lesson_10_flashcards

Question 1

What is Responsible AI?

Answer 1

The practice of building AI systems that are ethical, safe, transparent, and fair, while ensuring compliance with societal and legal standards.

Question 2

Why is Responsible AI important?

Answer 2

It mitigates risks, ensures public trust, complies with regulations, and aligns with moral imperatives.

Question 3

What are the three pillars of Responsible AI?

Answer 3

Safety, fairness, and accountability.

Question 4

What is bias in AI systems?

Answer 4

Unwanted or unfair preferences learned by models due to skewed training data or systemic issues in data collection.

Question 5

What is fairness in AI?

Answer 5

Ensuring equitable treatment and outcomes for all individuals or groups, often evaluated on dimensions like gender, race, or age.

Question 6

What is equity in AI?

Answer 6

Going beyond fairness to address historical and societal inequalities to create just outcomes.

Question 7

What is calibration in fairness?

Answer 7

A measure of how well a classifier’s predicted probabilities align with actual outcomes across different groups.

Question 8

What is the Fairness Impossibility Theorem?

Answer 8

A concept stating it is mathematically impossible to simultaneously achieve multiple fairness metrics (e.g., equal calibration and equal error rates) if groups differ.

Question 9

What is interpretability in AI?

Answer 9

Understanding which features a model uses and how they influence predictions.

Question 10

What is explainability in AI?

Answer 10

The ability to provide human-understandable explanations for specific model predictions.

Question 11

What is transparency in AI?

Answer 11

Providing insight into a model’s data, design, training process, and decision-making pipeline.

Question 12

What is governance in Responsible AI?

Answer 12

A structured process to ensure accountability, compliance, and ethical decision-making in AI system development and deployment.

Question 13

What is the role of provenance in AI systems?

Answer 13

Tracking the origin and usage of data to ensure auditability and accountability.

Question 14

What is differential privacy?

Answer 14

A privacy-preserving technique that adds noise to query results to prevent re-identification of individuals in datasets.

Question 15

What is redress in AI systems?

Answer 15

Mechanisms that allow users to seek remedies for errors or harms caused by AI predictions.

Question 16

What is GDPR and how does it impact AI?

Answer 16

The General Data Protection Regulation is an EU law that governs data privacy and requires compliance for systems handling EU citizens’ personal data globally.

Question 17

What is CCPA?

Answer 17

The California Consumer Privacy Act, a broad online privacy law giving California residents control over their personal data, impacting companies nationwide.

Question 18

What is the role of trust in AI privacy policies?

Answer 18

Trust ensures users understand and feel confident in how their data is collected, processed, and used by AI systems.

Question 19

What is the importance of transparency in privacy compliance?

Answer 19

Transparency helps users understand AI’s decision-making and data handling, ensuring trust and adherence to privacy laws.

Question 20

What is an example of bias in word embeddings?

Answer 20

Gender bias, where analogies like ‘man is to surgeon as woman is to nurse’ reflect societal stereotypes present in training data.

Question 21

What is federated learning?

Answer 21

A machine learning approach that trains models across decentralized devices, preserving user privacy by keeping data local.

Question 22

How do academic incentives affect the study of bias in AI?

Answer 22

Publishing pressures may lead to findings that don’t reproduce, highlighting the need for rigorous evaluation of biases in AI models.