The course will address the main approaches to artificial intelligence research, starting with the field’s more formal foundations (logic and search), and ascending ever-higher levels of complexity (reasoning and decision-making under uncertainty) before reaching more modern techniques in the area of machine learning. We will also pay particular attention to the links between the studies of artificial and biological intelligence – to what those links are, and what they could or should be. Most weeks will deal with a different topic, though adjacent weeks’ topics are typically related. Each week will begin with a lecture that surveys the most important features of the field, then follow on with a seminar-style session in which students will be able to discuss selected aspects of the field in more detail.

1. Identify key characteristics of the major sub-fields of artificial intelligence, including their preferred methods and the types of intelligent behavior that they most directly address.
2. Understand the convergences and divergences between different perspectives on research in artificial intelligence: demonstrate cross-disciplinary synthesis in critical analysis of the field.
3. Identify key methodological and conceptual weaknesses in current research on artificial intelligence, including how certain techniques address other techniques’ weaknesses, while introducing new limitations of their own.
4. Demonstrate information literacy in self-directed study: e.g., in the ability to identify your own references in a specified area, or to support a specified point of view.
5. Communicate complex, multi-disciplinary material coherently and concisely, both in oral and in written form.
6. Understand the most pressing, current challenges in the field (both technical and socio-political), and some modern attempts to meet those challenges.

Extra readings (to be introduced in the second session of each week, if time allows):

·         WEEK 1: Newitz (2022). The curious case of the AI and the lawyer. New Scientist, 255 (3396), page 28

·         WEEK 2: Ueon et al., (2012) Highly scalable graph search for the Graph500 benchmark | Proceedings of the 21st international symposium on High-Performance Parallel and Distributed Computing (acm.org)

·         WEEK 3: Liu et al., (2017). Planning Dynamically Feasible Trajectories for Quadrotors Using Safe Flight Corridors in 3-D Complex Environments | IEEE Journals & Magazine | IEEE Xplore

·         WEEK 4: Tzung-Pei et al. (2001). Adversarial Search by Evolutionary Computation | MIT Press Journals & Magazine | IEEE Xplore

·         WEEK 5: Pavlov et al. (2013). Exploring Automated Reasoning in First-Order Logic: Tools, Techniques and Application Areas | SpringerLink

·         WEEK 6: Fox et al. (2020). OpenClinical.net: Artificial intelligence and knowledge engineering at the point of care - PMC (nih.gov)

·         WEEK 7: Bonet & Gefner. (2001). Planning and Control in Artificial Intelligence: A Unifying Perspective (springer.com)

·         WEEK 9: Biederman & Taroni. (2006). Bayesian networks and probabilistic reasoning about scientific evidence when there is a lack of data - ScienceDirect

·         WEEK 10: Constantinou. (2018). Bayesian-Artificial-Intelligence-for-Decision-Making-under-Uncertainty.pdf (researchgate.net). N.B. This is a research proposal rather than a paper.

·         WEEK 11: Hope, T.M.H. (2020). Linear regression. In Mechelli A. & Vieira, S. (eds) Machine learning: methods and applications to brain disorders. Elsevier.

·         WEEK 12: Botvinick et al. (2020). Deep Reinforcement Learning and Its Neuroscientific Implications - ScienceDirect

·         WEEK 13: Roohani et al. (2018). (18) (PDF) Predicting Language Recovery after Stroke with Convolutional Networks on Stitched MRI (researchgate.net). NeurIPS Workshop on Machine Learning for Health (ML4H).

·         WEEK 14: Schrimpf, M., Blank, I. A., Tuckute, G., Kauf, C., Hosseini, E. A., Kanwisher, N., ... & Fedorenko, E. (2021). The neural architecture of language: Integrative modeling converges on predictive processing. Proceedings of the National Academy of Sciences, 118(45).