Editors' Choice

Editors’ Choice: Transformers from Scratch

Editor’s Summary: This post provides a detailed explanation of how transformers work. Transformers in this context refers to a tool for sequence transduction (converting one sequence of symbols to another) an essential tool for natural language processing. The author provides a step by step discussion of how transformers work in terms of language, including many […]

Editors' Choice

Editors’ Choice: Vibing Digital History

Editors’ Summary: This post considers how generative AI has broken down the barrier to entry for doing digital history. The author argues that while AI is bad at doing history, it can enable historians to do good digital history. He acknowledges that in some instances code generated by AI immediately becomes technical debt, especially for […]

Editors' Choice

Editors’ Choice: Hidden Constellations

Editors’ Summary: This project combines oral history interviews, digital storytelling, and GIS to map Sapphic areas of New York City. The author begins by visualizing all of the bars in NYC, and identifying the four bars that are explicitly lesbian bars. The author encourages the reader to move beyond the idea of bars as the […]

Editors' Choice

Editors’ Choice: Defactoring Pace of Change

Editors’ Summary: This article argues that bespoke code used in digital humanities research should be treated as a core part of scholarly output rather than treated as invisible technical labor. It introduces defactoring, a method of close reading and restructuring, to reveal a project’s underlying computational narrative. The authors demonstrate this approach by unpacking the […]

Editors' Choice

Editors’ Choice: Can AI Replace Social Science Researchers?

Editor’s Summary: This post responds to another viral essay “Academics Need to Wake Up on AI” that argued that academia is in denial about how good Claude Code is at producing the 6,000–8,000 word journal article. The author argues that the peer-reviewed journal article as the primary unit of academic production is likely dead, and […]

Editors' Choice

Editors’ Choice: No More Tools

Editors’ Summary: In this post, the author shows how the rise of AI has made the critical thinking component of using code in DH even more essential. He details his initial explorations using Claude Code to build DH web apps for use in the classroom. This post argues that the old tools of Digital Humanities […]

Editors' Choice

Editors’ Choice: Do all politicians sound the same? Comparing model explanations to human responses

Editor’s Summary: This article considers the bold claim that politicians from different parties really “all sound the same”. It trains an AI model on 20+ years of Finnish parliamentary speeches and compares its guesses about affiliation with that of 438 human readers. It turns out that the system is ‘better’ at telling parties apart. Humans […]

Editors' Choice

Editors’ Choice: More Strategies for Avoiding AI

Editors’ Summary: This post shares some practical ways college instructors can design courses so students are not completely dependent on AI. It highlights grading systems that reward process over product, moving from writing to problem solving to finally reaching a solution. It empowers the students into realising the value of thinking for themselves instead of […]

Editors' Choice

Editors’ Choice: Stories from Black Physicists in Our Collections

Editor’s Summary: Modern sciences in the United States has been a predominately white-dominated workplace. This project at the American Institute of Physics collects oral history interviews from Black and African American physicists. This project addresses the lack of narratives in the history of science, especially physics, from minority communities. The need to record the perspectives […]