AI Developers Bet on Different Moral Philosophies: Anthropic on Virtue Ethics, OpenAI on Deontology

geoffreyirving · x · 2026-08-23

Geoffrey Irving observes that different AI developers are implicitly betting on different systems of moral philosophy as their central approach to character training. Anthropic leans towards virtue ethics, while OpenAI tends towards deontology.

Related event: Geoffrey Irving on Character Training: Language Philosophy Challenges and Diverging Moral Philosophy Bets in AI Alignment(7 posts)→

Original post →

More from Companies & People

• Anthropic Staff Actually Take Vacation, OpenAI Does Not — JacquesThibs · 2026-08-23
• DOJ reportedly probing a16z over board seats at competing startups — HaktanSuren · 2026-08-23
• Anthropic CTO holding an MBA violates expectations — saranormous · 2026-08-23
• 2012: buy for users, 2021: for engineers, 2026: for training data — andrewchen · 2026-08-23
• Miles Brundage: Anthropic employees actually take vacation unlike OpenAI — Miles_Brundage · 2026-08-23
• Ox Alpha Revealed to Use Unreleased Zhipu GLM Model — py_blu · 2026-08-23

Companies & People channel →