Thursday, January 10, 2008
Konflating Kopyright and Kongestion
In June, 2007, AT&T held a meeting of Hollywood movie and music moguls at which it announced plans to monitor Internet traffic on its network "to police copyright infringement." I have blogged it several times, e.g., here.
New details of AT&T's private policing plan emerged this week at CES. [Here's one report.]
AT&T frames the "problem" in a way that conflates two separate issues. The first is network congestion. The second is copyright violation. An NBC representative makes this conflation obvious when he says, "The volume of peer-to-peer traffic online, dominated by copyrighted materials, is overwhelming . . ." [That quote's here.]
The two issues should be separated.
Network congestion has some very simple solutions. First, replace those few network nodes or facilities that are undersized. (See work by Odlyzko & Coffman, 1998, and Odlyzko, 2003 [.pdf] for supporting data.) Second, where up-sizing is not possible, or not cost effective ( which is almost always at the edge of the network) it is simple for an Internet access provider to explicitly count packets and either (a) throttle data rates, or (b) offer explicit tiers of service so that more throughput costs more. The carrier can do this without knowing what the data represent; in fact, it is easier and cheaper to simply count packets than it is to look inside them to infer what the ones and zeros represent.
The other problem, copyright violation, is more subtle. It is perfectly legal to send copyright materials over the Internet under the doctrine of fair use, and it is impossible to examine the materials to determine whether fair use applies. AT&T spokesman James Cicconi implies that AT&T will make that determination after alleged violations have been automatically flagged. Such a kill-them-all-and-let-God-sort-it-out approach would shift the costs of enforcement onto society at large, including the creative community that copyright was designed to protect in the first place. Even if it is done in a "friendly way," as AT&T's Cicconi proposes.
However, there are simple solutions to the copyright problem. If a flagrant copyright violator is, indeed, making money from materials he or she does not own, that violator would need to advertise or otherwise make a public offer to sell the offending stuff. The public offer is publicly detectable, and this is where copyright owners and their agents should act. There are other solutions too, e.g., takedown as mandated by the DMCA. (Whether or not you buy into the specifics of DMCA takedown, at least it addresses content at the application layer, without involving the network's infrastructure.) In any case, the question, "Do the benefits to music and movie moguls of let-God-sort-it-out enforcement out-weigh its costs to society?" is one that we can address independently of network congestion.
David Weinberger says "the carriers are the last people I trust to make decisions about what’s important and acceptable." But even he, author of truly great books, conflates congestion with copyright, and this draws a comment from Seth Finkelstein, who takes network congestion as a given (which it isn't!) and asks, "who do you nominate to make [the] decision" about what traffic to block or delay?
I nominate the end-user to make the decision. If, as I propose above, Internet access providers were to provide explicitly different tiers of service for different, explicitly laid out throughput plans charged at different rates, the user could make that decision. And it would be a free market decision that even a libertarian would love. The carrier would need to decide whether to (a) eat the cost of upgrading its infrastructure to allay the risk of losing customers to bad performance or (b) implement tiered service to save network upgrade costs but risk losing customers who don't like paying for tiered service. But, hey, that's business.
Carriers know all this. But, as it turns out, they don't want to frame the picture simply because they don't want to be in the Stupid Network business. They're addicted to adding value. They've convinced one class of customers, music and movie moguls, that they can add value for them. Importantly to the carriers, the value they propose to add is network-resident value; the return of the Intelligent Network. Network-resident value is business as usual, even when it imposes huge costs on the rest of society. This is why they persist in framing the two issues as inter-related.
When they persist, I propose we point this out by spelling the conflated issues "kopyright" and "kongestion."
We who appreciate the end-user value of the Internet should do our best to keep the two issues, congestion and copyright, separate.
UPDATE: Harold Feld weighs in.
New details of AT&T's private policing plan emerged this week at CES. [Here's one report.]
AT&T frames the "problem" in a way that conflates two separate issues. The first is network congestion. The second is copyright violation. An NBC representative makes this conflation obvious when he says, "The volume of peer-to-peer traffic online, dominated by copyrighted materials, is overwhelming . . ." [That quote's here.]
The two issues should be separated.
Network congestion has some very simple solutions. First, replace those few network nodes or facilities that are undersized. (See work by Odlyzko & Coffman, 1998, and Odlyzko, 2003 [.pdf] for supporting data.) Second, where up-sizing is not possible, or not cost effective ( which is almost always at the edge of the network) it is simple for an Internet access provider to explicitly count packets and either (a) throttle data rates, or (b) offer explicit tiers of service so that more throughput costs more. The carrier can do this without knowing what the data represent; in fact, it is easier and cheaper to simply count packets than it is to look inside them to infer what the ones and zeros represent.
The other problem, copyright violation, is more subtle. It is perfectly legal to send copyright materials over the Internet under the doctrine of fair use, and it is impossible to examine the materials to determine whether fair use applies. AT&T spokesman James Cicconi implies that AT&T will make that determination after alleged violations have been automatically flagged. Such a kill-them-all-and-let-God-sort-it-out approach would shift the costs of enforcement onto society at large, including the creative community that copyright was designed to protect in the first place. Even if it is done in a "friendly way," as AT&T's Cicconi proposes.
However, there are simple solutions to the copyright problem. If a flagrant copyright violator is, indeed, making money from materials he or she does not own, that violator would need to advertise or otherwise make a public offer to sell the offending stuff. The public offer is publicly detectable, and this is where copyright owners and their agents should act. There are other solutions too, e.g., takedown as mandated by the DMCA. (Whether or not you buy into the specifics of DMCA takedown, at least it addresses content at the application layer, without involving the network's infrastructure.) In any case, the question, "Do the benefits to music and movie moguls of let-God-sort-it-out enforcement out-weigh its costs to society?" is one that we can address independently of network congestion.
David Weinberger says "the carriers are the last people I trust to make decisions about what’s important and acceptable." But even he, author of truly great books, conflates congestion with copyright, and this draws a comment from Seth Finkelstein, who takes network congestion as a given (which it isn't!) and asks, "who do you nominate to make [the] decision" about what traffic to block or delay?
I nominate the end-user to make the decision. If, as I propose above, Internet access providers were to provide explicitly different tiers of service for different, explicitly laid out throughput plans charged at different rates, the user could make that decision. And it would be a free market decision that even a libertarian would love. The carrier would need to decide whether to (a) eat the cost of upgrading its infrastructure to allay the risk of losing customers to bad performance or (b) implement tiered service to save network upgrade costs but risk losing customers who don't like paying for tiered service. But, hey, that's business.
Carriers know all this. But, as it turns out, they don't want to frame the picture simply because they don't want to be in the Stupid Network business. They're addicted to adding value. They've convinced one class of customers, music and movie moguls, that they can add value for them. Importantly to the carriers, the value they propose to add is network-resident value; the return of the Intelligent Network. Network-resident value is business as usual, even when it imposes huge costs on the rest of society. This is why they persist in framing the two issues as inter-related.
When they persist, I propose we point this out by spelling the conflated issues "kopyright" and "kongestion."
We who appreciate the end-user value of the Internet should do our best to keep the two issues, congestion and copyright, separate.
UPDATE: Harold Feld weighs in.
Technorati Tags: AndrewOdlyzko, AT&T, Competition, Content-Conduit, Copyright, Customer, DavidWeinberger, DeepPacketInspection, FairUse, Framing, IncreasingReturns, NetworkNeutrality, ParadoxoftheBestNetwork, savetheinternet
Comments:
Post a Comment
