Group B, Poster #188, Community Capability Building (CCB)
Updating and Simplifying Great ShakeOut Regional Webpages with Artificial Intelligence-Assisted Processes
Poster Image:

Poster Presentation
2026 SCEC Annual Meeting, Poster #188, SCEC Contribution #15484 VIEW PDF
process: updating earthquake hazard descriptions for all U.S. regional ShakeOut pages (60 total, spanning all states and territories plus multi-state region overview pages), and employing AI models to develop a standardized, consolidated page format that facilitates efficient maintenance and simplifies migration to a more modern web server and content management system. Large Language Models, including Claude Opus 4.8 and GPT-5.5 Pro, were used to review regional pages, identify commonalities, create preliminary outlines and drafts, and synthesize a template for future pages. The resulting workflow combines LLM-assisted review, outlining, and drafting with manual verification of hazard information, statistics, references, and hyperlinks. This process reduces the time required to update regional earthquake information while promoting consistency, maintainability, and adaptability across ShakeOut.org. The resulting format scales the amount and organization of content to each region’s participation and information needs, including streamlined single-page formats for smaller regions. Overall, this project has demonstrated the potential of integrating artificial intelligence further into ShakeOut website modernization, and further refinement of prompts, input methods, and model parameters may improve processing speed and consistency.
SHOW MORE
SHOW MORE
