Responsible AI practices
Introduction
How the course is laid out
The story: A driving course has three parts: first the rules of the road and why they exist, then how to drive safely with the car's own safety features, and finally how to signal clearly so other people on the road understand what you're doing. There's a short quiz after each part.
In AI/AWS terms: The course has three sections, each ending with a knowledge check:
- What responsible AI is: its definition, the challenges it addresses, and its core dimensions (the rules of the road).
- Building responsible AI systems: AWS services and tools, responsible model selection, and responsible dataset preparation (driving safely with the car's features).
- Transparent and explainable models: transparency and explainability, safety versus transparency trade-offs, and human-centered design for explainable AI (signaling clearly).
For the exam: Responsible AI covers the principles, the AWS tools that support them, and transparency and explainability.