Inspired by Wilbert Baan’s Interactive Story Telling Experiment and a spare hour to code, I made another system using Flickr to generate image to a story. This time the system scrapes headlines from Google News, gets significant terms from Yahoo, and then queries those as tags in Flickr. This way Flickr provides random, though clarifying, pictures to the headlines - the photo editorial is generative but often illuminating (for example, getting this picture to the Rick Renzi story). I’ve made the page scroll down automatically and reload when all headlines have been flickrified so I can have my spare screen act as an always up-to-date, but augmented, rss-reader / issue-ticker. You can try it yourself at flickrNews US or flickrNews NL or flickrNews FR. The US version works best as the Yahoo Term Extraction Service is optimalized for English.

Just for aesthetic fun I made ‘4 interpretations of Alice in Wonderland by Flickr’. The method is simple: load the first chapter of Alice in Wonderland, for each new word get a picture from Flickr and display it, if there already was an image for that word we’ll use that one. The result is quite beautiful and reminds one of Burroughs’ cutup method: 1st interpretation, 2nd interpretation, 3th interpretation, 4th interpretation. Kinda feels like that ‘memory’ game. How quickly can you cognitively map the images to the words they belong to and read Alice without reading the words?