Envisioning Generative Artificial Intelligence in Cartography and Mapmaking

TL;DR

This paper explores how generative AI can revolutionize cartography by enhancing map design, interpretation, and evaluation, while also addressing ethical concerns and limitations of current AI applications.

Contribution

It provides a comprehensive overview of GenAI's potential in cartography, including case studies and a roadmap for future research in this interdisciplinary field.

Findings

GenAI benefits various cartographic design decisions.

Case studies on symbolization, map evaluation, and reading.

Identification of limitations and ethical considerations in GenAI applications.

Abstract

Generative artificial intelligence (GenAI), including large language models, diffusion-based image generation models, and GenAI agents, has provided new opportunities for advancements in mapping and cartography. Due to their characteristics including world knowledge and generalizability, artistic style and creativity, and multimodal integration, we envision that GenAI may benefit a variety of cartographic design decisions, from mapmaking (e.g., conceptualization, data preparation, map design, and map evaluation) to map use (such as map reading, interpretation, and analysis). This paper discusses several important topics regarding why and how GenAI benefits cartography with case studies including symbolization, map evaluation, and map reading. Despite its unprecedented potential, we identify key scenarios where GenAI may not be suitable, such as tasks that require a deep understanding of cartographic knowledge or prioritize precision and reliability. We also emphasize the need to consider ethical and social implications, such as concerns related to hallucination, reproducibility, bias, copyright, and explainability. This work lays the foundation for further exploration and provides a roadmap for future research at the intersection of GenAI and cartography.