Q&A (10 minutes)
Q&A about ethics, computing, and logistics of the course.
No question is too silly, naive, controversial, basic, or advanced.
Many definitions of fairness
- individual fairness: the goal of giving similar individuals similar decisions
- group fairness: the goal of giving similar decisions across demographic groups
Discrimination manifests itself in unequal treatment of groups.
From Friedler et al. (2021)
Classification Metrics
- Accuracy is the proportion of all classifications that were correct, whether positive or negative. \(( TP + TN) / all\)
- Recall, or true positive rate (TPR) is the proportion of all actual positives that were classified correctly as positives. \(TP / (TP + FN)\)
- False positive rate (FPR) is the proportion of all actual negatives that were classified incorrectly as positives (probability of false alarm). \(FP / (TN + FP)\)
- Precision is the proportion of all the positive classifications that are actually positive. \(TP / (TP + FP)\)
Google Developer’s ML Concepts
Practice example
Two groups of 100 people.
| Flagged high-risk |
50 |
20 |
| …Flagged that did reoffend (TP) |
40 |
16 |
| …Flagged did not reoffend (FP) |
10 |
4 |
| Not Flagged that did reoffend (FN) |
10 |
2 |
Calculate accuracy, precision, true positive rate, and false positive rate
Practice example
Two groups of 100 people.
| Accuracy (TP+TN / all) |
80/100 = 0.80 |
94/100 = 0.94 |
| Precision (TP / flagged) |
40/50 = 0.80 |
16/20 = 0.80 |
| True positive rate (TP / reoffend) |
40/50 = 0.80 |
16/18 = 0.88 |
| False positive rate (FP / did not reoffend) |
10/50 = 0.20 |
4/82 = 0.05 |
Same precision, similar true positive rate, but people in Group A who do not reoffend are 4x more likely to be wrongly flagged.
Group Discussion: Ethics frameworks
- Consequentialism: In the example, which error is worse: a false positive (wrongly flagged) or a false negative (missed)? Does your answer depend on the domain (bail, loans, medical care)?
- Deontology: Is it ever acceptable to use a protected attribute (e.g., race) to correct for bias?
- Contractualism: Which fairness definition could be justified to everyone affected, including the group with the higher base rate?
- Political economy: Who chooses the fairness metric? Who benefits from that choice?
The impossibility result
Kleinberg, Mullainathan & Raghavan (2016) and Chouldechova (2017) showed:
When base rates differ between groups, a (non-perfect) classifier cannot satisfy calibration/predictive parity and equal false positive and false negative rates at the same time.
- Fairness is not a bug to be fixed with better code
- Choosing a fairness definition is a value choice, not a technical one
- Someone must decide which errors matter most, and for whom
Fairness and the law
US anti-discrimination law (e.g., Title VII of the Civil Rights Act, Fair Housing Act, Equal Credit Opportunity Act) recognizes:
- Disparate treatment: intentionally treating people differently because of a protected characteristic
- Disparate impact: a neutral-looking practice that disproportionately harms a protected group (Griggs v. Duke Power Co., 1971)
- The EEOC’s four-fifths rule: a selection rate for one group below 80% of the rate for the highest group is evidence of adverse impact
Most algorithmic bias looks like disparate impact, not intent.
Fairness in generative AI
- Image generators asked for “a CEO” or “a doctor” produce mostly white men; “a fast-food worker” produces mostly people with darker skin (Bloomberg analysis of Stable Diffusion, 2023)
- Attempts to correct this can overcorrect: in 2024, Google paused Gemini’s image generation of people after it produced historically inaccurate images (e.g., racially diverse 1940s German soldiers)
- Language models can associate names, dialects (e.g., African American English), and identities with negative traits
What would “fair” output even look like for a generative model?
Group Discussion: Fairness
- Should algorithms be used for high-stakes decisions at all, if perfect fairness is mathematically impossible? Are human decision-makers any fairer?
- Who should get to decide which fairness definition a system uses: engineers, companies, regulators, or affected communities?
- Base rates often differ because of historical injustice. Should a model treat those base rates as “ground truth”?
References and Further Readings
- Barocas, S., Hardt, M., & Narayanan, A. (2023). Fairness and Machine Learning: Limitations and Opportunities. MIT Press.
- Chouldechova, A. (2017). Fair prediction with disparate impact. Big Data, 5(2), 153–163.
- Friedler, S. A., Scheidegger, C., & Venkatasubramanian, S. (2021). The (im) possibility of fairness: Different value systems require different mechanisms for fair decision making. Communications of the ACM, 64(4), 136-143.
- Kleinberg, J., Mullainathan, S., & Raghavan, M. (2016). Inherent trade-offs in the fair determination of risk scores.
Journal Writing
Reflect on what you liked and disliked about this lecture, what you learned, and what you wished you’d learned.
- What you learned from the lecture and discussion – this can be something that surprised you, something you had not considered before, a new way of thinking about something
- What you wish you’d learned from the discussion – this can be questions you still have and questions you would like to explore further
- Anything you liked or disliked