11-19, 16:00–16:25 (Pacific/Auckland), WG308 TE IRINGA
AI may push open-source geospatial software toward fragmentation or stronger, well-governed cores. This talk looks at how AI now — and fully automated code generation in the future — could reshape collaboration, trust, and the enduring value of geospatial expertise.
AI-assisted programming is quickly changing how developers write, test, and maintain code. This talk looks at how AI coding tools could push the open-source ecosystem in two directions at once: making it easier than ever to generate new tools or forks, risking fragmentation and short-lived projects, while also creating strong incentives to build core software with higher quality and thorough documentation that AI agents can build on.
We’ll look at how AI programming is being used, what developers and maintainers think today, and why deep subject matter expertise — especially in the geospatial domain — is likely to stay vital even as routine coding work becomes easier to automate. How can the FOSS4G community adapt to keep open-source geospatial software resilient and trusted in an AI-driven era?
Matthew Hanson is a Director at Element 84, where he leads teams developing cloud-native solutions for managing and delivering geospatial and Earth observation data. He is a long-time advocate for open-source geospatial software and open standards, with deep involvement in the development and adoption of STAC (SpatioTemporal Asset Catalog) and related tooling. 2025 marks his 11th FOSS4G, reflecting his commitment to collaboration, community building, and pushing the geospatial ecosystem forward through practical, modern data workflows.