I want to dilate slightly on a point I made in a recent post at the Cato blog thinking about the Megaupload takedown and online innovation:
The last innovation is always safe. That’s why it’s easy to claim concrete examples of the harm regulation might do are hyperbolic fearmongering: Nobody’s going to shut down YouTube or Twitter now, because we’ve already seen the incredible value creation they enable, even if they also make it a bit easier to infringe copyrights. And anyway, the success stories eventually get big enough to afford their own fancy lawyers. It’s the next platform that we risk strangling in the cradle, because every new medium starts out recapitulating old media content before it becomes truly generative. Early radio is full of people reading newspapers and books out loud. Early TV and film looks like what you get when someone points a camera at a stage play.
An important thing to understand about innovation that involves computing technology, annd in particular innovation in Web platforms, is that it’s user-driven to a much greater, much faster extent than we’re accustomed to. It is, of course, impossible to predict how a major new technology like the car or the cell phone will change society over the long term, but you have a pretty good idea what the technology is for at the time it’s introduced. I doubt Ford anticipated drive-in theaters, but he knew that the Model T was a machine for getting you from point A to point B.
Online innovation isn’t really like that. The most important new platforms of the Internet era have essentially opened up a space and seen what thousands or millions of users communicating with each other want to do with it. Think of the early phases of Twitter’s popularity, when there were plenty of pundits scoffing at this fad for letting people know what you had for breakfast—which at first was not a wholly inaccurate picture. Nobody anticipated that it would play a pivotal role in global anti-authoritarian revolutions, or that CNN would soon be mining it as a kind of realtime Zeitgeist thermometer during debates and primaries.
On the Internet, you don’t know what a new technology is for until you see what people do with it. Biologists have a saying: “Evolution is smarter than you are.” They mean that even a fundamentally blind process of trial and error, iterated over millions of years, will come up with better solutions to adaptive problems than even the cleverest primate product of that process. Smart tech entrepreneurs know how to take advantage of the corrolary: The Internet is smarter than you are. Online innovation is a collaborative process, and the most interesting uses the users find for your platform won’t neecessarily be the ones you intended. Indeed, that’s the guiding idea behind the “end to end principle” that has made the meta-platform of the Internet itself so incredibly generative. The pipes are, more or less, dumb: They know how to pass packets. Everything actually intereresting about the Internet arises from what people do on top of that dumb infrastructure, and it works because the pipes are basically open. A coder with a new idea doesn’t have to convince someone to enable the functionality they’re interested in implementing—a feed of 140 character messages? come on, who needs that?—they just try it, and the ideas that provide value spread like wildfire.
One corrolary of this is that the initial phase of deployment for a new platform is typically going to be the “what I had for breakfast” phase—the experimental phase when people are trying out a lot of different things and figuring out what the platform is good for. And very often, the first thing it will be good for is copyright infringement.
Consider YouTube for a minute. Seven years in, we see a dizzying array of creative purposes YouTube is good for. There’s professional quality programming produced specifically for YouTube, fantastically creative amateur mashups, lip dub videos, aspiring Biebers singing cover songs, amateur talk shows and game reviews and diaries, cute kittens and playful toddlers, kids showing off their light-saber skills, and a surprising number of people just unpacking the loot from their latest trip to the mall. We have an online video culture——a meme culture, if that’s not too redundant—that provides a thousand models for the kind of thing someone with a webcam and consumer video editing software might want to do with the platform… along with the realization that you, too, could even create a whole new genre.
In 2006—on YouTube, day one—none of that existed. We didn’t even have a mental model of what mass amateur video creation would look like, never mind the actual content to populate the platform. What did exist was a lot of copyrighted music and video sitting around on people’s hard drives. Needless to say, it would not have been feasible on day one for the platform to effectively pre-filter every user upload and determine which videos infringed someone’s copyright. A filter algorithm wouldn’t be able to tell the difference, up front, between Fair Use and simple infringement, even if it had amassed in advance an enormous and constantly updated database of content samples to compare uploads against. (And of course, when the platform isn’t limited to a specific type of content—like video or music—a filter can’t do much of anything with a compressed or encrypted file.) The only way for the platform to launch at all, in anything remotely like the form we know it, was to launch open.
I don’t clearly recall what all was on YouTube in those early days, but I’d wager the proportion of copyright infringing material was a good deal higher than it is today, because on day one, that’s most of what there was to upload. For those who figured out how to download or capture the underlying video files, though, those clips would provide the raw material for a lot of the creative, transformative fair use output that is now virtually synonymous with the platform.
We can even see something analogous in older industries. Japanese companies started by making cheap knock offs of American technology, then learned enough to start innovating themselves. One rather telling passage in a 2010 GAO report on counterfeiting and piracy suggests that the most important effect of “counterfeiting” is actually knowledge transfer: The thing that enables the copiers to rapidly transform themselves into legitimate competitors.
We can boil all this down to a few basic principles:
(1) In their early phases, new and open platforms for user-contributed content will tend to look a lot like tools for copyright infringement, because copying naturally precedes creation.
(2) Generative platforms need to “launch open” in order to fully harness the creative and innovative capabilities of a distributed user base.
(3) This creates a tradeoff between copyright enforcement and generativity. A regulatory regime that is more protective of one interest will tend to yield more “collateral damage” to the other.
(4) The collateral damage to copyright interests will always—and especially initially—be more visible than the damage to generativity. Copying that does happen is apparent, but there’s no way to detect user-driven innovation and creation that doesn’t happen. If you knew what to expect in advance, it wouldn’t be innovation.
(5) The value of existing intellectual property, its production mechanisms, and its distribution channels, is already apparent. The owners of those assets can readily identify themselves and each other, making it relatively easy to organize to advance their interests, and point to visible harms inflicted by open platforms. Beneficiaries of future generativity are less easy to organize because they don’t know who they are yet.
(6) This creates institutional and political bias in favor of protecting IP at the expense of generativity. The bias itself is undetectable because its primary manifestation is a relative absence of change.
(7) Attempts to strike a calculated balance between these interests on the basis of what we currently know about the uses of technology will tend to get it wrong. Hence SOPA, PIPA, PRO-IP, ACTA, and all the rest.