Nicolai Traasdahl Tarp, Gustavo Ostos Rios, and Tim van de Vathorst
Rivers and humans are deeply connected. Rivers sustain life on a planetary scale and underpin the economies and lifeways of human populations around the world (Anderson et al., 2019). However, despite this relationality, humans and rivers are often treated as a dichotomy, where human-centred approaches take precedence when designing new and sustainable ways of managing water (Allan et al., 2024).
We believe that there is a need to reconsider how rivers are represented and included when co-creating the future of river management.
Building on emerging literature exploring more-than-human representation and participation in decision-making processes, we explore the use of artificial intelligence (AI) as a mediator of more-than-human perspectives in systemic design processes.
Using novel Large Language Model (LLM) technology, existing real-time datasets and historical information; we have designed an AI persona that embodies the River Thames in the UK. It gives the river a voice and represents its perspectives, challenges, and needs to create an empathic connection with human co-creators. By turning available river catchment data about geology, hydrology, water quality, aquatic biota and human-river relations into stories and conversations shared by the river itself, AI offers a novel way to include the river as an active partner in decision-making and systemic design processes.
We end our presentation by reflecting on the usefulness and pitfalls of using AI for more-than-human representation and call for more work exploring new and interactive approaches of increasing more-than-human participation in systemic design processes.
Keywords: more-than-human, river health, water ecosystems, artificial intelligence, human-river relations