Stockholm Colloquium in Philosophy: Ruth Chang (Oxford)

Artificial Intelligence and Value Alignment

Abstract:

What might philosophers contribute to the 'value alignment' problem in artificial intelligence?  This is the problem, roughly speaking, of aligning machine outputs with human values. In this talk, I outline a conceptual framework for making progress on this problem, diagnosing ways in which I believe technologists who confront the problem go wrong and offering a proposal for designing machine systems that I argue is a sine qua non for machine-human value alignment.

Last updated: 2025-10-08

Source: Department of Philosophy