I first came to bots for one of my classes back in 2013, and in that time since, I’ve developed a small stable of botspawn that I’m relatively proud of. I’ve also developed a tool that helps people make bots in Google Spreadsheets without having to do write any code. Many people have used that spreadsheet tool with success, including my students. Next week (August 11, to be exact), I’ll be leading a short workshop in making bots at the 2016 Digital Pedagogy Lab Institute hosted here at UMW. It’ll be a quick, hands-on introduction to the world of bots and a demonstration of how to make them. Attendees need not have any prior coding or programming experience, but (if all goes well), they’ll leave 90 minutes later with a working Twitter bot of their own.

But why bots? Jesse suggested I title this workshop “Teaching with Bots”, so I’ve been thinking about some actual and some potential uses for bots in classrooms. Now, to be clear, I don’t mean bots that pretend to be human teaching assistants (someday I’ll get around to finishing my thinkpiece on why I think that’s a bad idea), but I am interested in bots as exploratory creative agents offering disruptions and interventions as paths toward understanding (or not.)

What, for instances, can we learn about Walt Whitman’s poetry from @WhitmanFML? This bot combines tweets hashtagged “#FML” with lines from Whitman to produce a chimeric, surprisingly coherent voice.

Or how do we understand authenticity and rhetoric when a parody bot accidentally produces an insight that, if not quite profound or true, could still be a conversation starter:

I’ve spoken before about how I think @ROM_TXT could be a tool for learning more about videogame textuality, though I think I still prefer it for its glitchy poetics:

These are just a few examples from existing bots. But let’s say you’re studying an author for a class, and you want to make a bot that mimics that author’s style? A Markov-chain algorithm would be a good place to start (there’s one built into my spreadsheet tool), but how do you decide which source text to feed it with? Are all of that author’s works representative, or are some more “typical” than others? What makes those other texts’ outliers, and can you automatically eliminate them in some way? Does the text need formatting or clean up before processing? And what if the output isn’t great? Should you try a different algorithm, or create a different approach entirely?

Ultimately, I believe that exploring these questions about an author’s work can lead students to ask and discover things about that work that they otherwise wouldn’t have, and that’s just one way that bots could be useful in teaching.

Interested? If you’ll be at DPL, come on out in person on Thursday. If not, I’ll prepare a handout of some sort to post online. Either way, how have you seen bots used in teaching contexts, or what could you imagine building?

[Robot image by Flickr user badjonni CC BY-SA 2.0]