As part of my work as a postdoc, I run a reading group for undergraduate and master's students on participatory AI systems. This is the reading list we used for spring semester 2023.
- Fairness, bias, and equity in AI
- Industry Practice and Human-AI Systems
- Global Implications of AI Systems
Fairness, bias, and equity in AI
Quantifying bias and fairness, part I
- ‼️Ochigame, “The Long History of Algorithmic Fairness”
- Friedman and Nissenbaum, “Bias in Computer Systems”
- ‼️Narayanan, 21 fairness definitions and their politics (55 min)
- Blodgett et al, “Language (Technology) is Power: A Critical Survey of “Bias” in NLP”
Quantifying fairness, part II
- ‼️Shah and Bender, “Situating Search” (CHIIR ‘22)
- Heaven, “Chatbots could one day replace search engines. Here’s why that’s a terrible idea”
- Bolukbasi et al, “Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings” (NEURIPS 2016)
- ‼️ Wachter et al, “Health care bias is dangerous. But so are ‘fairness’ algorithms.” https://www.wired.com/story/bias-statistics-artificial-intelligence-healthcare/
Medical racism and computational diagnosis
- Dawes et al, “Clinical versus actuarial judgment”
- Obermeyer et al, “Dissecting racial bias in an algorithm used to manage the health of populations”
- ‼️Barabas et al, “Interventions over predictions: reframing the ethical debate for actuarial risk assessment”
- Xiaowei Wang, “A New AI Lexicon: Care.” AI Now Institute, 2021.
Large language models and automated text
- ‼️Chiang, “Chat GPT is a Blurry JPEG of the Web,” New Yorker
- Counterpoint (please read the ensuing discussion, not just the thread): Andrew Lampinen on Twitter
- See also John Burns-Murdoch’s thread (and ensuing discussion)
- Narayanan, “Testing discrimination in practice” (on fairness criteria)
- ‼️Bender et al, “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜”
- Henderson et al, “Ethical Challenges in Data-Driven Dialogue Systems”
Industry Practice and Human-AI Systems
Practitioner perspectives on designing human-AI systems
- Bogdana Rakova et al., “Where Responsible AI Meets Reality: Practitioner Perspectives on Enablers for Shifting Organizational Practices,” Proceedings of the ACM on Human-Computer Interaction 5, no. CSCW1 (April 22, 2021): 7:1–7:23, https://doi.org/10.1145/3449081.
- Saleema Amershi et al., “Guidelines for Human-AI Interaction,” in Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (the 2019 CHI Conference, ACM Press, 2019), 1–13, https://doi.org/10.1145/3290605.3300233.
- ‼️Tom Zick, “Building better startups with responsible AI” (Techcrunch)
MIT goes to Congress
Advances in AI: Are We Ready For a Tech Revolution?
For this week, let’s read all of the prepared statements from the expert witnesses (see below). If you have time, watch the Q&A section from the hearing, which starts about halfway through.
- Dr. Eric Schmidt Chair, Special Competitive Studies Project
- Dr. Scott Crowder Vice President / CTO, IBM Quantum / IBM Systems, Technical Strategy, and Transformation
- Dr. Aleksander Mądry Director / Cadence Design Systems Professor of Computing, MIT Center for Deployable Machine Learning / Massachusetts Institute of Technology
- Ms. Merve Hickok Chair and Research Director, Center for AI and Digital Policy
“Hard core” tech work: layoffs, labor organizing, and the craft of software engineering
- ‼️Tarnoff, B., & Weigel, M. (2020). Voices from the Valley: Tech Workers Talk About What They Do--and How They Do It. Farrar, Straus and Giroux.
- This book _looks _really long (162 pages!) but the margins are pretty big and the entire thing is written really conversationally — it’s 7 long interviews by people who work in tech that have been edited down. I honestly don’t think it would take much longer than reading 2 (even 1) journal articles.
- Yung Au, “A New AI Lexicon: Exporting AI.” AI Now Institute (2021)
Global Implications of AI Systems
Ethics of NLP in Africa
- ‼️Birhane, “Algorithmic Colonization of Africa”
- Mohamed et al, “Decolonial AI: Decolonial Theory as Sociotechnical Foresight in Artificial Intelligence”
- Adelani et al, “MasakhaNER: Named Entity Recognition for African Languages”
- Mhlambi, “Ethical Implications of AI and Ubuntu as an Intervention”
Reframing AI Governance: Perspectives from Asia
Arun, “AI and the Global South: Designing for Other Worlds”
Png, “At the Tensions of South and North: Critical Roles of Global South Stakeholders in AI Governance”
- The Beginnings of AI and Data Governance: Landscape in Sri Lanka
- To What Extent Does Malaysia’s National Fourth Industrial Revolution Policy Address AI Security Risks?
- An Ill-advised Turn: AI Under India's e-Courts Proposal
- Chinese AI Governance in Transition: Past, Present, and Future of Chinese AI Regulation
- Myth of Data-Driven Authoritarianism in Asia
- Kampong Ethics
- Between Threat and Tool: The Poetics and Politics of AI Metaphors and Narratives in China
Ghost work and global data annotation
- Di Wu, “Good for tech: Disability expertise and labor in China's artificial intelligence sector.” First Monday (2022).
- Karishma Mehrotra, “Human Touch,” Fifty Two (2022).
- ‼️Mary Gray, “Ghost Work in Pandemic Times,” CUNY Graduate Center (2021).
- Note: The full video is a little over an hour long, and while I think it’s worth watching the whole thing, the most important part is the lecture (starts at 4:48, ends at 34:56). The entire video is also captioned, so it’s especially easy to watch at 2x speed if you are so inclined.