These are example discussion points for you to think about before class. You are not expected to engage with all of them — pick the ones that speak most directly to your own research, and bring two or three rough answers to the in-class session. The full description of how to use these pages, including what the question tags mean, is on the Week 1 Discussion page.
Sub-lessons
Ethical Frameworks and Four Lenses
- Calibrate Pick a specific AI-in-research decision you have actually made in the last six months (e.g. using a tool to summarise papers, drafting a section with AI assistance). Read it through each of the four lenses in turn (consequentialist, deontological, virtue, ubuntu). Where do the lenses converge, and where do they pull in opposite directions?
- Apply For your own field, which of the four lenses do you think the dominant ethical norms are actually using, even if they don't name it? What do those norms miss?
- Critical The lesson explicitly flags its own limits: the instructor is not an ethicist. Which parts of the four-lenses presentation would you want a professional ethicist to push back on, and what would you want them to add?
- Connect Week 3 introduced environmental costs and the labour conditions inside the AI supply chain. Pick one of the case-studies-in-waiting from Week 3 (e.g. cobalt mining, data-centre water, the labour conditions in data-labelling pipelines) and apply the four lenses to it. Does the case look the same across the four lenses, or do they pick out genuinely different problems?
Ubuntu, Relational Ethics, and the Just AI Framework
- Calibrate “I am because we are” is a famously compact statement of ubuntu. Pick one AI tool you use regularly and ask: who is the “we” that the tool is — and is not — designed for? What does that tell you about its appropriateness for your research community?
- Apply The RIA Just AI Framework is a structured set of questions to ask of any AI system. Apply it to one tool from your current research workflow. Where does the framework surface concerns you had not previously made explicit?
- Critical Relational ethics is sometimes critiqued as “not actionable” — rich for theorising but thin for practical decisions. Steel-man that critique, then answer it from your own perspective. Is the relational frame doing more work than its critics allow?
- Connect Week 3 made the supply chain behind AI hardware visible — the mines, the labour, the energy, the people who carry the costs but never see the benefits. Read the supply-chain picture through the lens of relational personhood from this lesson. Does the picture stay the same, or does the relational framing surface something the environmental framing alone did not?
Transparency, Authorship and Integrity
- Calibrate Pick the current journal policy on AI authorship and use from a journal you actually plan to publish in. What does it formally require, and what does it leave for your judgement? Which part of that judgement-space do you find hardest?
- Apply Draft a one-paragraph AI-use disclosure that you would be willing to attach to your next paper. What does it commit you to, and what does it deliberately not say?
- Critical The lesson notes that journal policies, institutional guidelines, and legal frameworks are all in flux. Is a researcher who waits for stable policy being prudent or evasive? When does “the rules are not yet clear” stop being a defence?
- Connect Week 2 covered training-data curation as a process most users never see. This lesson asks you to make your own AI use visible to your readers. Where on the “visibility” spectrum — from completely opaque (vendor labs about training) to completely transparent (you about your draft) — do you think research will and should settle, and what is your own personal commitment within that range?
The Broader Landscape of AI Ethics
- Calibrate The page names dimensions the course has not been able to cover (labour, surveillance, military use, corporate power). Pick the one that most directly intersects your own research area — in subject matter, in funding, or in tool provenance — and articulate the connection explicitly.
- Apply If you had to add one extra week to this course covering one of these broader dimensions, which would you add and what would the syllabus look like in outline?
- Critical “Acknowledging the territory is larger than any single week can cover” is intellectually honest but can also operate as a way of deferring difficult conversations. Where do you think the right balance lies between honest acknowledgement and active engagement?
- Connect The labour-and-exploitation discussion here connects to the critical-minerals supply chain in Week 3. Are these two surface symptoms of one underlying issue (the cost of AI is paid by people you don't see) or genuinely separate problems requiring different responses?
Applying Ethics: Case Studies and Your Framework
- Calibrate Pick the case study you find most personally uncomfortable to decide. Articulate what makes it uncomfortable. Is it (a) a values clash, (b) lack of information, (c) likely social consequences for you, or (d) something else?
- Apply Build your own one-page personal AI-ethics decision rule, drawing on what made sense from the four lenses, ubuntu, and the case studies. Make it specific enough that someone reading it could predict your decision on a new case.
- Critical The lesson says the cases are “genuinely debatable.” Pick the case where you most disagree with someone else in the cohort's likely position. Where exactly is the disagreement — on the facts, the lens, or the framing?
- Connect Hold your personal framework next to the AI-use decisions you implicitly made during Weeks 1, 2, and 3 — from the moment you first opened a model to read the History of AI lesson, through the way you read the energy and supply-chain material, to whether you used AI to take notes on any of it. Were the decisions you made retrospectively consistent with the framework you would now defend?