ETech Bionic Systems: We can make participants faster...stronger...better March 8, 2006 You Mon Tsang, CEO Dan Gartung, CTO YM is a lurker not submissions to wikipedia. no editing. dugg 6 stories 2 del. bookmarks 5 flickr photos has rated 500 movies rated 100% ebay transactions only 7% users read a blog 4% create content most users dont' participate participation lacks short term payoff, and that is why it suffers. 5% of the populace can create content. bt 5% can change the world. problem #2: those who create are shown a very fractured world to operate in. 50 photo sites, 100 blog aggregators and tagging sites, etc. kindness utillity incentive harvest wikipedia flickr epinions, pagerank squidoo technorati's authority memo newsvine? give them somehting back boxxet: planned for 3% participation add bionics to make it worth more bionics: amplify. participants are shy, nuanced, perceptive, passionate, undependanble, distracted. automation: tireless, obedient, thick skinned, but clueless combine the best of both. good practices of bionic systems: leveragre the participation that humans are uniquely qualified to handle utilize machine learining rather than programming give back to the user more than the system receives yet another "know it when i see it" reference. perferences hot or not digg biz360 filtering and categorization delicious audio visiual recognition a9 selecting photos riya (++) facial recognition actually learns what people look like emotive alert tell mood of the voicemail machine learning by example photo descirptions tone of an article (sarcasm, happy, sad, etc.) return > gave you rate 30 movies, get 200 reco's can find posative and negative journalists on a topic. suport vector machine. sentiment filtering prefernce demographics boxxet brings together best of all into one feed.