We're very excited to welcome you to this advanced NLP seminar! We will cover ethical and social issues in NLP, especially focusing on large language models/foundation models. We'll cover research ethics, IRB, data and participatory research, harms and their mitigation, what models we should build and who should decide, and how to apply NLP to social questions about the justice system, framing and dehumanization, disinformation and toxicity, and for counter speech.
This is an advanced seminar with a lot of reading! We'll meet in breakout groups, and then bring the discussion back to the whole group! Attendence is strictly required, no hybrid option.
Students who may need an academic accommodation based on the impact of a disability must initiate the request with the Office of Accessible Education (OAE). Professional staff will evaluate the request with required documentation, recommend reasonable accommodations, and prepare an Accommodation Letter for faculty dated in the current quarter in which the request is being made. Students should contact the OAE as soon as possible since timely notice is needed to coordinate accommodations. The OAE is located at 563 Salvatierra Walk (phone: 723-1066, URL: http://oae.stanford.edu). If you already have an Academic Accommodation Letter, we invite you to share your letter with us. Academic Accommodation Letters should be shared at the earliest possible opportunity so we may partner with you and OAE to identify any barriers to access and inclusion that might be encountered in your experience of this course.- Well-Being, Stress Management, and Mental Health: If you are experiencing personal, academic, or relationship problems and would like someone to talk to, reach out to the Counseling and Psychological Services (CAPS) on campus. CAPS is the university’s counseling center dedicated to student mental health and wellbeing. Phone assessment appointments can be made at CAPS by calling 650-723- 3785, or by accessing the VadenPatient portal through the Vaden website. For more information, visit: https://vaden.stanford.edu/caps-and-wellness.
Each week of this class focuses either on ways to avoid ethical or social problems in doing NLP research (we call these Red Weeks “(NLP Should) Do No Harm”) or on ways to apply NLP to help solve social or ethical problems (we call these Blue weeks “(NLP Should) Do Good”). We've tried to give you extra papers beyond the 2-3 we read for each session, the extra papers are designed to be useful for helping come up with project ideas or literature surveys.
Week | Date | Description | Course Materials | Deadlines |
---|---|---|---|---|
1 | April 6 Thursday |
Part I: 3:00-4:00 Introduction to the class and each other
[slides pptx, pdf]
Part II: 4:10-5:20 Where does the data come from? The Belmont Report, Participants, Labelers, and Data in NLP [slides pptx, pdf] |
Required Readings:
Plus either one of the following two papers:
Further Readings for Projects and Background I'll be lecturing on these first two papers:
Emily M. Bender and Batya Friedman. 2018. Data statements for NLP: Toward mitigating system bias and enabling better science. TACL 6, 587–604. Casey Fiesler and Nicholas Proferes. 2018. “Participant” Perceptions of Twitter Research Ethics. Social Media + Society, 4(1). 22 Timnit Gebru, Jamie Morgenstern, Briana Vecchione, Jennifer Wortman Vaughan, Hanna Wallach, Hal Daumé III, Kate Crawford. 2020. Datasheets for Datasets. Arxiv. Vitak, J., Shilton, K., Beyond the Belmont principles: Ethical challenges, practices, and beliefs in the online data research community. In Proceedings of the 19th ACM conference on computer-supported cooperative work & social computing (pp. 941-953). Williams, M. L., Burnap, P., Towards an Ethical Framework for Publishing Twitter Data in Social Research: Taking into Account Users’ Views, Online Context and Algorithmic Estimation. Sociology, 51(6), 1149–1168.
Shuster, Evelyne. 1997. Fifty years later: the significance of the Nuremberg Code." New England Journal of Medicine 337, 20: 1436-1440. The Common Rule: The Federal Policy for the Protection of Human Subjects. 45 CFR part 46, Kobi Leins and Jey Han Lau and Timothy Baldwin. 2020. Give Me Convenience and Give Her Death: Who Should Decide What Uses of NLP are Appropriate, and on What Basis?. ACL 2020 Rickford, John Russell. "Unequal partnership: Sociolinguistics and the African American speech community." Language in Society 26, no. 2 (1997): 161-197.
|
Read the Belmont plus one of the other papers before class. No need to write paragraphs for today. |
2 | April 13 Thursday |
Part I: 3:00-4:00 The role of the local community, participatory research, and decolonization Part II: 4:10-5:20 Harms of Language Models: Surveys and Background |
Part I: Read any two of these three papers:
Langdon Winner. 1980. “Do Artifacts have Politics?”, Daedalus,109 (1): 121-136 Leon Derczynski, Hannah Rose Kirk, Vidhisha Balachandran, Sachin Kumar, Yulia Tsvetkov, M. R. Leiser, Saif Mohammad. 2023. Assessing Language Model Deployment with Risk Cards.
Further Readings for Projects and Background:
Steven Bird. 2022. Local Languages, Third Spaces, and other High-Resource Scenarios. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 7817–7829, Dublin, Ireland. ACL. Shakir Mohamed, Marie-Therese Png, William Isaac, 2020. Decolonial AI: Decolonial Theory as Sociotechnical Foresight in Artificial Intelligence. Philosophy and Technology 33: 659–684 . Julia Kreutzer, Isaac Caswell, Lisa Wang, Ahsan Wahab, Daan van Esch, Nasanbayar Ulzii-Orshikh, Allahsera Tapo, Nishant Subramani, Artem Sokolov, Claytone Sikasote, Monang Setyawan, Supheakmungkol Sarin, Sokhar Samb, Benoît Sagot, Clara Rivera, Annette Rios, Isabel Papadimitriou, Salomey Osei, Pedro Ortiz Suarez, Iroro Orife, Kelechi Ogueji, Andre Niyongabo Rubungo, Toan Q. Nguyen, Mathias Müller, André Müller, Shamsuddeen Hassan Muhammad, Nanda Muhammad, Ayanda Mnyakeni, Jamshidbek Mirzakhalov, Tapiwanashe Matangira, Colin Leong, Nze Lawson, Sneha Kudugunta, Yacine Jernite, Mathias Jenny, Orhan Firat, Bonaventure F. P. Dossou, Sakhile Dlamini, Nisansa de Silva, Sakine Çabuk Ballı, Stella Biderman, Alessia Battisti, Ahmed Baruwa, Ankur Bapna, Pallavi Baljekar, Israel Abebe Azime, Ayodele Awokoya, Duygu Ataman, Orevaoghene Ahia, Oghenefego Ahia, Sweta Agrawal, Mofetoluwa Adeyemi. 2022. Quality at a Glance: An Audit of Web-Crawled Multilingual Datasets. Transactions of the Association for Computational Linguistics 2022; 10 50–72 Lane Schwartz. 2022. Primum Non Nocere: Before working with Indigenous data, the ACL must confront ongoing colonialism. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 724–731, Dublin, Ireland. Association for Computational Linguistics. https://aclanthology.org/2022.acl-short.82.pdf Sebastian Ruder, Ivan Vulić, and Anders Søgaard. 2022. Square One Bias in NLP: Towards a Multi-Dimensional Exploration of the Research Manifold. In Findings of the Association for Computational Linguistics: ACL 2022, pages 2340–2354, Dublin, Ireland. Association for Computational Linguistics. https://aclanthology.org/2022.findings-acl.184/ Surangika Ranathunga and Nisansa de Silva. 2022. Some Languages are More Equal than Others: Probing Deeper into the Linguistic Disparity in the NLP World. AACL/IJCNLP 2022, 823–848. Heather Lent, Kelechi Ogueji, Miryam de Lhoneux, Orevaoghene Ahia, and Anders Søgaard. 2022. What a Creole Wants, What a Creole Needs. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 6439–6449, Marseille, France. European Language Resources Association. David Jurgens, Yulia Tsvetkov, and Dan Jurafsky. 2017. Incorporating Dialectal Variability for Socially Equitable Language Identification. ACL 2017. Vithya Yogarajan, Gillian Dobbie, Henry Gouk. 20023. Effectiveness of Debiasing Techniques: An Indigenous Qualitative Analysis. ICLR TinyPaper 2023 Allison Koenecke, Andrew Nam, Emily Lake, Joe Nudell, Minnie Quartey, Zion Mengesha, Connor Toups, John Rickford, Dan Jurafsky, and Sharad Goel. 2020. Racial Disparities in Automated Speech Recognition. Proceedings of the National Academy of Sciences 117 (14) 7684-7689. Bender, Emily M., Timnit Gebru, Angelina McMillan-Major, and Shmargaret Shmitchell. 2021. On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?🦜. FAccT '21: Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency March 2021 Pages 610–623 Weidinger, Laura, Jonathan Uesato, Maribeth Rauh, Conor Griffin, Po-Sen Huang, John Mellor, Amelia Glaese, Myra Cheng, Borja Balle, Atoosa Kasirzadeh, Courtney Biles, Sasha Brown, Zac Kenton, Will Hawkins, Tom Stepleton, Abeba Birhane, Lisa Anne Hendricks, Laura Rimell, William Isaac, Julia Haas, Sean Legassick, Geoffrey Irving, and Iason Gabriel. 2022. Taxonomy of risks posed by language models. In 2022 ACM Conference on Fairness, Accountability, and Transparency, pp. 214-229. 2022. Laura Weidinger, John Mellor, Maribeth Rauh, Conor Griffin, Jonathan Uesato, Po-Sen Huang, Myra Cheng, Mia Glaese, Borja Balle, Atoosa Kasirzadeh, Zac Kenton, Sasha Brown, Will Hawkins, Tom Stepleton, Courtney Biles, Abeba Birhane, Julia Haas, Laura Rimell, Lisa Anne Hendricks, William Isaac, Sean Legassick, Geoffrey Irving, and Iason Gabriel. 2021. Ethical and social risks of harm from Language Models. arXiv:2112.04359 [cs] (Dec. 2021). Shelby, Renee, Shalaleh Rismani, Kathryn Henne, AJung Moon, Negar Rostamzadeh, Paul Nicholas, N'mah Yilla-Akbari, Jess Gallegos, Andrew Smart, Emilio Garcia, and Gurleen Virk. 2023. Identifying Sociotechnical Harms of Algorithmic Systems: Scoping a Taxonomy for Harm Reduction. arXiv preprint arXiv:2210.05791 (2022). A. Stevie Bergman, Gavin Abercrombie, Shannon Spruit, Dirk Hovy, Emily Dinan, Y-Lan Boureau, and Verena Rieser. 2022. Guiding the Release of Safer E2E Conversational AI through Value Sensitive Design. In Proceedings of the 23rd Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 39–52, Edinburgh, UK. Association for Computational Linguistics. Ganguli, Deep, Liane Lovitt, Jackson Kernion, Amanda Askell, Yuntao Bai, Saurav Kadavath, Ben Mann, Ethan Perez, Nicholas Schiefer, Kamal Ndousse, Andy Jones, Sam Bowman, Anna Chen, Tom Conerly, Nova DasSarma, Dawn Drain, Nelson Elhage, Sheer El-Showk, Stanislav Fort, Zac Hatfield-Dodds, Tom Henighan, Danny Hernandez, Tristan Hume, Josh Jacobson, Scott Johnston, Shauna Kravec, Catherine Olsson, Sam Ringer, Eli Tran-Johnson, Dario Amodei, Tom Brown, Nicholas Joseph, Sam McCandlish, Chris Olah, Jared Kaplan, Jack Clark. 2022. Red teaming language models to reduce harms: Methods, scaling behaviors, and lessons learned. arXiv preprint arXiv:2209.07858 (2022). |
Read and post paragraphs by 5pm Wednesday April 12. |
3 | April 20 Thursday |
Part I: 3:00-4:00 Harms of Language Models: Bias and Stereotype
Part II: 4:10-5:20 Harms of LLMs: Influence |
Part I: Read two of these three papers:
|
Read and post paragraphs by 5pm Wednesday April 18 |
4 | April 27 Thursday |
Part I: 3:00-4:00 Harms of Language Models: Intellectual Property
Part II: 4:10-5:20 Harms of Language Models: Privacy and Libel |
Part I: Read these two papers:
Part II: Read these 3 very short pieces (or in one case, a pair of pieces) and write 1 paragraph on Volokh/Kerr and 1 paragraph on the Italy case; don't write comments on the first reading (Section 5.4 "Legality"):
|
Read and post paragraphs by 5pm Wednesday April 26. |
5 | May 4 Thursday |
Dan Sick: Class Cancelled | ||
6 | May 11 Thursday |
Part I: 3:00-4:00: Regulation (of Large Language Models, and AI in general)
Part II: 4:10-5:20: The Role of the Technical Community: Should we build it? And other actions we can take. |
Part I: Do these readings and then instead of separate paragraphs on the readings, write the reflection as described in the Ed announcement
Part II: Do these readings and then instead of separate paragraphs on the readings, write the reflection as described in the Ed announcement
|
Read and post paragraphs by 5pm Wednesday May 10 |
7 | May 18 Thursday |
Part I: 3:00-4:00: AI for the people
Part II: 4:10-5:20: Dreaming up AI |
Part I: AI for the people
Part II: Dreaming up AI
|
Read and post paragraphs by 5pm Wednesday May 17. Lit Review due Mon May 15 5:00pm |
8 | May 25 Thursday |
Part I: 3:00-4:00: Using NLP for Social Change: Survey and Applications to Policing
Part II: 4:10-5:20: NLP for Social Good: Applications in Counter Speech |
Part I: Read these two papers:
Part II: Read this paper:
|
Read and post paragraphs by 5pm Wednesday May 24. Project Proposal due Friday May 26 5:00pm |
9 | May 29 Thursday in classroom |
Individual meetings with Dan, Ria, Peter on projects. | ||
10 | June 8 Thursday |
Special Class Presentation Day (regular classes end June 7). Other days this week: more individual meetings with Dan, Ria, Peter on projects. | Final Project Report due Mon June 12, 5:00pm |