On the Importance of Explainability
summary
The talk discusses the problem of explainability in AI: modern systems (e.g., face recognition and image classification) can outperform humans but cannot explain their reasoning. It outlines the importance of interpretable models and explanations for high-stakes domains like medicine and law and notes that upcoming regulations will require AI systems to provide such explanations.
AI-written summary · about
chapters (11)
- 0:00Explainability in AI
- 3:02Black boxes in ML
- 6:30AI responsibility
- 9:50Context and intuition
- 13:34Interactive ML collaboration
- 16:48Trust in AI
- 20:19Explaining AI models
- 23:19Regulations and ethics
- 26:57Cross-domain AI
- 28:29Human vs AI limits
- 30:02Debating autonomous AI
related talks
45mAI-404: Explanation Not Found
28mThe Need for Explainable AI
24mWhy and How to Build Explainability into your ML Workflow
35mWhy Do You Need an Explainable Ai?
43mBring in the Lawyers: Explainable AI Driven Decision-making
15mIntroduction to Explainable AI
48mExplainable Artificial Intelligence (XAI): Why, When, and How?
45mAI-404: explanation not found