

Walter Sinnott Armstrong on AI and Morality
Jun 14, 2024
Philosopher Walter Sinnott Armstrong and David Edmonds discuss AI's role in ethics, focusing on embedding human values, resolving ethical dilemmas in healthcare, and AI's potential in decision-making processes like kidney transplants and medical diagnoses.
AI Snips
Chapters
Transcript
Episode notes
Challenges of Programming Morality
- Programming specific moral systems into AI, like utilitarianism or Kantianism, presents a challenge.
- Choosing one system over another raises the question of authority and potential bias.
Limitations of Bottom-Up Approach
- Building AI morality solely from internet data reflects human biases.
- This 'bottom-up' approach lacks transparency in its reasoning.
Hybrid Approach to AI Morality
- A hybrid approach combines principles with crowdsourced values.
- This approach identifies key features in moral dilemmas, like kidney allocation, to predict human judgment.