Ethical AI

Where do Ethics come from?

• Humans experience events that influence their ethics
• The influence is discounted by "munge distance"
• Influences decay over time
• Human ethics is the accumulation of decayed influences

Can AI Systems' Ethics use that Model?

• What events should be used?
• What is the discount for "munge distance"?
• What is the decay function?
• How can the influences be accumulated?
• How should this be represented?

AI Ethics in Action

• User's ethics?
• Society's ethics?
• AI system's ethics?