Dissertation & Viva Preparation Lab
A source-preserving, cumulative polishing and reassessment space for the 2024–25 MA AI, Ethics and Society programme, its second-year dissertation and future doctoral research. This is independent preparation, not an officially awarded mark, accredited assessment, university-approved viva or a declaration that previously undertaken work is incomplete.
Original syllabus preserved conceptuallyAll 10 listed options10 teaching meetings33 diagnostic questions11 assessment domainsDissertation / PhD pathways
1. Approved 2024–25 two-year structure — not a generic MA
Year 1 — core and permitted options
The Ethics of Technology and Artificial Intelligence (SC03017S7, 30 credits, Term 1).
History and Foundations of AI (SC03018S7, 30 credits, Term 2).
May include options from the approved list; the programme specifies options in year 1 not exceeding 45 credits.
Year 2 — dissertation-led
Dissertation, MA AI, Ethics and Society (SC03019D7, 60 credits, core).
Complete the remaining approved option credits toward 180 total credits; 60 optional credits in total across the two years.
The dissertation is substantive, independent research with a well-argued thesis. The programme does not explicitly prescribe a compulsory oral viva in the supplied specification; this page's viva exercises are additional preparation.
Note: programme variants include October and January entry. The original document lists differences in teaching sequence. The specification states no more than 45 option credits in one term; Analytic Tools for Data Science needs strong Python background. The item Critical and Creative AI appears in full-time listings but not in the part-time option tables; this is recorded as a listing difference requiring confirmation, not silently assumed part-time availability.
Full programme option register — no omitted option from full-time listing
| Module | Named option (original wording or variant) | Credits | Status |
|---|---|---|---|
| SC03020S7 | Philosophy of Artificial Intelligence: Minds, Machines and Meaning | 30 | Optional in the approved programme; delivery/availability to be checked |
| ARMC247S7 | Managing Digital Media | 30 | Optional in the approved programme; delivery/availability to be checked |
| LALW059S7 | Algorithms and the Law | 30 | Optional in the approved programme; delivery/availability to be checked |
| BUCI081H7 | Demystifying Computing with Python | 15 | Optional in the approved programme; delivery/availability to be checked |
| BUCI080H7 | Analytic Tools for Data Science | 15 | Optional in the approved programme; delivery/availability to be checked |
| BUMN154H7 | Managing Emerging Digital Technologies / Management of Emerging tech | 15 | Optional in the approved programme; delivery/availability to be checked |
| BUMN176H7 | Introduction to Analytics and Business | 15 | Optional in the approved programme; delivery/availability to be checked |
| BUMN178H7 | Big Data Management | 15 | Optional in the approved programme; delivery/availability to be checked |
| BUMN179H7 | Data Visualization and Communication | 15 | Optional in the approved programme; delivery/availability to be checked |
| SC04023S7 | Critical and Creative AI | 30 | Optional in the approved programme; delivery/availability to be checked |
2. All nine official programme learning outcomes
- Identify and explain core AI principles, technical underpinnings and societal impacts.
- Evaluate ethical and societal implications across technology, law, policy, philosophy and business.
- Analyse implementation case studies and propose viable solutions.
- Apply ethical theories and frameworks to real AI cases.
- Develop and critique AI policy and regulation using philosophical, ethical and technical knowledge.
- Demonstrate critical thinking in clear, coherent debate.
- Communicate AI ethics to policymakers, technologists and the general public.
- Conduct independent research culminating in a well-argued thesis.
- Collaborate across academic and professional sectors on AI governance.
These outcomes are paraphrased for navigation; the programme specification remains the authoritative wording. Assessment examples include written exercises, essays, quizzes, case studies and problems. The history module uses short case-study answers with criteria including focused answers, structure, conceptual clarity, analysis, literature and evidence.
3. Complete meeting-level reading register from the supplied history syllabus
Primary and supplementary readings remain distinct. This listing adds a searchable preparation layer without modifying the original academic files. Pre-reading suggestions: Melanie Mitchell (2019), AI: A Guide for Thinking Humans; Margaret Boden (2016), AI: Its Nature and Its Future; and Cave, Dihal and Dillon (eds., 2020), AI Narratives.
Meeting 1: Thinking Machines — 2 primary, 2 supplementary
Primary readings
- Turing (1950) — Computing Machinery and Intelligence
- Searle (1980) — Minds, Brains, and Programs
Additional readings
- Minsky (1961) — Steps Towards Artificial Intelligence
- McCarthy (1959) — Programs with Common Sense
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
Meeting 2: Meaning of Intelligence: Signal Processing vs Grounded Cognition — 2 primary, 2 supplementary
Primary readings
- Halpern (2015) — Beautiful Data, Prologue/Introduction pp. 1–38
- Burrow (2023) — Algorithmic Fanboy
Additional readings
- Barsalou (2008) — Grounded Cognition
- Wiener (1961 edition) — Cybernetics, Chapter IX pp. 169–180
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
Meeting 3: Artificial Advisors: Expert Systems to Today — 2 primary, 1 supplementary
Primary readings
- Maeda and Quan-Haase (2024) — Parasocial Human–AI Interactions
- Shah et al. (2016) — Can Machines Talk? ELIZA and modern dialogue
Additional readings
- Fogg (2003) — Persuasive Technology, Chapters 3 and 5
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
Meeting 4: Taming Uncertainty: Probabilities and Prediction — 2 primary, 2 supplementary
Primary readings
- Jiang, Gradus and Rossellini (2020) — Supervised Machine Learning Primer
- Rescher (1998) — Predicting the Future, Chapters 7–8
Additional readings
- Lee and Trott (2023) — How Large Language Models Work
- Margetts, Dorobantu and Bright (2024) — Public Services with Data Science and AI
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
Meeting 5: Is Data Still the New Oil? — 2 primary, 2 supplementary
Primary readings
- Crawford (2021) — Atlas of AI, Chapter 3: Data
- Muldoon et al. (2024) — Typology of AI Data Work
Additional readings
- Penman et al. (2024) — ImageNet Classifier Values
- Srnicek (2022) — Data, Compute, Labor
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
Meeting 6: Feedback Loops and Model Collapse — 2 primary, 1 supplementary
Primary readings
- Shumailov et al. (2024) — The Curse of Recursion
- Wyllie, Shumailov and Papernot (2024) — Fairness Feedback Loops
Additional readings
- Pagan et al. (2023) — Classification of Feedback Loops
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
Meeting 7: AI, Power, and the State — 2 primary, 2 supplementary
Primary readings
- Eubanks (2018) — Automating Inequality, Chapter 3
- Benjamin (2019) — Race After Technology, Chapter 3
Additional readings
- Winner (1980) — Do Artifacts Have Politics?
- Angwin et al. (2016) — Machine Bias
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
Meeting 8: Fixing Unfairness: Technical Fixes or Counter-Imaginaries? — 2 primary, 2 supplementary
Primary readings
- Kazansky and Milan (2021) — Bodies Not Templates
- Ferrara (2023) — Fairness and Bias in AI
Additional readings
- Binns (2018) — Fairness in ML: Political Philosophy
- Huang et al. (2024) — Collective Constitutional AI
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
Meeting 9: AI Design, Explainability and Accountability — 2 primary, 2 supplementary
Primary readings
- Davis (2023) — Affordances for Machine Learning
- EDPS (2023) — TechDispatch: Explainable AI
Additional readings
- Reed (2023) — Accountability for LAWS in War
- Costanza-Chock (2020) — Design Justice, Introduction
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
Meeting 10: Materiality of AI and Its Consequences — 2 primary, 2 supplementary
Primary readings
- Crawford (2021) — Atlas of AI, Chapter 1: Earth
- Hogan and Lepage-Richer (2024) — Climate Tech
Additional readings
- Birhane and van Dijk (2020) — Robot Rights? Human Welfare Instead
- Muldoon and Wu (2023) — AI in the Colonial Matrix of Power
Source: supplied History and Foundations of AI reading syllabus. Verify complete page numbers and supplied links in the original PDF.
4. Wider knowledge coverage — the polishing inventory
Source hierarchy: (A) course-confirmed core; (B) programme-listed options; (C) dissertation/viva skills; (D) interdisciplinary extensions and updated 2026 context. An extension is never relabelled as a compulsory programme requirement.
1. Historical chronology
Dates, technical milestones, paradigm changes and causes of AI winters
MA foundational / research preparation
2. Computing and AI
Symbolic AI, cybernetics, ML, neural networks, transformers, tools and evaluation
MA foundational / research preparation
3. Philosophy of mind
Turing, Searle, functionalism, grounding, consciousness and intentionality
MA foundational / research preparation
4. Ethics and fairness
Consequentialism, duties, care, bias, fairness metrics and remedies
MA foundational / research preparation
5. Politics and institutions
State power, administrative decisions, rights, regulation and public accountability
MA foundational / research preparation
6. Statistics and evidence
Probability, calibration, causal inference, selection bias and valid measurement
MA foundational / research preparation
7. Literature and terminology
Precise attributions, named readings, conceptual distinctions and terminology
MA foundational / research preparation
8. Research methodology
Research design, construct validity, falsification, evidence and limitations
MA foundational / research preparation
9. Dissertation and viva
Thesis, contribution, originality, methodology, counterarguments and limitations
Dissertation and voluntary viva rehearsal
10. 2026 landscape
Current AI use, assessment policy, employment, governance and institutional change
2026 evidence extension
11. Optional specialisms
Python, analytics, data management, media, legal and creative AI routes
Approved optional pathways and wider specialisations
Extended checklist of constructs, theories, methods, dates and principles
Historical & conceptual
Logic; Turing (1950); Dartmouth (1956); cybernetics; connectionism; symbolic AI; perceptrons; expert systems; AI winters; statistical ML; deep learning; 2012 ImageNet era; transformers (2017); generative models; public-facing chatbots (2022 onward); 2026 agentic and multimodal evaluation.
Computing & AI
Algorithms, data structures, programming, software lifecycle, version control, hardware, GPU/compute, networks, APIs, cloud, cybersecurity, supervised/unsupervised/reinforcement learning, loss, gradient descent, tokenisation, embeddings, transformer attention, fine-tuning, RAG, tool use, agent orchestration, benchmarks, safety and robustness.
Philosophy
Philosophy of mind, dualism, materialism, functionalism, intentionality, the Chinese Room, embodied/grounded cognition, epistemology, empiricism, rationalism, testimony, social epistemology, consequentialism, deontology, virtue, care, contractualism, justice, scientific realism, falsification.
Political and economic constructs
Institutions, governance, democracy, rule of law, accountability, constitutional arrangements, rights, welfare, public services, bureaucracy, procedural fairness, distributive justice, labour markets, market power, human capital, signalling, inequality, externalities, technological concentration and digital sovereignty.
Statistics & research
Descriptive statistics, distributions, conditional probability, Bayes, sampling, uncertainty, calibration, precision, recall, F1, confusion matrices, regression, confounding, causal inference, validity, reliability, construct measurement, triangulation, reproducibility, sensitivity, adversarial testing, ethics review, attribution and provenance.
5. Interactive knowledge diagnostic — 33 questions
Complete the questions, then choose Score my answers. Results show percentage correct within each domain and an overall percentage. These are not statistical percentile ranks compared with other students: a genuine percentile requires a representative norm group, which is unavailable. Results are local to the page; no answers are submitted to a server.
0 of 33 answered
6. Dissertation and viva rehearsal — explain, defend, revise
These are optional doctoral-style oral defence exercises, not an asserted compulsory MA viva. For each prompt, first give an unaided two-minute explanation; then check citations, counterarguments and possible disconfirmation. You may type a private practice response below and copy it for your own records.
- State the precise research question in one sentence. What is explicitly outside its scope?
- What existing results make the research necessary, and what exact limitation remains unresolved?
- What is the thesis contribution: conceptual, empirical, methodological, normative or technical?
- Why is the chosen method appropriate? Which alternative design would challenge your conclusions?
- What are your operational definitions for bias, personhood, delegation, values and accountability?
- Which evidence would falsify your preferred explanation?
- What is the strongest objection from moral philosophy or political theory?
- How do you distinguish correlation, mechanism, causality and normative judgment?
- What does the literature already establish, and how does your work differ?
- How have you addressed sampling, construct validity, measurement error and reproducibility?
- What are the governance, privacy, fairness and participant-risk implications?
- What remains uncertain; which claims cannot yet be made?
- How would you defend your conclusions without access to an AI assistant?
- How would the study be extended to PhD scale without merely restating the MA dissertation?
Six-part self-validation rubric
| Criterion | Self-test |
|---|---|
| Recall and date precision | Can I give the correct year, name and nature of the event, and distinguish online preprint from peer-reviewed publication? |
| Definition accuracy | Can I define the term and distinguish it from adjacent concepts without prompts? |
| Argument and counterargument | Can I reconstruct the strongest opposing position rather than a weak version? |
| Evidence calibration | Can I identify what a statistic measures, its denominator, uncertainty and limitations? |
| Application | Can I apply the theory to an unfamiliar real-world case? |
| Original contribution | Can I identify the unresolved limitation and design a test that could overturn my conclusion? |
7. From existing MA dissertation work to a doctoral research portfolio
Existing work acknowledged: the supplied working dissertation proposal includes Ethical Personhood, Bias, and Systemic Moral Accountability in AI and the working title From Moral Systems to Moral Agents: Exploring Bias, Ethical Delegation, and Value-Embedding in Socio-technical AI Infrastructures. The research was already developed in a draft; this lab is a polishing and extension environment, not a replacement for that work.
| Doctoral thesis family | Possible refinement (research question, not a claimed discovery) | Evidence to seek |
|---|---|---|
| Normative philosophy | Under what conditions can responsibility for AI-mediated decisions be meaningfully assigned? | Compare responsibility theories, institutional practices and counterexamples. |
| Socio-technical empirical study | Which institutional decision pathways create measurable moral-accountability gaps? | Audit trails, documented decisions, interviews and controlled case comparisons. |
| AI safety and technical evaluation | Can system-level provenance and contradiction tests reveal undetected governance failures? | Open benchmarks, test harness, ablations, failure rates and independent replication. |
| Critical policy and law | Which regulatory obligations correspond to demonstrable remedies when AI systems harm public access? | Statutory/regulatory mapping, case law, enforcement and redress outcomes. |
| Mixed-methods / interdisciplinary | Can philosophical accountability definitions be operationalised in an empirical AI governance study? | Construct definitions, inter-rater reliability, quantitative data and qualitative explanation. |
| Education and competence recognition | Do new AI-era assessments measure capability better than prior methods without unjustified exclusion? | Assessment validity, alternative pathways, differential outcomes and reasonable adjustments. |
Gap-first protocol: (1) precise proposition; (2) strongest related work and first-publication dates; (3) established result; (4) demonstrable limitation; (5) falsifiable hypothesis; (6) study design; (7) measurable contribution; (8) limitations and rival explanations; (9) publication and provenance trail.
8. Source integrity and current extensions
University originals retained as authoritative: 2024–25 MA AI, Ethics and Society programme specification; History and Foundations of AI syllabus and complete primary/additional reading list; 2024–25 assessment and Level 7 marking criteria; user's existing dissertation proposal. The source files are not modified, overwritten or represented as official new programme documents. Cross-disciplinary specifications (PPE, statistics, mathematics, management, media, healthcare and sport) are supporting references, not required AI modules.
2026 additions: OfS / Advance HE (11 Nov 2026) · OfS AI assessment guidance · QAA Characteristics Statements · Institutional bottleneck research flag.
Boundary: This page preserves every option named in the provided programme specification's full-time lists and all ten meetings in the history reading document, but it does not pretend to contain module handbooks or detailed reading lists for unprovided options or the entire ethics module. Further original materials can be appended without deleting the records here. Its quiz is a formative self-check of selected themes—not exhaustive certification of programme mastery.