Marc Tuters

When I first saw a project like this at the Whitney 2000 Bitsreams (using an internet image keyword search entered by gallery goers and projected around the walls), I appreciated the randomness of the results along with their accuracy mixed with noise, however, Google image search erased much of that novelty. With flickr, the quality of images inherently increases, and as it's gained size and steam, the tags have allowed the collective beast to really leverage it's knowledgebase. At this point I think it's time to use it to investigate more meaningful datamining and associations beyond the random. "Experimental Colr Pickr"

http://www.krazydad.com/colrpickr/index.php?group=squaredcircle

by Jim Bumgardner that uses the "Squared Circle" set is an example, but the content is a little dry. I think there's opportunity to do something much more socialy provocative and organized around locations, events, and what not.

I haven't focused as much as I'd like to on social software potentials/theory, but if anyone has some writings to refer to on these topics, please send me some links.