsholders must enumerate their rights to be respected. We have to have registries if we have to ask permission. Harry Fox Agency has less than half the content and won’t even stand behind the list it offers if you get sued. We have no registry and until we enumerate our rights we can hardly expect them to be protected.

It is essential we give rightsholders antitrust and competition relief to work together, coming together into pools that make licensing easier, faster, and simpler. (Comment: not cheaper? No, didn’t think so.) Sports leagues have exemptions, but not music publishers—even though you can’t find half of the relevant publishers. A recipe for disaster.

It’s not just music or ideas. Working with Mendelay: software that treats PDFs the way the rest of the world treats mp3s. Let’s help people organize files and find/share scholarly research. You know when you see a popular song on a network that it is probably owned and shouldn’t be shared without permission, but a scholarly work is different: may be open access per its funder; may be owned by the author who wants it shared; etc. There have been no lawsuits over sharing scholarly materials. And it’s going 3D: a kid in Sweden scanned his father’s handcuff key and released it on a filesharing network: now anyone with a fab lab ($3000) can make that key. Now apply that to medical instruments or medical tests. Should we stand in the way of sharing ideas for medical tests? Don’t let focus on music and movies distract from the overall environment.

RCA Victor created the lateral cut disc for music; decided jazz was not music. Little company sued for the right to make lateral cut discs. Learned Hand said plaintiff deserved a license; result: first two albums were Jelly Roll Morton and Louis Armstrong. We’ve seen this problem solved before; the anomaly will be if we don’t use known methods of solving it. Acousticàelectric was a far greater change than analogàdigital; in an acoustic world, an artist could only be heard in the same room. Loudspeakers, then radio, then TV, then cable and satellite: every single medium was licensed with a collective license, a pool and a way to split it up, not with control. All paid for without control: actuarial blanket licenses. We should address copyrisk the same way we address other risks: pooling money as in an insurance fund. We can’t stop copying or count every copy that gets made.

My Q: Given that content in the charts DePriest showed seemed to migrate to China, does it show that tech works or that law is the key?

DePriest: the UGC sites that elected to abide by UGC principles, primarily in US and Western Europe, and then our content moved to China. Localizing content in China is not so bad because then it can go government-to-government.

Burger: it’s a game of whack-a-mole, though. The issue is product v. service, and the problem is that we’re used to huge profits from products. Asking universities to put in physical controls that kids will work around is a waste of time and money.

Griffin: attack the motive and not the mechanism. We destroyed people’s motives to copy with widespread licensing—you could now get Star Trek easily without copying at home.

DePriest: DVRs?

Griffin: we let it go after we watch it. We don’t make huge video collections anymore.

DePriest: video is the only justification for a terabyte hard drive. Whack-a-mole is a fact of life. Of course tech isn’t static; why should infringement tech be any different from virus software or any machine with an operating system?

Burger: the problem with whack-a-mole is who pays for it? Why shouldn’t the content companies pay for whacking the mole, not the university? Usually you pay to protect your own property. The bus company isn’t asked to pay for security guards at the mall because the bus brings shoplifters.

Griffin: problem is that you can’t find outlier copyright owners; major content owners are willing to grant site licenses, but they’re not blanket enough to be worthwhile in many cases. (This conversation is the mirror image of the GBS discussion previously.)

DePriest: it’s your network as an educational institution, you should pay to protect it. Flows on the network need protection too. (Are they your flows?) NSF gives away hundreds of millions a year to protect networks, and nothing to protect the data on them. An interstate highway with no state troopers. (Note that the state trooper analogy is a bit different—it’s about using law to change behavior in the physical space; speed bumps are a different type of response to behavior you don’t like. I cover this a little in My Library.)

Q: why single out higher ed? We’re 4% of file sharing in America; why single us out? About 90% of the notices I get (identifies as DMCA agent) are for activity on the wireless network. He doesn’t actually have to do anything about that; can really only identify 5-10% of those people; and those people presumably go home to other network providers like Verizon, but those providers aren’t under the same constraints as campuses.

Burger: because they could.

DePriest: We’re all in this together. Need cooperation from ISPs, whether universities or Verizon, and device makers. We want Verizon to hold its subscribers accountable. You need some authentication on your wireless network to extend the teachable moments.

Q: Worried about licensing substituting for fair use: why should you have to pay if the use is fair?

Me: I agree that this is a major theoretical, even moral, problem with some proposals. On the other hand, a system that roughly tracks copying could probably get acceptable practical results. You should always be able to avoid paying for your parody of Gone with the Wind or Sarah Palin's book.

Burger: Most of these proposals are really targeted at private use (rather than transformation).

No comments: