Reviewed by
Christopher Watts
Collective intelligence is the learning, decision-making, sense-making and problem-solving capabilities of social groups, firms, NGOs, governments, societies and humanity in general, in contrast to the capabilities that individual human beings have in isolation. Interest in knowledge capabilities and governance may be as old as Plato, but during our lifetime a number of recent developments have led to the emergence of a new discipline, combining behavioural and cognitive sciences with an interest in the potential consequences of technological innovations, including the Internet and World Wide Web, AI, Big Data analytics and machine learning, and collaborative projects such as Open Source software and Wikipedia. Collectives can include human-machine hybrids, and potentially they can also give voice to local stakeholders, and redress the balance between political elites and those whom their policies will affect. The new book, “Big Mind”, is an attempt to theorise the new discipline in this socio-technical context.Its author, Geoff Mulgan, pulls in ideas from a variety of disciplines, especially Psychology and Computer Science, to present his own theory of collective intelligence, and then in later chapters illustrates their application, for which he can draw upon his own experience as founder of two UK-based think tanks (DEMOS, NESTA), and participant in many projects to improve the intelligence expressed through policy making at local and national administrative levels.
Readers with only partial knowledge of some of the background may find the book a valuable introduction to the subject. At 280 pages long, it attains a reasonable balance between covering a lot of diverse material, and remaining appealingly short and readable for a non-specialist readership. Some might have preferred a narrower focus, allowing for a stronger empirical case to be built. Instead, Mulgan tends to illustrate his points from many disparate subjects and sources, including poets and ancient wisdom alongside the sciences, and readers may be distracted by the jumps rather than impressed at the breadth of his reading.
This reviewer, however, was more intrigued by what it omitted, and by what it was not.
For JASSS readers, the main thing to note is that we are not in it. Apart from some brief namechecks (Simon, March, Helbing), there are no references to Social Simulation. As readers may know, simulation models have been applied in the studies of organisational learning, knowledge dynamics and innovation networks, peer review and publication in science, and the philosophy of social epistemology. All of this would seem relevant as examples of studying collective intelligence. On the evidence of this book, however, we appear to have more work to do in reaching both think tanks and policy makers.
Despite its University Press publication, the book is not an academic literature review – the namechecking in the Afterword is far too brief for this, and the notes section is helpful for some arguments and silent on others. See instead Malone and Bernstein (2015) for a review that covers much of the ground.
The book is also not a practical, “how-to” guidebook for managers. Typical of Mulgan’s rhetorical style is argue that an organisation can have too much of something (for example, exploration rather than exploitation), then also cite cases of it having too little. The implication is that there exists some ideal, golden mean between these extremes. (Mulgan might have referenced Aristotle on this point, or even Goldilocks.) But no practical guidance is offered as to how managers are to identify on which side of the optimum their organisations currently lie: for example, do they need more or less exploration of new ideas?
The book omits any history of its themes, but they have been discussed before. Kline (2015) or Rid (2016) are good recent accounts of how 1940s Cybernetics, the study of governance systems in the age of human-machine interactions, evolved into the Information Age. Collins (1998) traces the workings of the “invisible college” by which we inherit intellectually.
Also missing is any mention of alternative theories of collective intelligence. Appreciation for Mulgan’s own theorisation will be easier when it has some rivals. For this reviewer’s suggestions, alternative treatments might:
• Focus on just one type of collective. It seems over-optimistic of Mulgan to assume that the one theory will account for groups, firms and societies, using many different media.
• Consider how the collective’s environment adapts to the collective’s learning. As a stable collective emerges in knowledge and power, how does this affect the object of knowledge?
• Focus less on inputs to intelligence, i.e. data, and more on how policies become enacted. (A curious omission from a book by the founder of NESTA, formerly known as the UK government’s “Nudge Unit”.) How do the results of analyses motivate?
• Give mention to studies of the role played by narratives in sense making, idea generation, problem solving and building social capital in organisations (Brown and Duguid, 2000).
• Consider the possibility of conflict between the intelligence of the collective, and the individual’s intelligence gained via membership of a collective. Mulgan occasionally confuses the two, but, like the goals of organisations and their staff, these can often conflict.
• Examine more the role of political, technocratic and intellectual elites (including think tanks!) Why is problem solving not done at the local level? Why does collective intelligence need levels and hierarchies? Are there perhaps useful analogies to be had with Deep Learning Neural Nets or with the concept of predictive processing in recent Cognitive Science?
A final, pessimistic thought. Mulgan’s enthusiasm for the potential of collective intelligence may already be out of date. In the age of fake news, twitter bots, and Internet surveillance funded by both governments and billionaires, does information technology still make us collectively smarter? Does it rather enable those with extreme views to find and echo each other? Does Popper’s and Soros’s Open Society survive when states and the 1% can buy so much share of voice?
Perhaps an agent-based modeller can address these issues? Conflict energises, and if Mulgan’s book prompts some critical responses, we shall hopefully be (collectively) better off.
ReferencesCOLLINS, R. (1998). The sociology of philosophies: a global theory of intellectual change. Cambridge, Mass.; London: Belknap Press of Harvard University Press.
KLINE, R. R. (2015). The cybernetics moment, or, why we call our age the information age. New Studies in American Intellectual and Cultural History (pp. 1 online resource (351 pages).). Retrieved from http://ebookcentral.proquest.com/lib/warw/detail.action?docID=3433433 Connect to Proquest e-book
MALONE, T. W. and Bernstein, M. S. (2015). Handbook of collective intelligence. Cambridge, Massachusetts: MIT Press.
RID, T. (2016). Rise of the machines: the lost history of cybernetics. New York: W. W. Norton.
Return to Contents
of this issue