Explainability
Domain: Safety & Ethics
🎯 Motivation
The ability to explain why an AI made a specific decision. Critical for trust, especially in high-stakes domains.
📋 Concrete Example
AI rejects a loan. Explainability: ‘The rejection was based on: low credit score (60% weight), high debt ratio (30% weight), short employment history (10% weight).’ Reasons, not just a yes/no.
🔗 Analogy
A teacher explaining a grade - ‘You got a B because: homework was excellent (40%) but the exam was average (60%).’ Explainability gives reasons for AI decisions.