EthicsSafetyAccountabilitycs.CY
Biospheric AI
Marcin Korecki
January 31, 2024
arXiv.org
1 citations
Key Takeaways
-
Current AI ethics frameworks are too narrow: The paper argues that existing AI ethics approaches focus exclusively on human values, which is insufficient for protecting the broader environment. Even recent attempts to include animal sentience don't adequately address the complexity of ecosystems and biospheric health.
-
A new "ecocentric" approach is needed: Rather than centering human or animal interests, the paper proposes "Biospheric AI" that prioritizes the health of entire ecosystems and the biosphere as a whole, recognizing that AI systems can have cascading environmental impacts beyond individual species.
-
AI design can be reoriented toward ecological goals: The paper suggests that modern AI models can be deliberately designed and applied in ways that consider and protect biospheric interests, though it acknowledges this requires new research directions and methodologies.
-
The stakes extend beyond traditional ethics concerns: By framing AI's relationship with the biosphere as a central concern, the paper highlights that irresponsible AI development could cause environmental damage at scale—making ecological considerations as important as human value alignment in responsible AI development.
Abstract
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.