WLIA 2025 Annual Conference Appleton

Generative AI is becoming an invaluable tool for GIS professionals, streamlining everyday tasks such as project planning, workflow automation, and documentation. This presentation explores how AI can assist with automating workflows, drafting and reviewing scripts, answering geoprocessing questions, and refining written content. We will examine the strengths and limitations of AI models like ChatGPT, Gemini, and GovAI, while also sharing best practices for crafting effective prompts. Additionally, we will address key concerns such as data privacy and job security. By focusing on practical applications, this session will demonstrate how GIS professionals can leverage AI to boost productivity and efficiency in their daily work.

On January 12, 2024 seven proposals for Wisconsin’s legislative districts were submitted to the Wisconsin Supreme Court in the case Clarke v. Wisconsin Elections Commission. One was generated by a computer algorithm, and six were created by expert human mapmakers. In this presentation, we examine a major development from the case: the map proposal generated by the FastMap algorithm significantly outperformed the other proposals. Relevant background information, an algorithm summary, and map visualizations/comparisons are provided. The case likely marks the moment when computer algorithms surpassed humans in overall mapmaking ability, but it was not the first time that FastMap outperformed a group of expert human mapmakers. A previous version of the algorithm did the same in December 2021 when it created a map proposal for Wisconsin’s legislative districts that better adhered to “least change” principles than any of the six proposals that were submitted to the Wisconsin Supreme Court in the case Johnson v. Wisconsin Elections Commission.

Going forward, algorithmic mapmaking promises to enhance American democracy by ensuring that congressional, state legislative, and local election districts better reflect constitutional requirements, the will of the voters, and other criteria specified by courts and legal teams. However, better maps will become the norm only if high-quality redistricting data is readily available and if redistricting commissions, nonprofit organizations, courts, lawmakers, and legal teams are interested in fairness and are open to learning from experts in mathematical optimization and algorithm design.