Friday, August 8, 2008

Exploiting socio-cognitive network effects will involve a new science

A variety of authors like Yochai Benkler have argued that the rise of decentralized networks of knowledge creation, information processing, knowledge sharing, etc. are just an early indicator of a deep radical change in the economic and social realms.

Decentralized network-based systems challenge centralized hierarchical systems. Seti@home distributed computation challenges supercomputer makers like IBM and NEC; Web-as-a-platform systems and open source code development challenge Microsoft; p2p distribution challenges media giants like the recording industry; Wikipedia challenges Britannica; the Iowa Electronic Markets challenge expert political forecasters; and the list goes on.

I recently gave a short talk about the idea that the businesses that will succeed this disruptive wave will be the ones that figure out how to harness more people, more efficiently, in more sophisticated and creative ways, and this will require some new forms of science. A link to a video of the narrated slideshow can be found in this blogpost.


Jed Harris said...

The audio seems to stop just after the slide with the red and blue blog clusters. The video continues.

Ed H. Chi said...

We were just made aware of this problem and are attempting to fix the audio problem.

Peter Pirolli said...

The audio problem is something we just discovered in Keynote that is apparently something the iWork team thought had been fixed. They're working on it.

Anonymous said...

It's an interesting topic to ponder. I believe that most of Benkler's examples can be divided up according to three basic incentives. Handing over excess CPU cycles for something like drug discovery addresses a need for people to act charitably. I see open source as being driven primarily by career concerns and professional reputation. Note that more than 99% of open source contributors are male. This matches gender participation in professional software development pretty well. Wikipedia seems to share a lot in common with the current trend of political dirt digging on blogs. This can be seen as driven by persuasion and group membership. Some of Wikipedia's authorship can also be driven by career concerns - someone may want to learn about an esoteric topic by researching it and writing an article.

What's important, I believe, in all of this is that some level of inter-party trust has to develop for socio-cognitive network effects to emerge. People don't like to contribute to a cause where they feel that freeloading is taking place. A lot of technological development has a winner take all structure. In group cognition this could result in a small number of key participants being overwhelmed with communication requests from a large number of attention starved individuals. That's a pretty good model for how a lot of things currently work, so maybe hierarchies have an advantage. How can this be altered? I'd love to know.