This module gave a multidisciplinary overview of intelligent behaviour in humans, machines and animals. It was delivered from the perspective of a cognitive psychologist looking for common ground and recurring issues. These topics fitted together well, and a broad understanding across domains provided intriguing insights within them. There were no course pre-requisites, although background in introductory cognitive psychology was desirable. No other background, e.g. in philosophy or computer programming, was assumed.
Includes details of syllabus and reading.
Tools for Intelligent Behaviour
Problem solving strategies for well-defined problems, mental models for reasoning and inference, and schemas
as structured knowledge for problem solving. Limitations
in human performance, and recurring themes and issues.
Theories of Human Intelligence
Domain-general versus domain-specific theories of differences in intelligence, working memory capacity accounts, intelligence as inhibition of incorrect answers. Low-level cognition and brain structure and function explanations of differences in intelligence. Synthesis
and more recurring themes.
What is Artificial Intelligence?
Artificial Intelligence: process- behaviour- and task-based definitions. The Symbol System Hypothesis and brain-computer equivalence. Intelligent behaviour as heuristic-driven navigation of problem state spaces. The Symbolic Search Space Paradigm as the dominant approach to Artificial Intelligence.
Artificial Intelligence Landmarks
A survey of the famous programs; solving games and puzzles, natural language processing and expert systems. Knowledge-light versus knowledge-heavy approaches. The harsh evaluation of the first 25 years of attempts to create Artificial Intelligence.
Artificial Intelligence: Observations, Objections
The Turing Test as a behavioural benchmark. The five barriers to true Artificial Intelligence; the Logic Problem, the Knowledge Problem, the Frame Problem, the Intentionality Problem and the Symbol Grounding
Problem. The human brain versus the computer.
Artificial Intelligence: Where Next?
CYC as the last standing for traditional Artificial Intelligence, connectionism, situated cognition &
Artificial Life.
Animal Intelligence Tests
Pitfalls in studying animal intelligence. Understanding and measuring individual differences in intelligence between species, and individual differences within species. Working memory capacity and attentional focus as the key to understanding differences in ability between species (including humans).
Tools, Puzzles, Beliefs & Intentions
Natural tool use, problem solving experiments in the laboratory, the nature of causal beliefs and intentions.
Animal Communication
Communication in the wild, attempts to teach animals sign or artificial languages. Implications of their failure for theories of intelligence: words as symbols versus words as tools, and the cognitive fringe-benefits of language.
Animal Theory of Mind & Deception
Cognitive capacity versus modularity as explanations for social behaviour. Animal (lack of) awareness of other minds. Deception in the wild and in the laboratory. The evolution of general intelligence as a response to deadly technology, not social complexity. Looking back across the ten lectures.