GeoAI Agency Primitives

TL;DR

This paper proposes a set of nine agency primitives and a benchmark to enhance human-GIS collaboration with foundation models, aiming to improve productivity in geospatial workflows.

Contribution

It introduces a novel vocabulary of agency primitives and a benchmark to facilitate implementable and testable agentic assistance in GIS workflows.

Findings

Proposed a vocabulary of 9 agency primitives for GeoAI assistants.

Developed a benchmark to measure human productivity in GIS tasks.

Aims to enable comparable and testable agentic assistance in geospatial workflows.

Abstract

We present ongoing research on agency primitives for GeoAI assistants -- core capabilities that connect Foundation models to the artifact-centric, human-in-the-loop workflows where GIS practitioners actually work. Despite advances in satellite image captioning, visual question answering, and promptable segmentation, these capabilities have not translated into productivity gains for practitioners who spend most of their time producing vector layers, raster maps, and cartographic products. The gap is not model capability alone but the absence of an agency layer that supports iterative collaboration. We propose a vocabulary of $9$ primitives for such a layer -- including navigation, perception, geo-referenced memory, and dual modeling -- along with a benchmark that measures human productivity. Our goal is a vocabulary that makes agentic assistance in GIS implementable, testable, and comparable.