In all honesty, I proposed my upcoming class on AI and Digital Storytelling because of enrollment issues. Enrollment in the English department ... and our professional writing program ... is at an all-time low.

Most of this is a perception problem. Analyzing stories and writing for multiple audiences is fun and extremely useful in every single field. So I proposed an entry level English course with the special topic of AI & Digital Storytelling to pique interest and show students what it means to be a writer in the 21st century.

In my writing, I've primarily been using AI to develop new poems ... treating AI much like a Surrealist poetry machine. Surrealism was a 20th century art movement that sought to break out our usual ways of thinking. Our minds get stuck in the ruts of repeated thought patterns. This is especially true with poetry. Surrealists developed many techniques and technologies to break out of these usual ways of thinking ... that is the best way to treat GPT-3.

The obvious next step is to explore how GPT-3 tells stories. In their book, Story Machines, Mike Sharples and Rafael Péréz y Péréz (2022) point out that most of our learning as children happens via storytelling. Telling stories is a great way to grab people's attention, illustrate how concepts and events can be linked, and show how abstract topics affect our everyday lives (p. 17). 

Those disjointed stories you hear my twin toddlers telling with their Duplo toys ... that's them learning how to make connections and communicate their experiences. 

Notice that this learning occurs during play ... not in the classroom.

This is why I'm using digital storytelling this semester to teach students the right use of AI writing tools. Storytelling with AI gives students an opportunity to explore language play, critically analyze AI output, and understand the social nature of writing.

Though we should approach academic writing playfully, it is difficult. Stakes are high, and students don't always see the point. 

Storytelling brings playful writing back to the classroom, offering more opportunity for experimenting with language in ways that engage readers. The ability for students to generate stories that are creative, imaginative, and even persuasive is a great way to practice using AI tools for innovation.

The fun is in the revising and rewriting ... not in the generation of text. Academic essays too often start and end with the generation of text.

AI doesn't really know how to play on its own. Its job is to guess what comes next. Sometimes that's useful ... sometimes that's even entertaining, but it is never purposeful without humans in the game. As a result, AI-generated stories tend to be predictable, boring, or without meaning. This will be obvious to most students, because they engage with stories every day.

Students may not always know what a mediocre essay looks like ... but they will know a bad story when they see one. Their challenge will be to use that output to make something interesting.

Because AI scrapes it's data indiscriminately from the internet, all the structural biases (like gender and race) are embedded in most of the stories AI constructs on its own. The larger the dataset, the worse it gets. Tell AI to write a love story ... it is probably going to be hetero-normative. Ask AI to bring in a wise guru to your story ... it's probably going to be a man.

Focusing on stories helps students to critically analyze all the stories they encounter (or write) ... not just AI stories. Our culture and values are mostly formulated and passed on through stories, telling us how to behave in implicit ways.

Students (and teachers) are often obsessed with who owns what language. Real creativity requires multiple writers and multiple language sources. Even in academic writing, these ideas and sources are not always easily traceable.

My goal is to provide a safe space for students to share their stories and explore the potential of AI through digital storytelling. They don't have to worry about "getting caught" using AI, and we can freely discuss how we are using AI, because we are not writing non-fiction, academic essays, or argumentative papers.

George Lucas never had to cite all the movies he ripped off to make Star Wars.

Much of our discussions on AI revolve around plagiarism and content generation. With the introduction of ChatGPT3, we still see most people trying to get the AI to generate entire essays. While these are important conversations, we should spend more time learning how students can use these tools to fuel creativity ... rather than figuring out how they can be used for plagiarism.

As AI becomes increasingly prevalent in our daily lives, it is important for students to understand how it works and the implications of its use. Digital storytelling is an effective way to explore these topics in a fun and creative way by developing a healthy relationship with AI that enhances their writing ... instead of taking it over. Writing stories can be a lot of fun, often requires collaboration, and makes students care about the quality of content, whether it is AI or human-generated.

My class still isn't fully enrolled ... but it's close. I'm hoping that I can lay a foundation with these students for future classes that will draw them into the wonders of writing ... and AI-proof their work, not just for future classes, but for future professions.

---
Lance Cummings
Professor of English |  🔗 Linkedin Profile | 🔗 Medium Page | 🔗 Twitter