Responsible AI practicesFull notesSummaryRingkasanStoriesPracticePrinciples of Human-Centered Design for Explainable AI8 exam-style questions on this lesson.Question 1 of 8Emergency room doctors use an AI tool to make high-stakes decisions under time pressure. Which human-centered design principle is most relevant?ADesign for human and AI learningBDesign for unbiased decision-makingCDesign for amplified decision-makingDDesign for maximum automationCheck answerQuestion 2 of 8In amplified decision-making, what does reflexivity mean?AOutputs are mirrored on two screensBThe system reacts faster than the userCThe model retrains itself automaticallyDThe design prompts users to reflect on their decisionsCheck answerQuestion 3 of 8Which steps belong to designing for unbiased decision-making?AFind biases, design fair processes, and train decision-makersBHide the decision process so users can't game itCRemove all humans from the decision processDUse only the largest model available for decisionsCheck answerQuestion 4 of 8An AI system learns from expert radiologists, the way an apprentice learns from a mentor. Which concept is this?AReflexivityBCognitive apprenticeshipCPersonalizationDData augmentationCheck answerQuestion 5 of 8A learning platform is designed to be accessible to people with disabilities and language barriers. Which aspect of human and AI learning is this?AAmplified decision-makingBCognitive apprenticeshipCUser-centered designDControllabilityCheck answerQuestion 6 of 8What does reinforcement learning from human feedback (RLHF) do?AReduces the number of features in the training dataBEncrypts training data with human-managed keysCRemoves humans from the training loop entirelyDUses human feedback as the reward to align the modelCheck answerQuestion 7 of 8A company wants human annotators to rank model responses so the rankings can train a reward model for RLHF. Which AWS service fits?AAmazon SageMaker Ground TruthBSageMaker ClarifyCSageMaker Model MonitorDAmazon A2ICheck answerQuestion 8 of 8. Select two.Which TWO are key aspects of designing for unbiased decision-making? (Select TWO.)ATrainingBHidden processesCFairnessDFaster inferenceEMaximum autonomyCheck answer0 of 2 selectedModel Trade-Offs9 exam-style questions on this lesson.Machine Learning Development Lifecycle10 exam-style questions on this lesson.