Moral Tractability

Moral tractability explores the computational feasibility of solving ethical problems, aiming to understand the limits of both human and artificial moral reasoning. Current research focuses on identifying tractable subclasses of problems within various ethical frameworks (consequentialism, deontology, virtue ethics) and developing efficient algorithms for specific model types, such as probabilistic circuits and certain classes of neural networks. This research is significant for advancing both theoretical understanding of ethical decision-making and the development of reliable and explainable AI systems capable of navigating complex moral dilemmas.