Michael Wooldridge
University of Oxford
- Research field
- Multi-agent systems, the foundations of artificial intelligence
Michael Wooldridge is a professor of computer science at the University of Oxford and has worked on multi-agent systems since the early 1990s; with Nicholas Jennings he wrote the 1995 survey that fixed the field's definition of an agent as autonomous, reactive, proactive, and social, and he wrote its standard textbook. In a 2025 position paper with Emanuele La Malfa, Gabriele La Malfa, and colleagues at Oxford, King's College London, and Sussex, he examined the systems of several language models that are now called multi-agent systems and argued that most of them lack the field's defining properties: the models are not trained to cooperate or compete, their environments and messages are prose, they run in sequence under an orchestrator, and claims of emergent collective behaviour are not measured. The unit presents this argument after its section on agents built from several models.
In the masterclass
Section 5, after the subsection on several models together, presents the paper's four discrepancies between multi-agent theory and systems of language models.
The research is presented in Unit 15: Reasoning, Tools, and Agents, in a research spotlight box that explains the result at the level of the unit and links it to what you have just learned.