Reading Group Discussion Leader (0 Points)

Purpose, Task, and Criteria

Purpose: To make you a leader and a critical consumer of the AI conversation by preparing, framing, and facilitating a real discussion for your peers.

Task: Select a timely, course-relevant source, frame it in a five-minute summary, facilitate a class discussion with three to five scaffolded questions, and submit a short reflection.

Criteria: Assessed on the relevance of your source, your framing, the depth and sequencing of your questions, your facilitation, and your reflection; see the rubric below. This assignment is extra credit, and the rubric scales the credit earned for your slot.

Assignment Goals

The goals of this assignment are:

Select and summarize a current event or reading relevant to course themes (AI systems, responsible AI, societal impact, technical foundations)
Craft three to five focused discussion questions that move from comprehension to analysis to personal judgment
Facilitate a class discussion that draws out multiple perspectives and reaches a synthesis conclusion
Reflect in writing on what you learned from leading the discussion and how it connects to course material

The Assignment

Overview

Each week, one or two students lead a short discussion on a current event or reading related to AI. This is entirely extra credit — there is no penalty for not doing it, but going earlier in the term earns significantly more credit.

The session structure is:

5 minutes: Student presenter summarizes the source and frames the central question.
10 minutes: Class discussion, facilitated by the student presenter.
1 minute: Student closes with a one-sentence synthesis of what the class concluded.

After your session, you submit a 300–500 word reflection write-up within one week.

For the Audience: Your Participation Counts

Leading a session is extra credit; being a thoughtful member of its audience is ordinary, expected participation, and it counts toward your Class Activities and Participation grade. A student-led discussion is only as good as the room it is held in, so when a classmate presents, you have a job.

Before the session, post a brief reading response to the presenter’s source — one takeaway, one question, and one connection to something we have built or run — or, when the source is circulated the same day, arrive with one genuine question ready. During the session, engage: build on the presenter’s framing, offer a counter-view, or connect their source to a system we have worked with. The Reading Responses guide explains the format, and audience engagement is assessed within the participation rubric. This is what makes the student-led sessions worth holding — and it means everyone, not only the presenter, has a stake in them.

Extra Credit Scale

Week	Extra Credit Points
Week 0 (first class meeting)	20
Week 1	18
Week 2	16
Week 3	14
Week 4	12
Week 5	10
Week 6	9
Week 7	8
Week 8	7
Week 9	6
Week 10	5
Week 11	4
Week 12	3
Week 13	2
Week 14	2
Week 15	1

Sign-up is first-come, first-served via the course sign-up sheet (link on the course LMS). Two students may sign up per class meeting. If a slot is full, choose the next available slot.

Important: You earn the extra credit points only if you complete both the in-class discussion and the written reflection on time. Earning the slot without presenting forfeits the credit.

What to Present On

You may choose any of the following:

Option A — Current Event: A news story, blog post, preprint, or policy announcement from the past 60 days that relates to AI. Examples: a new model release with notable capabilities or risks, a regulatory action, a high-profile AI failure or success, a research result.

Option B — Course-Adjacent Reading: A book chapter, article, or essay from the seed list below, or another source you propose (check with the instructor at least one week before your slot).

Option C — Determinism / Automation Bias Focus: A reading or event specifically about how humans interact with, over-trust, or mis-calibrate their trust in automated systems. This option connects directly to the Deterministic and Probabilistic Computing activity and is particularly welcome in the early weeks.

Alternative Format Option: Competing Texts Session

Any of the options above may instead be led as a Competing Texts Session: rather than presenting a single source, you stage two opposing sources on the same AI question — for example, an optimist piece and a skeptic piece on the same capability, risk, or policy. If you choose this format, your session must include:

A steelman summary of each source: present each author’s argument in its strongest form — the version its author would endorse — not a caricature.
One probing question per side: one question that presses on the optimist source’s weakest assumption, and one that presses on the skeptic source’s weakest assumption.
A closing reflection on which claims survive: close by identifying which claims from each text survive contact with the other text, and which do not.

The session structure and timing are unchanged: your 5-minute summary covers both texts (steelmanned), the probing questions count toward your three to five discussion questions, and your one-sentence closing synthesis states which claims survived. The same rubric applies.

Seed Reading List

These are suggested sources. You are not limited to this list — propose others with instructor approval.

Foundational Books (chapters or excerpts)

Crawford, K. (2021). Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press. (Chapters 1, 6, or 7)
Benjamin, R. (2019). Race After Technology: Abolitionist Tools for the New Jim Code. Polity. (Chapters 1–2)
Carr, N. (2014). The Glass Cage: How Our Computers Are Changing Us. W. W. Norton. (Chapters 2–3 on automation bias)
Christian, B. (2020). The Alignment Problem. W. W. Norton. (Any chapter)
Marcus, G. & Davis, E. (2019). Rebooting AI. Pantheon. (Chapters 3–4)
Wooldridge, M. (2021). A Brief History of Artificial Intelligence. Flatiron Books. (Any chapter)

Core and Technical Textbooks (chapters or excerpts)

These are the technical books assigned across the course schedule — a chapter you present here can double as a deeper dive on a topic we cover in a lecture.

Mitchell, M. (2019). Artificial Intelligence: A Guide for Thinking Humans. Farrar, Straus and Giroux. (Any chapter; e.g. Ch 1–2 on agents and history, Ch 3 on how models “understand,” Ch 4 on knowledge and reasoning, Ch 8 on meaning.) https://melaniemitchell.me/aibook/
Russell, S. & Norvig, P. (2020). Artificial Intelligence: A Modern Approach (4th ed.). Pearson. (Chapter 2 on intelligent agents; later chapters by topic.) https://aima.cs.berkeley.edu/
Jurafsky, D. & Martin, J. H. Speech and Language Processing (3rd ed. draft). (Ch 3 on n-gram language models; later chapters on embeddings and transformers.) https://web.stanford.edu/~jurafsky/slp3/
Nielsen, M. Neural Networks and Deep Learning. (Chapter 1, neural nets from first principles.) http://neuralnetworksanddeeplearning.com/
Sutton, R. & Barto, A. (2018). Reinforcement Learning: An Introduction (2nd ed.). MIT Press. (Chapters 1–6 for the RLHF thread.)
Huyen, C. (2022). Designing Machine Learning Systems. O’Reilly. (Chapter 7 on model deployment and prediction services.)

Seminal Papers

Bender, E. M. et al. (2021). On the dangers of stochastic parrots: Can language models be too big? FAccT 2021.
Gebru, T. et al. (2021). Datasheets for datasets. Communications of the ACM, 64(12).
Parasuraman, R. & Manzey, D. H. (2010). Complacency and bias in human use of automation. Human Factors, 52(3).
Skitka, L. J., Mosier, K., & Burdick, M. (1999). Does automation bias decision-making? International Journal of Human-Computer Studies, 51(5).
Bowman, S. R. (2023). Eight things to know about large language models. arXiv:2304.00612.
Chalmers, D. J. What we talk to when we talk to language models. https://philarchive.org/rec/CHAWWT-8

Policy and Reports

NIST AI Risk Management Framework (AI RMF 1.0), 2023.
EU AI Act — Executive Summary and Annex III (High-Risk Systems List), 2024.
AI Now Institute Annual Report (most recent year).
Partnership on AI — Guidance on Responsible Practices for Synthetic Media.

Essays and Journalism

Wallace, B. (2026, July 5). The revenge of the philosophy majors. The New York Times. https://www.nytimes.com/2026/07/05/business/philosophy-majors-ai-jobs.html

Discussion Questions You Must Prepare

Prepare exactly three to five discussion questions. They should progress in depth:

Comprehension — What did the source say? (Ensures everyone is on the same page.)
Analysis — What does this imply? What are the assumptions behind it?
Synthesis or Position — What should we do, or what do you now believe? (At least one question must be at this level.)

At least one question should connect to something specific from the course — a lab, a reading, a lecture concept — and you should call that connection out explicitly.

If you chose the Competing Texts Session format, two of your questions must be the required probing questions — one per side.

Reflection Write-Up

Submit a 300–500 word reflection within one week of your discussion date, via the course LMS. Your reflection must address:

What worked: One specific moment in the discussion that you are proud of.
What to improve: One thing you would do differently if leading this discussion again.
Course connection: The main insight from the discussion and the specific course concept (by name) it connects to or complicates.
Your position: One sentence stating what you now believe about the topic that you did not believe (or had not thought about) before preparing.

If you chose the Competing Texts Session format, your reflection must also include the closing reflection on which claims from each text survived contact with the other.

Grading

This assignment is scored on the rubric above. Points earned are added as extra credit to your final course grade. The rubric has five criteria; each criterion is worth one-fifth of the extra credit points for your slot. A student presenting in Week 1 (18 EC points) earns up to 18/5 × rubric score per criterion.

Example: Week 2 slot (16 EC points), student earns Proficient (3/4) on all five criteria → 16 × (3/4) = 12 extra credit points added to final grade.

Submission

In your submission, please include answers to any questions asked on the assignment page, as well as the questions listed below, in your README file. If you wrote code as part of this assignment, please describe your design, approach, and implementation in a separate document prepared using a word processor or typesetting program such as LaTeX. This document should include specific instructions on how to build and run your code, and a description of each code module or function that you created suitable for re-use by a colleague. In your README, please include answers to the following questions:

Describe what you did, how you did it, what challenges you encountered, and how you solved them.
Please answer any questions found throughout the narrative of this assignment.
If collaboration with a buddy was permitted, did you work with a buddy on this assignment? If so, who? If not, do you certify that this submission represents your own original work?
Please identify any and all portions of your submission that were not originally written by you (for example, code originally written by your buddy, or anything taken or adapted from a non-classroom resource). It is always OK to use your textbook and instructor notes; however, you are certifying that any portions not designated as coming from an outside person or source are your own original work.
Approximately how many hours it took you to finish this assignment (I will not judge you for this at all...I am simply using it to gauge if the assignments are too easy or hard)?
Your overall impression of the assignment. Did you love it, hate it, or were you neutral? One word answers are fine, but if you have any suggestions for the future let me know.
Using the grading specifications on this page, discuss briefly the grade you would give yourself and why. Discuss each item in the grading specification.

Any other concerns that you have. For instance, if you have a bug that you were unable to solve but you made progress, write that here. The more you articulate the problem the more partial credit you will receive (it is fine to leave this blank).

Assignment Rubric

Description Pre-Emerging (< 50%) Beginning (50%) Progressing (85%) Proficient (100%)

Reading or Event Selection — relevance and quality of the chosen source (20%) The source is tangentially related to AI or so recent it has not been analyzed in credible outlets. The source is relevant to AI but does not connect clearly to course themes or is a general news summary without depth. The source is directly relevant to a course theme (probabilistic computing, responsible AI, agent design, societal impact, etc.), comes from a credible outlet, and has enough substance to support 15 minutes of discussion. The source is timely, substantive, credibly sourced, and connects to at least two course themes. The student contextualizes it within something the class has studied.

Summary and Framing (5-minute presentation) — quality of the student-led introduction (20%) The summary reads the source aloud or is missing key context; the class does not know why the topic matters. The summary conveys the main point but does not frame a tension or open question for discussion. The 5-minute summary accurately conveys the source's main argument, identifies the relevant tension or question, and closes with a clear framing that sets up the discussion questions. The summary is concise and engaging, draws on at least one concrete example or data point from the source, and situates the topic within the course arc in a way that makes students want to engage.

Discussion Questions — quality and depth of the three to five questions (20%) Questions are yes/no, recall-only, or so broad ('Is AI good or bad?') that they cannot generate structured discussion. Questions are open-ended but do not scaffold from comprehension to analysis; most require no course knowledge to answer. Three to five questions progress from 'What did the source say?' through 'What does this imply?' to 'What should we do or think about it?', and at least one question requires course-specific knowledge to answer well. Questions are well-sequenced, one explicitly connects to a concept from course readings or labs, and the final question invites students to take and defend a position.

Discussion Facilitation — effectiveness of leading the class discussion (20%) The student reads questions aloud and waits; the discussion stalls or is dominated by one or two voices. The student follows up on some responses but does not redirect tangents or synthesize across contributions. The student draws out quieter participants, redirects off-topic contributions, connects two or more student responses to each other, and closes the discussion with a one-sentence synthesis. The student demonstrates active listening throughout, explicitly builds on prior student contributions ('Building on what X said...'), and the closing synthesis captures genuine tension or insight from the discussion, not just a restatement of the source.

Reflection Write-Up — quality of the 300–500 word post-discussion reflection (20%) The reflection is a plot summary of the source or missing. The reflection describes what happened but does not analyze why the discussion went the way it did. The reflection identifies one thing that went well, one thing to do differently, and connects the discussion's main insight to a specific concept from the course (naming the lecture, activity, or reading it connects to). The reflection is analytically sharp: it identifies a moment where the discussion revealed a genuine disagreement, explains what underlying assumption drove the disagreement, and states what the student now thinks about that assumption.

Please refer to the Style Guide for code quality examples and guidelines.

CS357

Foundations of Artificial Intelligence