Introductory description

This module provides a general, hands-on, and critical introduction to recent developments in the field of contemporary artificial intelligence (AI) research, specifically focusing on the use of multilayered or “deep” artificial neural networks for productive or “generative” systems which create new artifacts, such as text, images, music, and speech; examples of these proliferating interactive artifacts include ChatGPT, which deploys generative text models in a dialogical mode and has gained widespread use for applications in communication, marketing, business, and e-learning; and StableDiffusion, which can generate an expansive variety of digital art given textual descriptions. Instead of treating these systems as radically novel inventions, students will learn to place them in their historical, philosophical, political, and ethical contexts of North American, European, and now global scientific and commercial research cultures as well as in cultural imaginaries of AI in popular literature and film; and while there are no prerequisites, students will have the opportunity to experiment with these models directly in an interactive environment, and thus gain a deeper—and increasingly crucial—technical understanding of both their powers and limitations. Finally, because the impact of Generative AI is an ongoing and evolving issue, the module will engage with up-to-date news media and political/regulatory developments in order to help students develop their ability to produce novel assessments and critiques of Generative AI-related impacts and controversies as they emerge and unfold.

Module aims

Outline syllabus

This is an indicative module outline only to give an indication of the sort of topics that may be covered. Actual sessions held may differ.

• Mechanical Production and Reproduction: Introduction to the Module plus a History of Generative Art, Language, and Music to the 1950s.
• From ELIZA to AARON: A History of Generative Art, Language and Music, Part II: 1950s—2000s.
• Artificial Neural Networks from the Perceptron to the Deep Learning Era.
• Early Neurography: The ImageNet Dataset, Convolutional Neural Networks, and Creative AI.
• Signals, Music, and Text: Wavenet, Jukebox, and Generative Pre-trained Transformers (GPTs).
• Presentation and Discussion Week for Generative AI Controversies.
• Multimodal Neurography: Text-to-Image Models (DALL-E, StableDiffusion).
• Instruction Tuning, Human Feedback, and the Dialogical Transformer: Deconstructing ChatGPT.
• Multimodal Dialogical Models: Consequences and Futures for Politics, Education, and Culture.

Learning outcomes

By the end of the module, students should be able to:

• Demonstrate in-depth knowledge of the conceptual foundations, cultural background, and historical impacts of generative systems for various forms of image/text/signal media
• Demonstrate a critical understanding of the assumptions and present-day limitations of contemporary generative artificial intelligence (AI) models
• Demonstrate a practical ability to interact with these systems in a hands-on and exploratory manner to help answer questions about generative AI
• Demonstrate an appreciation and deeper understanding of the social, ethical, cultural, and philosophical implications and impacts of recent developments in generative AI
• Demonstrate skill in analysis and interpretation of various media content around subjects in generative AI

Indicative reading list

Agüera y Arcas, B. (2022). Do Large Language Models Understand Us? Daedalus.

Bakhtin, M. M. (1982). The Dialogic Imagination: Four Essays (M. Holquist, Ed.; C. Emerson, Trans.). University of Texas Press.

Barthes, R. (1977). The Death of the Author. In Image, Music, Text (pp. 142–148). Fontana.

Bender, E. M., Gebru, T. et al. (2021). On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, 610–623.

Benjamin, W. (1968). The Work of Art in the Age of Mechanical Reproduction. In Illuminations: Essays and Reflections. Schocken.

Boden, M. A. (1998). Creativity and artificial intelligence. Artificial Intelligence, 103(1), 347–356.

Bourdieu, P. (1996). The Rules of Art: Genesis and Structure of the Literary Field (New Ed edition). Polity Press.

Brown, T. B. et al. (2020). Language Models are Few-Shot Learners. ArXiv:2005.14165 [Cs].

Castelle, M. (2020). The social lives of generative adversarial networks. Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, 413.

Cohen, H. (1974). On Purpose: An enquiry into the possible roles of the computer in art. Studio International, 187(962), 11–16.

Derrida, J. (1971). Signature Event Context. In Limited Inc. (pp. 1–24). Northwestern University Press.

Foster, D. (2023). Generative Deep Learning: Teaching Machines to Paint, Write, Compose, and Play (2nd ed.). O’Reilly Media.

Gatys, L. A., Ecker, A. S., & Bethge, M. (2015). A Neural Algorithm of Artistic Style. ArXiv:1508.06576 [Cs, q-Bio].

Globus, G. G. (1995). The Postmodern Brain. John Benjamins Publishing.

Goodfellow, I. et al. (2014). Generative Adversarial Networks. ArXiv:1406.2661 [Cs, Stat].

Jacobsen, B. N. (2023). Machine learning and the politics of synthetic data. Big Data & Society, 10(1).

Luccioni, A. S. et al. (2023). Stable Bias: Analyzing Societal Representations in Diffusion Models (arXiv:2303.11408).

McQuillan, D. (2022). Resisting AI: An Anti-fascist Approach to Artificial Intelligence. Bristol University Press.

Miller, A. I. (2019). The artist in the machine: The world of AI-powered creativity. Mit Press.

OpenAI. (2023). GPT-4 Technical Report (arXiv:2303.08774). arXiv. https://doi.org/10.48550/arXiv.2303.08774

Ouyang, L. et al. (2022). Training language models to follow instructions with human feedback (arXiv:2203.02155).

Park, J. S., O’Brien, J. C., Cai, C. J., Morris, M. R., Liang, P., & Bernstein, M. S. (2023). Generative Agents: Interactive Simulacra of Human Behavior (arXiv:2304.03442).

Solaiman, I. (2023). The Gradient of Generative AI Release: Methods and Considerations (arXiv:2302.04844).

Stark, L., & Crawford, K. (2019). The Work of Art in the Age of Artificial Intelligence: What Artists Can Teach Us About the Ethics of Data Practice. Surveillance & Society, 17(3/4), 442–455.

Tunstall, L., Werra, L. von, & Wolf, T. (2022). Natural Language Processing with Transformers: Building Language Applications with Hugging Face. O’Reilly Media.

Weizenbaum, J. (1966). ELIZA—a computer program for the study of natural language communication between man and machine. Communications of the ACM, 9(1), 36–45.

Wiezenbaum, J. (1976). Computer Power and Human Reason: From Judgment to Calculation. W.H. Freeman and Company.

Research element

• Exhibiting the capacity for critical thinking, methodological reflection and analysis;
• Being able to apply critical research and enquiry skills;
• Taking responsibility for once's own learning and development;
• Effectively use research tools and resources

Interdisciplinary

• To appreciate the value of understanding different disciplinary approaches and perspectives;
• Leverage a confidence and competence in interdisciplinarity specifically in relation to understanding the variety of ways that Generative AI and societies interact at various political, philosophical, and ethical levels;
• Communicating ideas effectively in different ways and to people with different disciplinary backgrounds.

International

• To appreciate the local, regional, national, and global issues that permeate the topic of Generative AI;
• To appreciate the global reach that technological developments in Generative AI have in different parts of different societies.

Subject specific skills

• Methodological skills in analysing and interpreting media controversies on the topic of Generative AI;
• Practical skills in using, interacting, and experimenting with Generative AI tools and software directly;
• Conceptual and methodological skills in combining different disciplinary approaches to Generative AI in contemporary society;
• Conceptual skills in combining different different disciplinary approaches and perspectives on Generative AI in society in relation to their subject specialism.

Transferable skills

• The ability to communicate with peers and academics within and outside of one’s primary discipline;
• The practical and technical ability to use and critically engage with new forms of software and digital resources;
• Collaborative skills to work in mixed-competency teams to complete methodological and practical exercises;
• Academic writing skills and the ability to account for individual and group-based analytical processes;
• Skills to articulate arguments orally, in writing and visually, supported by reading and research;
• Practical skills to manage time to meet deadlines as an individual and team member;
• Analytical skills to connect theoretical ideas with methodological and practical applications;
• Ability to solve problems creatively, specifically as these relate to the benefits and limitations of Generative AI;
• Being a self-motivated and independent learner.