skip to primary navigationskip to content

Department of Computer Science and Technology

Masters

 

Course pages 2024–25

Computing for Collective Intelligence

Principal lecturer: Prof Amanda Prorok
Taken by: MPhil ACS, Part III
Code: R181
Term: Lent
Hours: 16 (8 x 1 hour lectures; 8 x 1 hour reading groups)
Format: In-person lectures
Class limit: max. 18 students
Prerequisites: Familiarity with core machine learning paradigms - Basic calculus (analytical skills) - Basic discretre optimization (combinatorics)
Moodle, timetable

Objectives

There is a substantial body of academic work demonstrating that complex life is built by cooperation across scales: collections of genes cooperate to produce organisms, cells cooperate to produce multi-cellular organisms and multicellular animals cooperate to form complex social groups. Arguably, all intelligence is collective intelligence. Yet, to-date, many artificially intelligent agents (both embodied and virtual), are generally not conceived from the ground up to interact with other intelligent agents (be it machines or humans). The canonical AI problem is that of a monolithic and solitary machine confronting a non-social environment. This course aims to balance this trend by (i) equipping students with conceptual and practical knowledge on collective intelligence from a computational standpoint, and (ii) by conveying various computational paradigms by which collective intelligence can be modelled as well as synthesized.

Assessment

The assessment will be based on a reading-group paper presentation (individual or in pairs), accompanied by a 2-page summary, and a technical position paper (individual work) handed in after the course. The position paper will be assessed based on its technical correctness, strength or arguments,
and clarity of research vision, and will be accompanied by a brief 5-minute pre-recorded talk that summarizes key arguments (any slides used are also submitted as part of the project).

  • Paper presentation and 1-page summary: 25%
  • Technical position paper: 75% (4000 word limit)

Recommended reading

  • Swarm Intelligence : From Natural to Artificial Systems, Bonabeau et al (1999)
  • Joined-Up Thinking, Hannah Critchlow, (2024)
  • Supercooperators, Martin Nowak (2011)
  • A Course on Cooperative Game Theory, Chakravarty et al., (2015)
  • Governing the Commons: The Evolution of Institutions for Collective Action. Ostrom (1990).