Automated text generation is interesting and provocative for folks interested in writing; I am thinking of provocations like the material created for NaNoGenMo; but more often, this sort of text generation appears in shorter forms, often tweets, from “bots.” Mark Sample has written about the power and political potential of such bots.
Whether folks are interested, and how they’re interested, would determine the shape of this session: we could talk about, or compare, people’s favorite existing “bots” (or note some of the challenges and problems with them; my favorite is pentametron), or we could work collectively to develop a bot. I’ll try to have the infrastructure ready so that we could get a basic Twitter bot off the ground and tweeting in the course of a session, if that is something people would be interested in; there’s a lot of unfun fiddly-business when setting up a “twitter” bot (or the mechanics of how to get to a script to run routinely or whatever), which I am happy to talk with folks about, but is way less interesting, than trying to imagine what such a project/device can do.
(For folks with previous bot writing experience, I have never used a Markov chain… but if anyone had such experience, and wanted to share, I would love to learn!)
