As part of my work as a postdoc, I ran 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
- 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
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
- 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.