Ulrich Engelke, Dirk Slawinski and Tisham Dhar: Designing AI for Earth Science
Artificial intelligence (AI) systems have seen an incredible uptake in recent years across a wide range of applications, including scientific use cases where the AI system is often considered to be a co-scientist working alongside the human scientist. By combining human reasoning and decision-making strengths with the power of contemporary AI systems, such as scalable analysis and global data access, we can expect to achieve scientific outcomes that go well beyond what each entity can achieve individually. We are already seeing strong evidence of the success of such collaboration. One may, however, wonder whether the benefit of scientist-AI collaboration is expected to be of a similar degree across scientific disciplines and what system design parameters and contextual factors influence the success of scientific discovery. Effective user interfaces, for instance, carefully designed through user experience (UX) principles are an integral component for users to deeply engage with AI and for them to remain in control of the system. In this presentation, we discuss the scientist-AI collaboration from a human centric viewpoint with a specific focus on earth science. We will provide a brief overview of current research at the intersection between AI, UX and earth sciences, and we will pose some open challenges to provide food for discussion at the workshop and avenues for further research for the ACCESS community.