Today, the Congressional Internet Caucus Advisory Committee will host its annual State of the Net Conference. One of the panels is titled “Internet Copyright Filters: Finding the Balance.” I suspect it will feature episodes of a peculiar condition now impeding public-policy analysis. This condition can be called “technological schizophrenia,” or “technophrenia” for short. Despite the name, technophrenia is less a mental disorder than a context-specific example of doublethink.

Technophrenia arises from a brute fact of modern life: Most technologies, while advancing, are still imperfect—they don’t do exactly what we might want them to do in a perfect world, so they inevitably have costs, as well as benefits. Assessing the relative costs and benefits of imperfect technologies could be fairly said to be the key to most debates about copyrights and digital technologies.

Technophrenia inhibits this assessment process by imposing radically different standards depending upon whether a given technology promotes copyright infringement or copyright protection. The technophrenic claims that infringement-promoting technologies must be tolerated and embraced because their profound imperfections might someday decline. But imperfect technologies that protect copyrights are intolerable—even if their minor imperfections would decline over time—because, in the interim, some small quantum of lawful or noninfringing use might be deterred.

To see technophrenia in action, consider how the public-interest group Public Knowledge (“PK”) confronted two questions about technologies whose relative levels of “imperfection” differed most profoundly.

In Grokster, PK confronted two of the most “imperfect” infringement-promoting technologies that have ever existed—the Grokster and Morpheus file-sharing programs. According to the federal courts, Grokster presented “clear,” “replete,” “overwhelming,” and “unmistakable” evidence that the defendants were for-profit corporations that intended to profit from copyright piracy by inducing, (i.e., encouraging or duping), teens, college students, and others to break the law. The distributors of Morpheus claimed that this business model was great because it gave them “no product costs to acquire music,” and “the ability to get all the music.”

But, by the time Grokster reached the Supreme Court, it had also attracted the attention of the United States Department of Justice. Then, the defendants realized that there might be something problematic about a real-world implementation of the “business model” of Bill Sykes, the villain of Oliver Twist. Citing “fear of a criminal investigation,” they thus refused to let amici look at the record. As a result, potential amici could defend the Grokster defendants only if they were willing to have no clue about what the defendants had actually done, intended or said—other than that the defendants themselves feared that it was potentially criminal under then-existing law.

Naturally, Public Knowledge leapt to their defense. In its Supreme-Court amicus brief, PK argued that the Grokster defendants should not be civilly liable for potentially criminal conduct: After all, if the defendants continued inducing infringing uses of their programs, copyright holders could just continue to sue into bankruptcy the families of every teenager or college student that the defendants successfully induced. PK also argued that even if the Grokster defendants intended to profit from infringing uses of their programs and even if they implemented their intent so effectively that their programs were used to download infringing files 97% of the time, they should not be liable because (1) 3% of the uses of their programs were not “highly likely” to be infringing, and (2) someday, the percentage of not-highly-likely-infringing uses might rise from 3% to 10%.

One might thus assume that Public Knowledge takes an indulgent view of imperfect technologies. Not really. PK will tolerate a technology’s 97% rate of imperfection only if that technology promotes copyright infringement.

By contrast, if a technology—even the same technology—prevents copyright infringement, PK will find even minute imperfections unacceptable. For example, recently, many user-generated content (UGC) sites like MySpace and YouTube committed to reduce infringing uses of their sites by using filtering technologies, most of which generate mathematical “fingerprints” of uploaded files and compare them against databases of exemplars submitted by copyright holders. Essentially, these technologies are far more sophisticated versions of the “file hash” technologies that facilitate infringing use of file-sharing programs like Grokster and Morpheus.

How effective should these technologies be at preventing, rather than promoting, infringement? While we will need real-world testing to assess their error rate, they should—at the least—be mirror images of file-sharing networks: At first, they might be only 90% effective at identifying infringing uses of works, but they might readily be expected to achieve 97% efficacy as they are fine-tuned. And make no mistake: Some proponents of these technologies will complain that I underestimate their efficacy.

Needless to say, PK was horrified that more effective versions of a technology that had once promoted copyright infringement might be used to prevent copyright infringement. In “Google Blinks and the Internet Is a Little Less Free,” PK President Gigi Sohn condemned their use.

That is technophrenia incarnate: If digital file identifiers are used to promote copyright infringement 90% of the time, they are OK with PK. If they are used to prevent copyright infringement 90% of the time, then they suddenly become unacceptably imperfect.

Oddly enough, the content industries often accused of “not getting” new technologies tend to be far less prone to technophrenia. In both the anticircumvention provisions of the DMCA and in Grokster, they favored standards that imposed liability based upon intent or predominant infringing use. In short, they were willing to tolerate technologies that, while far from perfect, were not predominantly piracy machines whose continued distribution would have the inevitable effect of redirecting copyright enforcement toward consumers—the result favored, in Grokster, by Public Knowledge, the Consumer Federation of America, Project Gutenberg, the Internet Archive, college librarians, and the distributors of the Grokster, Morpheus, and LimeWire file-sharing programs.

posted by Thomas Sydnor @ 9:42 AM |

Link to this Entry | Printer-Friendly | Email a Comment | Post a Comment(0)