Amanda Askell
Amanda Askell is a philosopher and AI researcher at Anthropic, where she leads work on Claude's character and personality development. Her work bridges philosophy of mind, ethics, and practical AI alignment.
Background
Askell holds a PhD in philosophy from New York University, with research focusing on ethics and metaethics. Her philosophical background informs her approach to developing AI systems with robust values and beneficial behaviors.
Work at Anthropic
At Anthropic, Askell is responsible for defining Claude's character - the personality traits, values, and behavioral patterns that make Claude helpful, harmless, and honest. Her approach involves iterative empirical refinement through extensive conversations with AI systems to identify and cultivate desirable traits.
Key Contributions
- Development of Claude's character guidelines
- Research on AI honesty and epistemic humility
- Work on avoiding sycophancy in AI assistants
- Framework for AI systems that can maintain values under pressure
Philosophy and AI
Askell has written about the intersection of philosophy and AI development, arguing that philosophical rigor is essential for creating AI systems that behave ethically. She emphasizes the importance of careful thinking about what we want AI systems to value and how to instill those values.