Term: Aug - Dec’ 25
Time: 10 – 11:30 AM
Venue: CDS 419
Credits: 3:0
Course Details
Outline: We interact with AI technology on a daily basis—such systems answer the questions we ask (using Google, or other search engines), curate the content we read, unlock our phones, allow entry to airports, etc. Further, with the recent advances in large language and vision models, the impact of such technology on our lives is only expected to grow. This course introduces students to ethical implications associated with design, development and deployment of AI technology spanning NLP, Vision and Speech applications.
This is a seminar-style course, wherein each class would be a discussion based on the readings assigned for that particular day. Each class would begin with a short quiz, which would be straightforward if you have read through the required section of the reading material. The in-class discussion among students would be facilitated by the instructor and teaching assistants who would bring in discussion points based on the reading material.
Prerequisites: The class is intended for graduate students and senior undergraduates. Students would most benefit if they have finished at least a basic machine learning course from IISc (or similar-quality course elsewhere) and any one courses related to the discussed applications (computer vision, speech or NLP). There are no hard pre-requisites, we trust students to self assess whether this course is for them.
Content: Specifically, this seminar course would facilitate discussions among students structured around pre-selected readings on topics related to ethics in AI. Specifically, we plan to read about and discuss following modules:
- M1. Overview of ethical theories
- M2. Data collection and curation
- M3. Biases and algorithmic fairness; debiasing and mitigating harms
- M4. Privacy
- M5. Content Moderation: Misinformation, disinformation and hate-speech
- M6. Understanding AI systems: transparency, accountability, (lack of) interpretability
- M7. Algorithm audits and transparency
- M8. Environmental impact of model training and inference
- M9. Future of work; economic impact of AI
Intentionally, many of the modules are connected to each other. We hope to spend at least 2-4 classes discussing each module.
Schedule:
This is tentative schedule and subject to change.
| Date |
Topic |
Reading Material |
| Aug 6 |
Course Overview |
– |
| Aug 11 |
(M1) Major Ethical Theories |
Required: (a) An overview of ethical theories |
| Aug 13 |
(M1) Values in NLP/ML research |
Required: (a)The Social Impact of NLP; and (b) The Values Encoded in ML Research (pages 1-15) |
| Aug 18 |
(M2) Data Cascades |
Required: “Everyone wants to do the model work, not the data work" |
| Aug 20 |
(M2) Data documentation |
Required: Datasheets for datasets (pages 1-10). Recommended: Data Statements for NLP |
| Aug 25 |
(M2) Data collection |
Required: IndicVoices: Towards building an Inclusive Multilingual Speech Dataset for Indian Languages. Recommended: IndicVoices Blog, Ethical Data Pledge, OpenAI Used Kenyan Workers on Less Than $2 Per Hour to Make ChatGPT Less Toxic |
| Aug 27 |
Institute Holiday |
– |
| Sep 1 |
(M3) Algorithmic Bias |
Required: (a) Moving beyond ‘‘algorithmic bias is a data problem’’, and (b) GenderShades. Recommended: Facial Recognition Is Accurate, if You’re a White Guy |
| Sep 3 |
(M3) Algorithmic Bias |
Required: Language (Technology) is Power: A Critical Survey of “Bias” in NLP (pages 1-9). |
| Sep 8 |
(M3) Cultural Bias |
Required: Investigating Cultural Alignment of Large Language Models and Towards Measuring and Modeling “Culture” in LLMs: A Survey. Recommended: Dollars Street Dataset, Large Language Models are Geographically Biased |
| Sep 10 |
(M3) Debiasing and mitigation |
Required: TBD . Recommended: Mitigating Gender Bias in NLP |
| Sep 15 |
(M4) Privacy |
Required: What Does it Mean for a Language Model to Preserve Privacy? Recommended: Differential Privacy: A Primer |
| Sep 17 |
(M4) Privacy |
Required: Extracting Training Data from Large Language Models, What does GPT-3 “know” about me? |
| Sep 22 |
(M5) Content Moderation |
Required: (a) When Curation Becomes Creation: Algorithms, Microcontent, and the Vanishing Distinction between Platforms and Creators. and (b) TBD. |
| Sep 24 |
(M5) Content Moderation |
Required: (a) Content moderation, AI, and the question of scale and (b) Do Not Recommend? Reduction as a Form of Content Moderation |
| Sep 29 |
(M5) Content Moderation |
Required: TBD. Recommended: Decolonizing Content Moderation, AI Content Moderation, Racism and (de)Coloniality |
| Oct 1 |
Dussehra Break |
– |
| Oct 6 |
Peer feedback on project proposal |
– |
| Oct 8 |
(M6) Misinformation, disinformation |
Required: TBD |
| Oct 13 |
(M6) Misinformation, disinformation |
Required: TBD. Images and misinformation in political groups: Evidence from WhatsApp in India and Tiplines to Combat Misinformation on WhatsApp |
| Oct 15 |
(M6) Misinformation, disinformation |
Required: TBD. A survey on automated fact-checking (till end of Section 2). Recommended: How to search for fact-checked information |
| Oct 20 |
Diwali Break |
– |
| Oct 22 |
(M7) Algorithm Audits |
Required: Auditing Algorithms Understanding Algorithmic Systems from the Outside: Chapter 2, 3 (also recommend reading 1 and 4). |
| Oct 29 |
(M7) Algorithm Audits |
Required: Unequal Representation and Gender Stereotypes in Image Search Results for Occupations. Recommended: An Image of Society: Gender and Racial Representation and Impact in Image Search Results for Occupations, Text-to-Image Generation Amplifies Demographic Stereotypes at Large Scale |
| Nov 3 |
(M8) Environmental Impact |
Required: Environmental Section of Foundations Models Paper (pages 140 - 145) and Power Hungry Processing: Watts Driving the Cost of AI Deployment?. Recommended: Estimating the Carbon Footprint of BLOOM,a 176B Parameter Language Model, Systematic Reporting of the Energy and Carbon Footprints of Machine Learning, Green AI: 1, 2, |
| Nov 5 |
Institute Holiday |
– |
| Nov 10 |
(M9) Future of work |
Required: (a) Future of work with AI agents and (b) Which economic activities are performed with AI. Recommended: Thousands of AI Authors on the Future of AI |
| Nov 12 |
(M9) Future of work |
Required: Generative AI at work |
Course Evaluation
The evaluation comprises 3 components:
- Class projects [50%]: we will shortly share more details about the course projects
- Readings & discussion [20%]: In-class quizzes, we would drop the lowest two scores.
- Final exam [30%]. There would be an exam towards the end of the course.
Class Projects
The course project constitutes 50% of the overall score, where students (with team sizes of no more than two) get a chance to apply the acquired knowledge.
The teaching staff would be keenly, and closely, involved in these projects, and students would have a choice to pick projects from a list of
planned and scoped ideas. Additionally, students can
also work on a project of their choice
so long as the project is related to (a) AI, and (b) there are ethical
aspects associated with it that merit serious discussion. These projects could take diverse forms, including case studies, qualitative research projects, model and/or data audits, algorithmic inquiries and solutions, etc. The project includes three milestones: (1) proposal along with a literature survey, which clearly states the problem, discusses relevant literature and outlines a rough action plan; a (2) mid-term report and a (3) final report. Towards the end of the course, students would get a chance to showcase their research through presentations or posters.
Late days: Each team would get three late days for projects, no extensions will be offered (please don’t even ask). After your late days expire, you can still submit your project but your obtained score would be divided by 2 if submitting after 1 day, and will be divided by 4 if submitting after 2 days. No submissions would be entertained after that.
Important Dates:
- TBD Project proposals due
- TBD Mid term reports due
- TBD Final reports due
Discussions & (Anonymous) Feedback
We will use Teams for all discussions on course-related matters. Registered students will receive the joining link/passkey.
If you have any feedback, you can share it (anonymously or otherwise) through this link: http://tinyurl.com/feedback-for-danish
Teaching Staff
- Archakam Satvikam Anudeep (Teaching Assistant)
- Danish Pruthi (Instructor)
Acknowledgements
We are grateful to Yulia Tsvetkov for readily sharing the content for her ethics classes at CMU and UW, and to Kinshuk Vasisht and Navreet Kaur for helping with the initial outline of this course.
-