FAQ
Frequently Asked Questions
Quick answers about Katherine Elkins’s research, projects, methods, programs, and where to find more detailed information.
Definitions
Research & Methods
What is human-centered AI?
Human-centered AI treats language, culture, judgment, institutions, and human values as central to the design, evaluation, and governance of AI systems. In Elkins’s research and teaching, this includes model evaluation, narrative and language analysis, ethical reasoning, creativity, cultural preservation, and institutional governance.
What is middle reading?
Middle reading works at more than one analytical distance: computational trajectories locate structure across texts, while close reading of the relevant passage determines what the pattern means. See Narrative, Translation & Transmission.
What is the confidence-scoring method?
Introduced in Informed AI Regulation, the confidence score measures how firmly a language model commits to a moral judgment rather than hesitating. It helps compare normative certainty across outputs and supports audits of framing fragility, negation sensitivity, and evaluation design. See AI safety research.
What is SentimentArcs?
SentimentArcs is an ensemble method developed by Jon Chun for modeling emotional trajectories across narratives using multiple sentiment-analysis approaches. It uses dynamic time warping to compare arcs of unequal length. See The Shapes of Stories.
Names and relationships
Projects & Programs
What is Archival Intelligence?
Archival Intelligence is a Schmidt Sciences-supported research project exploring how AI can help preserve, organize, and interpret endangered cultural archives while retaining the local knowledge and context that make those records meaningful. Visit the project site.
What is the difference between the Human-Centered AI Lab, Kenyon AI CoLab, the human-centered AI curriculum, and IPHS?
- Human-Centered AI Lab
- Katherine Elkins’s nonprofit organization, supporting broader human-centered AI research, collaborations, and public-facing work.
- Kenyon AI CoLab
- The student research lab at Kenyon, supporting experimental projects and mentored work in AI.
- Human-centered AI curriculum
- The interdisciplinary curricular program developed beginning in 2016: courses and curricular structure, not an organization.
- IPHS
- Kenyon’s Integrated Program in Humane Studies, an academic program and context for some of this interdisciplinary work.
Scholarly reception
Citations and Reach
In what fields is Katherine Elkins's work cited?
Across AI safety and model evaluation, AI governance and law, technical AI and NLP, computational literary study, education, translation studies, philosophy, medicine, and literary and cultural scholarship. Computer scientists, legal scholars, social scientists, and humanists all cite it.
What is Katherine Elkins's most-cited work?
"Can GPT-3 Pass a Writer's Turing Test?" (Elkins and Chun, 2020) is the most-cited article in the history of the Journal of Cultural Analytics. "The Shapes of Stories" is the most-cited Element in Cambridge's Digital Literary Studies series.
How is her AI-regulation work used?
Her three-regime map of EU, Chinese, and US AI regulation serves as shared shorthand in comparative law, international relations, and political theory. It has been carried into energy engineering, spiritual care, and journalism studies, and into industry AI-safety research from a Microsoft AI team.
Which of her methods do other researchers build on?
Her sentiment-arc method for measuring narrative structure is used in translation studies, corpus linguistics, disinformation analysis, computational literary history, and affective-computing research. Her ethics-based audit anchors work on evaluating moral reasoning in language models and has been taken up in business ethics.
What has she written on narrative and AI?
"What the Rise of AI Means for Narrative Studies" responds to the claim that computers cannot read or write literature. Narrative Inquiry, the field's flagship journal, has published two 2026 articles that take up the argument.
Who founded human-centered AI at Kenyon?
Katherine Elkins and Jon Chun founded Kenyon's human-centered AI curriculum and AI Lab in 2016.
Site guide
Finding the Right Page
Where can I find citations, reviews, and scholarly uptake?
Reception documents scholarly citation, adoption, debate, reviews, and later uses of Katherine Elkins’s work.
Where can I find talks and keynotes?
Speaking lists major topic areas and selected recent and upcoming engagements.
Where can I find interviews and press coverage?
Media collects recent coverage, interviews, podcasts, and press resources.
Where can I find the complete publication record?
The CV provides the most complete formal record. Books highlights the major volumes; Google Scholar and ORCID provide external publication profiles.
Practical routes
For Journalists, Organizers & Collaborators
I’m a journalist. Where should I start?
Start with Media for coverage and press resources, Speaking for public topics, and Research for the underlying work. Use Contact for interview or expert-comment requests.
I’m organizing an event. Where should I start?
Speaking outlines major topics and selected engagements. Use Contact to discuss an invitation, audience, or format.
I’m interested in research collaboration. Where should I start?
Research provides the current project map, and About gives the broader intellectual context. Use Contact for a specific collaboration inquiry.
I’m reviewing Katherine for a grant, search, fellowship, or institutional opportunity. Where should I start?
About establishes her research identity and research leadership; Research covers current work; Reception documents scholarly uptake; Recognition records major roles and external support; and the CV provides the complete formal record.