AI is poised to have an enormous impact on society. What should that impact be and who should get to decide it? The goal of this course is to critically examine AI research and deployment pipelines, with in-depth examinations of how we need to understand social structures to understand impact. In application domains, we will examine questions like “who are key stakeholders?”, “who is affected by this technology?” and “who benefits from this technology?”. We will also conversely examine: how can AI help us learn about these domains, and can we build from this knowledge to design AI for "social good"? As a graduate-level course, topics will focus on current research including development and deployment of technologies like large language models and decision support tools, and students will conduct a final research project.

Prerequisites: At least one graduate-level computer science course in Artificial Intelligence or Machine Learning (including NLP, Computer Vision, etc.), two preferred, or permission of the instructor. Students must be comfortable with reading recent research papers and discussing key concepts and ideas.

Acknowledgements Thank you to Dan Jurafsky for sharing course materials and to Daniel Khashabi for sharing the course website template!

Schedule

The current class schedule is below. The schedule is subject to change, particularly the specific readings:

Date Topic Readings Work Due
Tue Aug 27 Introduction, Origins and Data [slides]
  1. Avoiding Past Mistakes in Unethical Human Subjects Research: Moving From Artificial Intelligence Principles to Practice
Thu Aug 29 No class
Tue Sept 3 Data: Ownership [slides]
  1. Lundberg, Ian, et al. "Privacy, ethics, and data access: A case study of the Fragile Families Challenge." Socius 5 (2019)
  2. Buolamwini, Joy, and Timnit Gebru. "Gender shades: Intersectional accuracy disparities in commercial gender classification." FAccT. PMLR, 2018
  3. "Facial recognition's 'dirty little secret': Millions of online photos scraped without consent", NBC article
Reading Responses by 8pm Monday; course goals survey by class 9/10
Thu Sept 5 Data: Privacy
  1. Carlini, Nicolas, et al. "Extracting training data from diffusion models." 32nd USENIX Security Symposium (USENIX Security 23). 2023.
  2. Hannah Brown, Katherine Lee, Fatemehsadat Mireshghalla, Reza Shokri, Florian Tramèr. "What Does it Mean for a Language Model to Preserve Privacy?" FAccT 2022.
Reading Responses by 8pm Wednesday
Tue Sept 10 Data: Crowdsourcing
  1. Boaz Shmueli, Jan Fell, Soumya Ray, and Lun-Wei Ku. 2021. Beyond Fair Pay: Ethical Implications of NLP Crowdsourcing. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 3758–3769, Online. Association for Computational Linguistics.
  2. "Exclusive: OpenAI Used Kenyan Workers on Less Than $2 Per Hour to Make ChatGPT Less Toxic", Time
Reading Responses by 8pm Monday
Thu Sept 12 Fairness, Bias and Stereotypes: Fairness metrics
  1. Dwork, Cynthia, et al. "Fairness through awareness." Proceedings of the 3rd innovations in theoretical computer science conference. 2012.
  2. Chouldechova, Alexandra. "Fair prediction with disparate impact: A study of bias in recidivism prediction instruments", Big Data, Special issue on Social and Technical Trade-Offs. 2017.
  3. (optional) Corbe-Davies, Sam et al. "Algorithmic Decision Making and the Cost of Fairness", KDD. 2017.
  4. (optional) Mehrabi, Ninareh et al. "A Survey on Bias and Fairness in Machine Learning", ACM Computing Surveys. 2021.
Reading Responses by 8pm Wednesday
Tue Sept 17 Fairness, Bias and Stereotypes: Bias in Classification [slides]
  1. Zhao et al. "Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints", EMNLP. 2017.
  2. Obermeyer et al. "Dissecting racial bias in an algorithm used to manage the health of populations," Science, 2019.
  3. (optional) Sap et al. "The Risk of Racial Bias in Hate Speech Detection", ACL. 2019.
Reading Responses by 8pm Monday
Thu Sept 19 Fairness, Bias and Stereotypes: Stereotypes in Generation [slides]
  1. Feng et al. "From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Models", ACL 2023.
  2. Bianchi, Federico, et al. "Easily accessible text-to-image generation amplifies demographic stereotypes at large scale." FAccT 2023.
  3. (optional) Myra Cheng, Esin Durmus, and Dan Jurafsky. "Marked Personas: Using Natural Language Prompts to Measure Stereotypes in Language Models", ACL 2023.
Reading Responses by 8pm Wednesday
Tue Sept 24 Values and Design: Value sensitive design
  1. Friedman, Batya, et al. "Value sensitive design and information systems." Early engagement and new technologies: Opening up the laboratory (2013): 55-95.
  2. Umbrello, Steven, and Ibo Van de Poel. "Mapping value sensitive design onto AI for social good principles." AI and Ethics 1.3 2021.
Reading Responses by 8pm Monday
Thu Sept 26 Values and Design: Participatory design
  1. Brown, et al. 2019. Toward Algorithmic Accountability in Public Services: A Qualitative Study of Affected Community Perspectives on Algorithmic Decision-making in Child Welfare Services. CHI 2019.
  2. Sloane et al. Participation is not a Design Fix for Machine Learning. EAAMO 2022.
Reading Responses by 8pm Wednesday
Tue Oct 1 Values and Design: Surveyed Values
  1. Birhane et al. The Values Encoded in Machine Learning Research. FAccT 2022.
  2. Widder et al. It’s about power: What ethical concerns do software engineers have, and what do they (feel they can) do about them? FAccT 2023.
Reading Responses by 8pm Monday
Thu Oct 3 Societal Impact: Human-in-the-loop Decision Making
  1. Dasha Pruss. Ghosting the Machine: Judicial Resistance to a Recidivism Risk Assessment Instrument. FAccT 2023.
  2. Alex Albright. If You Give a Judge a Risk Score: Evidence from Kentucky Bail Decisions. 2019 [Only need to read the Abstract and Introduction]
Reading Responses by 8pm Wednesday
Tue Oct 8 Societal Impact: Hallucinations and Misinformation
  1. Kreps, Sarah, R. Miles McCain, and Miles Brundage. "All the news that’s fit to fabricate: AI-generated text as a tool of media misinformation." Journal of experimental political science 9.1 (2022): 104-117.
  2. Koenecke, Allison, et al. "Careless Whisper: Speech-to-Text Hallucination Harms." FAccT 2024.
Reading Responses by 8pm Monday
Thu Oct 10 Societal Impact: Overview
  1. Sayash Kapoor*, Rishi Bommasani* et al. "On the Societal Impact of Open Foundation Models." ICML 2024
  2. Harry H. Jiang et al. "AI Art and its Impact on Artists." AIES 2023
Reading Responses by 8pm Wednesday
Tue Oct 15 Proposal Presentations [Prepare for proposal presentations]
Thu Oct 17 Fall Break
Tue Oct 22 Societal Impact: Environmental Impact
  1. Strubell, Emma, Ananya Ganesh, and Andrew McCallum. "Energy and Policy Considerations for Deep Learning in NLP." Association for Computational Linguistics. 2019.
  2. Kaack, Lynn H., et al. "Aligning artificial intelligence with climate change mitigation." Nature Climate Change 12.6 (2022): 518-527.
Reading Responses by 8pm Monday
Thu Oct 24 Policy and Regulation: Research Codes of Ethics
  1. ACM Code of Ethics
  2. NeurIPS Code of Ethics
  3. Code of ethics from one other venue of your choice
Reading Responses by 8pm Wednesday
Tue Oct 29 Policy and Regulation: Company internal ethics review
  1. Emily Bender et al. "On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?." In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, 2021.
  2. "Meta created a ‘Supreme Court’ for content. Then it threatened its funds.", Washington Post article
Reading Responses by 8pm Monday
Thu Oct 31 Policy and Regulation: United States
  1. AI Bill of Rights
  2. White House executive order fact sheet
  3. White House executive order, sections 1, 2, and 4 only
Reading Responses by 8pm Wednesday
Tue Nov 5 Policy and Regulation: Europe and China
  1. European Union AI Act briefing
  2. "China Wants to Regulate Its Artificial Intelligence Sector Without Crushing It", Times article
  3. "Four things to know about China’s new AI rules in 2024", MIT Technology Review article
Reading Responses by 8pm Monday
Thu Nov 7 AI Opportunities: Policing
  1. Rob Voigt et al. Language from police body camera footage shows racial disparities in officer respect. PNAS 2017
  2. Matt Franchi et al. Detecting disparities in police deployments using dashcam data. FAccT 2023
Reading Responses by 8pm Wednesday
Tue Nov 12 AI Opportunities: Healthcare
  1. Emma Pierson et al. An algorithmic approach to reducing unexplained pain disparities in underserved populations. Nature Medicine 2021
  2. Wolf et al. The SEE Study: Safety, Efficacy, and Equity of Implementing Autonomous Artificial Intelligence for Diagnosing Diabetic Retinopathy in Youth. Diabetes Care. 2021.[Read just the first few paragraphs before the "RESEARCH DESIGN AND METHODS" section. This is to provide some additional context for #3]
  3. Wolf et al. Clinical Implementation of Autonomous Artificial Intelligence Systems for Diabetic Eye Exams: Considerations for Success. Practical Pointers. 2023
Reading Responses by 8pm Monday
Thu Nov 14 AI Opportunities: Education
  1. Han et al. Teachers, Parents, and Students' perspectives on Integrating Generative AI into Elementary Literacy Education. CHI 2024.
  2. "In Classrooms, Teachers Put A.I. Tutoring Bots to the Test", New York Times article
Reading Responses by 8pm Wednesday
Tue Nov 19 Guest Lecture: Translating Ethical Principles into AI Governance, Debra Matthews
  1. Imagining Governance for Emerging Technologies, Issues in Science and Technology. (optional)
  2. Governance of Emerging Technologies in Health and Medicine — Creating a New Framework, The New England Journal of Medicine. (optional)
Thu Nov 21 Guest Lecture: Kadija Ferryman
  1. Considering Biased Data as Informative Artifacts in AI-Assisted Health Care, The New England Journal of Medicine, 2023.

Policies

Attendance policy This is a graduate-level course revolving around in-person discussion. Students are expected to attend class and may notify instructors if there are extenuating circumstances.

Course Conduct This is a discussion class focused on controversial topics. All students are expected to respect everyone's perspective and input and to contribute towards creating a welcoming and inclusive climate. We the instructors will strive to make this classroom an inclusive space for all students, and we welcome feedback on ways to improve.

Academic Integrity This course will have a zero-tolerance philosophy regarding plagiarism or other forms of cheating, and incidents of academic dishonesty will be reported. A student who has doubts about how the Honor Code applies to this course should obtain specific guidance from the course instructor before submitting the respective assignment.

Discrimination and Harrasment The Johns Hopkins University is committed to equal opportunity for its faculty, staff, and students. To that end, the university does not discriminate on the basis of sex, gender, marital status, pregnancy, race, color, ethnicity, national origin, age, disability, religion, sexual orientation, gender identity or expression, veteran status, military status, immigration status or other legally protected characteristic. The University's Discrimination and Harassment Policy and Procedures provides information on how to report or file a complaint of discrimination or harassment based on any of the protected statuses listed in the earlier sentence, and the University’s prompt and equitable response to such complaints.

Personal Well-being Take care of yourself! Being a student can be challenging and your physical and mental health is important. If you need support, please seek it out. Here are several of the many helpful resources on campus: