Biospheric AI
This work addresses the need for broader ethical frameworks in AI to prevent harm to the biosphere, representing a foundational shift rather than an incremental improvement.
The paper tackles the problem of anthropocentric bias in AI ethics and value alignment by proposing a new ecocentric paradigm called Biospheric AI, which aims to ensure AI does not damage the biosphere and outlines initial research directions for its design and application.
The dominant paradigm in AI ethics and value alignment is highly anthropocentric. The focus of these disciplines is strictly on human values which limits the depth and breadth of their insights. Recently, attempts to expand to a sentientist perspective have been initiated. We argue that neither of these outlooks is sufficient to capture the actual complexity of the biosphere and ensure that AI does not damage it. Thus, we propose a new paradigm -- Biospheric AI that assumes an ecocentric perspective. We discuss hypothetical ways in which such an AI might be designed. Moreover, we give directions for research and application of the modern AI models that would be consistent with the biospheric interests. All in all, this work attempts to take first steps towards a comprehensive program of research that focuses on the interactions between AI and the biosphere.