EconTalk Can Artificial Intelligence Be Moral? (with Paul Bloom)
79 snips
Dec 25, 2023 
Join psychologist Paul Bloom from the University of Toronto as he challenges the idea of moral artificial intelligence. Bloom argues that pursuing moral AI may not only be futile but detrimental. He dives into the complexities of aligning AI with human values, the ethical dilemmas involved, and how these affect personal freedom. The conversation touches on parenting parallels, moral progress, and generational attitudes towards technology, providing thought-provoking insights on what it means to be moral in an AI-driven world.
AI Snips
Chapters
Books
Transcript
Episode notes
Alignment Paradox
- AI alignment with human values is complex because people want different things.
- There's no single morality to align with.
Intelligence vs. Morality
- Intelligence doesn't equate to morality.
- Russ Roberts questions whether smarter people make better moral decisions.
Origins of Morality
- Paul Bloom believes morality is partly innate and partly shaped by culture and experience.
- He notes that even horrific acts are often motivated by a twisted sense of morality.
