Week 1: AI & Inclusion

Time: Monday, July 6 at 11am & 7pm (Zoom link will be shared on Slack at in the Announcements channel on the day of the session)
Session Leads: Lis Sylvan & Lance Eaton

This session is a kick-off session where we'll do some introductions, some overview, some administrative details, and answer questions about the program. Be sure to attend one of these 2 sessions.

Time: Monday, July 6 at 3pm (Zoom link will be shared on Slack at in the Announcements channel on the day of the session)
Session Leads: Jessica Fjeld

Guiding Questions

1. • 1. The Executive Summary of the Principled AI doc (listed below) says that principles are only part of a larger governance regime. What's their role? As you learn about these documents, do you think that they are likely to have an impact on the technology's development and deployment? If so, how?

     2. At least one principle from the Fairness and Non-discrimination theme was referenced by every document in the study, but the documents recommended very different actions to avoid or remedy potentially discriminatory effects of AI. What approaches do you find most promising, and why?

     3. What information from the white paper is included in the visualization (p.8-9) and what is left out? How do we create visualizations -- or written summaries -- that balance robust substance with clarity and accessibility?

     4. How did the study methodology impact the findings? If you had designed this research, would you have made different choices?

Required readings

• Principled Artificial Intelligence: Mapping Consensus in Ethical and Rights-Based Approaches to Principles for AI

• P. 9-16 AI and Human Development (IDRC)

Additional readings

1. • Review the Principled AI report’s visualization and timeline and this Q & A with Jess Fjeld

   • Review Symposium reading list (AI topics)

Time: Tuesday, July 7 at 11am (Zoom link will be shared on Slack at in the Announcements channel on the day of the session)
Session Leads: Momin Malik

Guiding Questions

• • 1. On a technical level, what actually is AI, and related terms? (Machine learning, data mining, data science, pattern recognition, statistics, statistical learning, computational learning).

    2. What is AI imagined or thought to be in the general public?

    3. What is important about the discrepancy?

    4. What are correlations, and what can/can't they be used to accomplish?

Required readings

• • Vladan Joler and Matteo Pasquinelli. 2020. "The Nooscope Manifested: AI as Instrument of Knowledge Extractivism."
    AND/OR Yarden Katz. 2017. "Manufacturing an Artificial Intelligence Revolution."

  • Stephanie Yee and Tony Chu. "A Visual Introduction to Machine Learning, Part 1."

Additional readings

• • Stephanie Yee and Tony Chu. "A Visual Introduction to Machine Learning, Part 2."

  • [Hidden history of ML] Matthew L. Jones. 2018. "How We Became Instrumentalists (Again): Data Positivism since World War II." Historical Studies in the Natural Sciences 48 (5): 673–684.

  • [Showing the unrecognized connection between risk/credit industries and ML] Rodrigo Ochigame, Chelsea Barabas, Karthik Dinakar, Madars Virza, and Joichi Ito. 2018. "Beyond Legitimation: Rethinking Fairness, Interpretability, and Accuracy in Machine Learning." In ICML 2018: The Debates, at the 35th International Conference on Machine Learning.

  • [Activist letter, with great summary of underlying issues] Chelsea Barabas, Audrey Beard, Theodora Dryer, Beth Semel, and Sonja Solomun. 2020. "Abolish the #TechToPrisonPipeline: Crime prediction technology reproduces injustices and causes real harm."

  • [Long but comprehensive review] Momin M. Malik. 2020. "A Hierarchy of Limitations in Machine Learning."

  • [Connection to quantitative social science] Sendhil Mullainathan and Jann Spiess. 2017. "Machine Learning: An Applied Econometric Approach." Journal of Economic Perspectives 31 (2): 87–106.

  • [Review of issues in data] Alexandra Olteanu, Carlos Castillo, Fernando Diaz, and Emre Kıcıman. 2019. "Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries." Frontiers in Big Data 2.

  • [Philosophical angle] David Watson. 2019. “The Rhetoric and Reality of Anthropomorphism in Artificial Intelligence.” Minds and Machines 29 (3): 417–440.

Time: Wednesday, July 8 at 11am (Zoom link will be shared on Slack at in the Announcements channel on the day of the session)
Session Leads: Padmashree Gehl Sampath

Description: This lecture will look at how current technological trends, especially big data and AI related, reinforce structural inequality in societies. It will highlight how many of these changes are in a persistent state of conflict with the human rights guarantees contained in the UN human rights Charter. The seminar will end by providing students a discussion of alternate frameworks in which to locate and address the role of AI in structural inequality.

Guiding Questions: For these readings, keep an open mind about what draws your attention and interest; this will be important for the discussion in this session.

Required readings

• Unboxing Artificial Intelligence: 10 steps to protect Human Rights

• Artificial Intelligence Needs Human Rights

• Sampath - Global Justice in the 2020s

Time: Thursday, July 9 at 11am (Zoom link will be shared on Slack at in the Announcements channel on the day of the session)
Session Leads: Amar Ashar & Nagla Rizk

Guiding Questions

1. How do different communities define “inclusion” and “Artificial Intelligence” and how does it influence perceptions about the future of AI? How should we define ‘digital inclusion’ broadly vs. inclusion in the context of AI?

2. How can we design more inclusive AI systems?

3. How can data capture cultural norms in such a way that enhances inclusiveness?

4. Should we include those in AI-based systems who may not want to be included?

5. What are examples of how AI-based systems may exacerbate inequality in your community?

Required Readings

• Nagla’s Oxford Ethics Handbook piece (attached, read p. 635-644 and skim the introduction/conclusion): Note: This is an unpublished draft and should not be circulated.

• Why Inclusion Matters for the Future of AI

• AI and Human Development (read p. 9-16 )

• Browse/Skim https://aiandinclusion.org/

Optional/further reading if students are interested

• Africa, Access to Opportunity, and AI

• The State of Open Data: Histories and Horizons (Chapter 34)

• Urban Transport in the Sharing Economy Era

• AI Bias

• Many Facial Recognition Systems are Biased (NYTimes)

Time: Thursday, July 9 at 3pm (Zoom link will be shared on Slack at in the Announcements channel on the day of the session)
Session Leads: Bruce Schneier

Description: Writing for a popular audience is essential to getting your work and ideas into the public sphere. Bruce Schneier will talk about writing short opinionated essays: both the essays themselves and the business of getting them published.

Time: Friday, July 10 at 3pm (Zoom link will be shared on Slack at in the Announcements channel on the day of the session)
Session Leads: Chinmayi Arun

Guiding Questions

1. Bias and discrimination are problems that the world grapples with in many contexts. How is bias in AI different and why should we be concerned about it?

2. Consider the dimensions and scale of impact that bias in AI might have

3. Is bias in AI inevitable, or is it possible to take steps to mitigate it? What kinds of intervention do you think might help?

Required Readings

• • Langdon Winner's Do Artifacts Have Politics

  • Jonathan Zittrain, The Hidden Costs of Automated Thinking

  • West et. al, Discriminating Systems

  • Chinmayi Arun, AI & the Global South (SSRN link so doesn't go straight to a pdf)

  • Safiya Noble, Missed Connections: What Search Engines Say about Women (Trigger Warning)

Additional Readings

• Barocas & Sebst, Big Data's Disparate Impact

• Ajunwa et al, Hiring by Algorithm

• Frank Pasquale, Data Informed Duties in AI Development

• Rashida Richardson et al, Dirty Data, Bad Predictions