Writing Environmental Nonfiction with AI
Alan MacEachern and William J. Turkel are organizing two events this fall on the intersections of Artificial Intelligence and Environmental History. They have given them the title Artificial|Natural. The first is a series of talks and a roundtable discussion on October 7 with Dagomar Degroot, Jo Guldi, and Marnie Hughes-Warrington. The second is a hands-on workshop on "Writing Environmental Nonfiction with AI". It will be held on the morning of October 8 and conclude by noon.
Workshop participants will be coming from history, journalism, STS, environmental studies, and other fields. Working in pairs or small teams, participants will take on a project of their choice. We won't assume any particular level of expertise. For people who are just getting started, we can provide the scaffolding to get up and running by using one of the large language models through the browser. For participants with some coding experience, more ambitious projects might make use of techniques like RAG (retrieval augmented generation), automating linked open data queries through a SPARQL endpoint, or working with multimodal systems. The workflows are firmly grounded in the tools and techniques of the digital humanities: text encoding, the semantic web, linked open data, bibliography, databases, web APIs, text analysis, and text mining. There are two key differences, however. Rather than working with off-the-shelf tools or building them from the ground up, generative AI allows us to work from the top down. We can also make use of the models' ability to co-author code as well as prose.
Please email william.j.turkel@gmail.com to register.
