File sharing
File sharing and social software
Clay Shirky's latest essay, "File-sharing Goes Social" took me by surprise when it showed up in my mail on sunday. Clay's essay resonates with me -- I had been thinking about the same situation for weeks. I even submitted a talk proposal to the O'Reilly Emerging Technology Conference about this topic (I'll include it at the bottom of this post).
Coincidence aside, Clay is right on the money. He analyzes the RIAA attack model and outlines how file sharing systems are likely going to use social software models to evade the RIAA:
The disadvantage of social sharing is simple -- limited membership means fewer files. The advantage is equally simple -- a socially bounded system is more effective than nothing, and safer than Kazaa.
Limited search horizons are not necessarily all bad -- they do undercut the current model of searching for anything in this vast virtual library. But as it was pointed out in the Napster case, you have to know what you're looking for, otherwise Napster (and Kazaa to a degree) are useless. In light of Clay's previous essay "The Music Business and the Big Flip" the Internet community needs to take on the A&R functions and help one another with new music discovery. Having a limited search horizon bounded by the tastes of your friends will allow you to peruse the music in the social networks. This gives users the opportunity to discover new music on their own, rather than having it shoved down their throats by the music industry. Simply put, it is a different approach to finding music -- one that might be even more dangerous to the RIAA.
Furthermore, given the vast number of people in my Friendster network that I enjoy through a couple dozen friends, I would be inclined to think that the search horizon of a traditional P2P network isn't all that different from a search horizon offered by a social software P2P network. What difference are we talking about? A factor of 2? An order of magnitude??
Even if it is an order of magnitude, social software systems can be changed to expand the search horizon. What if the social software P2P systems used an approach closer to tribe.net where you can trade files with anyone in your tribe and with tribes that are closely allied with your tribe? My group of friends has several other sister social groups with whom I would feel comfortable trading music. The same rules that apply to forming relationships between two people can also be applied to forming a relationship between two groups. And this can drastically expand the search horizon.
In any case, I hope that O'Reilly accepts my proposal so I will have a chance to brainstorm about this cool topic a the next ETech conference.
Here is the abstract of my talk proposal:
Abstract:
Social software applications like Friendster, Tribe.net and LinkedIn have
proven that it is easy to build an online network of friends. P2P
applications like Gnutella, Kazaa and eDonkey have shown how effective
file trading applications can be.
However, in the times when the RIAA is suing the world for downloading
music off the net, it's time to combine a social software applications
like Tribe.net with a file sharing application like Kazaa. The
Tribe.net aspects of this application allow users to create networks of
their friends and then allow file sharing between members of their network.
The primary goal of the social software aspects is to keep the
RIAA and MPAA out of your social network in order to avoid detection.
Fringe benefits from this approach include music discovery features --
basic search facilities in file sharing applications assume that you
know what you are looking for, but social networks can allow users to
highlight their favorite music and build communities based on musical
tastes.
To avoid detection, the file sharing application should use common
off the shelf technology to avoid detection by the bad guys. All
connections should be tunneled over SSH and port numbers should change
frequently. Using these thechniques, the bad guys won't be able to
tell the difference between a legit VPN session and a Britney Spears
track being swapped.
This talk will introduce the concepts above and present social network
models, detection avoidance strategies, attack strategies and a real
world safety evaluation of such systems.
Posted by Mayhem at October 14, 2003 12:13 PM
the link seems to be to the mac article you like so much instead of what you wanted. I agree that social file sharing is going to limit your search horizon. However, some people in the group are going to be that 20% that provide 80% of the new music, either through purchasing or through risking a Kazaa-like service. Hopefully, they also have the bandwidth and storage to make this accessable to the remainder of the group. Or, the group has some capability to provide them with the bandwidth and storage elsewhere.
On a related note, what do you think about a tribe (or even cross-tribe) "wishlist"? Various members of the group could put out track titles they are looking for but can't find in the social network and someone could scratch an itch and cross it off for them. Dunno, the idea probably needs some work, but it might be interesting.
