Kristina Šekrst

Introduction to AI (CogSci syllabus, 2026/2027)

Introduction to Artificial Intelligence

A fifteen-unit introductory course in artificial intelligence for cognitive science students, covering the historical and conceptual foundations of the field, the symbolic, connectionist, and probabilistic paradigms, machine learning and deep learning, transformers and large language models, interpretability, ethics, and artificial consciousness.

Syllabus

  1. Artificial intelligence, mind, and consciousness
    Artificial intelligence as an engineering discipline and as a model of cognition. Searle’s Chinese Room argument: syntax and semantics, weak and strong artificial intelligence, and intentionality. The Turing Test and the Imitation Game. The attribution of mental states to artificial systems.
  2. Logic and complexity
    Propositional and predicate logic: syntax, semantics, and inference. Knowledge representation through symbols, rules, and logical inference. Soundness, completeness, and decidability; the Church-Turing thesis and the notion of computation. Computational complexity and the tractability of inference: P and NP classes, satisfiability, NP-hard and NP-complete problems. AI-completeness.
  3. Cybernetics and information theory
    Cybernetics as the study of control and communication. Feedback, homeostasis, adaptation, and goal-directed behavior. Turing, Wiener, and Ashby. The relationship between organism, machine, and environment. Information theory: Shannon’s theorems, entropy, information content and coding.
  4. The symbolic paradigm: thought as computation over representations
    The physical symbol system hypothesis as an empirical claim about minds. Problem solving as state-space search: means-ends analysis and the General Problem Solver; combinatorial explosion as complexity in practice. Expert systems and their brittleness. The frame problem, the qualification problem, and common-sense knowledge. The Chinese Room revisited: what symbol manipulation can and cannot explain.
  5. The connectionist paradigm: thought as emergent from networks
    The formal neuron and the McCulloch-Pitts model; Rosenblatt’s perceptron and learning from examples. Linear separability, logic, and the XOR problem. Multilayer perceptrons and the idea of backpropagation. What connectionism claims about the mind that symbolic AI denies.
  6. The probabilistic paradigm: thought as inference under uncertainty
    Probability and Bayes’ rule; generative models and probabilistic inference. Perception as unconscious inference, from Helmholtz to Bayesian models of perception, categorization, and causal learning. NP-hardness in belief networks. The probabilistic turn as the road to machine learning.
  7. Machine learning: supervised and unsupervised
    Features, labels, datasets, models, objective functions, and optimization. Supervised learning, classification, and regression. Unsupervised learning, clustering, dimensionality reduction, and latent structure. Training, validation, and test sets. Generalization, overfitting, underfitting, and model evaluation. Inductive bias.
  8. Reinforcement learning and decision-making
    Agents, environments, states, actions, rewards, and policies. Markov decision processes. Value functions and expected reward. Exploration and exploitation. Q-learning and temporal-difference learning. Reward models. Goal-directed behavior in humans and other animals.
  9. Deep neural networks
    The development of modern deep learning. Deep feedforward networks, optimization, and hierarchical feature learning. Convolutional neural networks and computer vision. Recurrent neural networks, sequence processing, and long-range dependencies. Autoencoders.
  10. Transformers: architecture and foundations
    Linear algebra basics. From word2vec to embeddings. Tokenization and positional encoding. The attention mechanism. Residual connections and layer normalization. Encoder and decoder architectures.
  11. Large language models and generative AI
    Pretraining: next-token prediction as self-supervised learning; scaling. Fine-tuning and reinforcement learning from human feedback. In-context learning and prompt engineering (zero-shot and few-shot). Hallucinations, grounding, and chain-of-thought reasoning. Prompt injections and adversarial prompting.
  12. Retrieval, memory, and agentic AI
    Retrieval-augmented generation and vector databases as external memory. Cognitive offloading and the extended mind. AI agents: planning, tool use, and self-monitoring. Agentic systems design case studies. Agents as cognitive architectures: production systems revisited.
  13. Explainability, interpretability, and machine reasoning
    Mechanistic interpretability: features, circuits, and probing. Explainable AI and what counts as an explanation of a model’s behavior. Chain-of-thought reasoning and the question of faithfulness.
  14. AI ethics and the social role of AI
    Algorithmic bias and fairness. Transparency, accountability, and the demand for explanation in high-stakes decisions. Alignment: specifying and verifying what AI systems should do. Privacy, labor, misinformation, and the concentration of AI capabilities.
  15. Artificial consciousness and final remarks
    Artificial consciousness, the attribution of mental states, and the moral status of artificial systems: the Turing Test and the Chinese Room revisited.

Required reading

  1. Russell, S., & Norvig, P. (2021). Artificial Intelligence: A Modern Approach (4th ed.). Pearson. (selected chapters)
  2. Boden, M. A. (2016). AI: Its Nature and Future. Oxford University Press. (selected chapters)
  3. Šekrst, K. (2025). The Illusion Engine: The Quest for Machine Consciousness. Springer. (selected chapters) · DOI
  4. Selected chapters of weekly readings

Supplementary reading

  1. Turing, A. M. (1950). Computing machinery and intelligence. Mind, 59(236), 433–460. · DOI
  2. Searle, J. R. (1980). Minds, brains, and programs. Behavioral and Brain Sciences, 3(3), 417–424. · DOI
  3. Aaronson, S. (2013). Why philosophers should care about computational complexity. In B. J. Copeland et al. (Eds.), Computability: Turing, Gödel, Church, and Beyond. MIT Press. · PDF
  4. van Rooij, I. (2008). The tractable cognition thesis. Cognitive Science, 32(6), 939–984. · DOI
  5. Wiener, N. (1948). Cybernetics: Or Control and Communication in the Animal and the Machine. MIT Press. · DOI
  6. Shannon, C. E. (1948). A mathematical theory of communication. Bell System Technical Journal, 27, 379–423. · DOI
  7. Newell, A., & Simon, H. A. (1976). Computer science as empirical inquiry: Symbols and search. Communications of the ACM, 19(3), 113–126. · DOI
  8. Dennett, D. C. (1984). Cognitive wheels: The frame problem of AI. In C. Hookway (Ed.), Minds, Machines and Evolution. Cambridge University Press.
  9. McCarthy, J., & Hayes, P. J. (1969). Some philosophical problems from the standpoint of artificial intelligence. In B. L. Webber & N. J. Nilsson (Eds.), Readings in Artificial Intelligence (pp. 431–450). Morgan Kaufmann. · DOI
  10. McCulloch, W. S., & Pitts, W. (1943). A logical calculus of the ideas immanent in nervous activity. Bulletin of Mathematical Biophysics, 5, 115–133. · DOI
  11. Rumelhart, D.; Hinton, G.; Williams, R. (1986). Learning representations by back-propagating errors. Nature, 323, 533–536. · DOI
  12. Fodor, J. A., & Pylyshyn, Z. W. (1988). Connectionism and cognitive architecture: A critical analysis. Cognition, 28(1–2), 3–71. · DOI
  13. Tenenbaum, J. B.; Kemp, C.; Griffiths, T. L.; Goodman, N. D. (2011). How to grow a mind: Statistics, structure, and abstraction. Science, 331(6022), 1279–1285. · DOI
  14. Chater, N.; Tenenbaum, J. B.; Yuille, A. (2006). Probabilistic models of cognition: Conceptual foundations. Trends in Cognitive Sciences, 10(7), 287–291. · DOI
  15. Jordan, M. I., & Mitchell, T. M. (2015). Machine learning: Trends, perspectives, and prospects. Science, 349(6245), 255–260. · DOI
  16. Sutton, R. S., & Barto, A. G. (2018). Reinforcement Learning: An Introduction (2nd ed.). MIT Press.
  17. LeCun, Y.; Bengio, Y.; Hinton, G. (2015). Deep learning. Nature, 521(7553), 436–444. · DOI
  18. Skansi, S., & Šekrst, K. (2026). Introduction to Deep Learning: Neural Networks, Large Language Models and Agentic AI. Springer. · Springer
  19. Vaswani, A., et al. (2017). Attention is all you need. NIPS’17: Proceedings of the 31st International Conference on Neural Information Processing Systems, 6000–6010. · ACM
  20. Botvinick, M., et al. (2019). Reinforcement learning, fast and slow. Trends in Cognitive Sciences, 23(5), 408–422. · DOI
  21. Brown, T. B., et al. (2020). Language models are few-shot learners. NIPS’20: Proceedings of the 34th International Conference on Neural Information Processing Systems, 1877–1901. · ACM
  22. Lewis, P., et al. (2020). Retrieval-augmented generation for knowledge-intensive NLP tasks. NIPS’20: Proceedings of the 34th International Conference on Neural Information Processing Systems, 9459–9474. · ACM
  23. Wei, J., et al. (2022). Chain-of-thought prompting elicits reasoning in large language models. NIPS’22: Proceedings of the 36th International Conference on Neural Information Processing Systems, 24824–24837. · ACM
  24. Longo, L., et al. (2024). Explainable Artificial Intelligence (XAI) 2.0: A manifesto of open challenges and interdisciplinary research directions. Information Fusion, 106, 102301. · DOI
  25. Lindsey, J., et al. (2025). On the biology of a large language model. Transformer Circuits Thread. · Full text
  26. Chalmers, D. J. (2023). Could a large language model be conscious? Boston Review. · Full text
  27. Müller, V. C. (2026). Ethics of artificial intelligence and robotics. In E. N. Zalta & U. Nodelman (Eds.), The Stanford Encyclopedia of Philosophy (Summer 2026 ed.). · Full text
  28. Seth, A. K. (2025). Conscious artificial intelligence and biological naturalism. Behavioral and Brain Sciences, 1–42. · DOI

Readings by course unit

  1. Artificial intelligence, mind, and consciousness
    Turing (1950); Searle (1980); additionally: Šekrst (2025): chapters 5 and 6
  2. Logic and complexity
    Aaronson (2013): sections 1–4; van Rooij (2008): sections 1–3; Šekrst (2025): sections 7.1–7.4
  3. Cybernetics and information theory
    Wiener (1948): Introduction; Shannon (1948): Introduction; additionally: Šekrst (2025): chapter 4
  4. The symbolic paradigm: thought as computation over representations
    Newell & Simon (1976); McCarthy & Hayes (1969): sections 1, 2.1, and 4.3 or Dennett (1984)
  5. The connectionist paradigm: thought as emergent from networks
    McCulloch & Pitts (1943): sections 1 and assumptions in section 2; Rumelhart et al. (1986); Fodor & Pylyshyn (1988): Part I and II; Šekrst (2025): section 8.1
  6. The probabilistic paradigm: thought as inference under uncertainty
    Chater, Tenenbaum, & Yuille (2006); Tenenbaum, Kemp, Griffiths, & Goodman (2011)
  7. Machine learning: supervised and unsupervised
    Jordan & Mitchell (2015); Šekrst (2025): chapter 8
  8. Reinforcement learning and decision-making
    Sutton & Barto (2018): Introduction; Botvinick et al. (2019)
  9. Deep neural networks
    LeCun et al. (2015); Šekrst (2025): chapter 9
  10. Transformers: architecture and foundations
    Vaswani et al. (2017); Šekrst (2025): chapter 10
  11. Large language models and generative AI
    Skansi & Šekrst (2026): sections 14.1–14.5.2; Šekrst (2025): chapter 12; Brown et al. (2020): Introduction
  12. Retrieval, memory, and agentic AI
    Skansi & Šekrst (2026): sections 15.1–15.3 (without code) or Lewis et al. (2020): Introduction and Methods
  13. Explainability, interpretability, and machine reasoning
    Šekrst (2025): chapter 14 or the following: Longo et al. (2024): sections 1 and 3, and Wei et al. (2022): sections 1, 2, 3.2; Lindsey et al. (2025): Introductory Example, Planning in Poems, Chain-of-Thought Faithfulness, Uncovering Hidden Goals in a Misaligned Model
  14. AI ethics and the social role of AI
    Müller (2026): Main Debates; Šekrst (2025): chapter 16
  15. Artificial consciousness and final remarks
    Chalmers (2023); Šekrst (2025): chapter 13 or Seth (2025): sections 1, 2, and 5

Learning outcomes

On successful completion of the course, students will be able to:

  1. Explain the principal historical and conceptual approaches to artificial intelligence.
  2. Describe the formal foundations of artificial intelligence, including logical inference, computational complexity, and information theory, and explain how tractability constrains theories of cognition.
  3. Compare the symbolic, connectionist, and probabilistic paradigms in terms of their central claims about the mind, their characteristic achievements, and their characteristic failures.
  4. Explain the main methods of machine learning, reinforcement learning, and deep learning, and identify the problems each is suited to solve.
  5. Describe the architecture of transformer models and explain how large language models are trained, adapted, and deployed in generative and agentic systems.
  6. Evaluate empirical claims about the cognitive capacities of artificial systems, and assess the methods used to support such claims.
  7. Analyze the interpretability of artificial systems and evaluate what would count as an explanation of a model’s behavior.
  8. Discuss the ethical and social consequences of artificial intelligence, and critically assess arguments concerning artificial consciousness and the moral status of artificial systems.