Appropriate AI by Design : Nurturing Sustainable, Trustworthy and Responsible AI Futures

AI is often promoted as a resource efficient technology which helps mitigate the unsustainable impacts of today’s ever-expanding data-driven ecosystem. However, AI systems are themselves extremely resource hungry, energy inefficient and create huge amounts of CO2 that contribute to digital technology’s growing global footprint. This implicit incongruity calls into question the sustainable trustworthiness of AI. Consequently, there is an urgent need to innovate more sustainable, trustworthy, and responsible futures for these technologies – particularly for energy actors keen to deliver governmental Net Zero promises. Using Schumacher's notion of Appropriate Technologies as a lens, this paper describes the development of a novel, design-led Participatory Futuring approach which helps researcher-practitioners to critically and creatively engage with stakeholders (e.g., technologists, policymakers, citizens) regards AI implementation. Building upon More-than-Human-Centred Design and Systems Thinking techniques, the paper demonstrates how this unique approach can nurture collaborative discursive spaces. Here, stakeholders can work together to co-create alternative, appropriate design pathways that better negotiate the ecological trade-offs and unintended consequences of adopting AI futures.