What is Collective Intelligence?
Collective intelligence can be defined as “the shared intelligence of a group”. In the book “Superminds” by Thomas W Malone, from MIT’s Centre for Collective Intelligence, he refers to collective intelligence as a “supermind”. Defining superminds as “a group of individuals acting together in a way that seems intelligent”.
On the surface, this is a simple concept, however, how you design groups of people to work together can massively vary the collective intelligence of the group, hence it is an important topic to study. In general, the intelligence of a group correlates with the intelligence of individuals in the group, but there are also a multitude of different factors that influence it. The most prominent factors tend to be the social intelligence of group members, and allowing all members to equally participate.
Furthermore, how groups are designed to interact with each other will affect the type of decisions they are good at making. In Malone’s book he splits all groups into five different “superminds”: hierarchies, democracies, markets, communities, and ecosystems. These systems all work very differently, and the “collective intelligence” of these systems will vary greatly depending on how they are designed.
AI and Collective Intelligence
Humans are not the only individuals that can make up a group with collective intelligence, computers can also be part of a group. Computers have already had massive effects on creating more collectively intelligent groups, for example by enabling increased group size or decreasing the cost of decision making (e.g. no longer have to meet in person to make decisions).
However, the effect of computers on the collective intelligence of a group may be in the course of radically changing with the emergence of generative AI and large language models (LLMs). One of our Butterflies, Martin, recently attended a conference at MIT, exploring the impact of these technologies. He was also lucky enough to have a chance to meet with members of the Centre for Collective Intelligence (CCI) to discuss how these methods can be implemented in the work that Butterfly Data do. He even wrote a blog article on it which you can check out here.
One thing that particularly stood out from meeting with CCI was their development of tools using ChatGPT (which they actually had access to before the public!), that enable groups to come up with better ideas and solutions. One of their tools is called the Supermind Ideator, which will help come up with novel ideas and solutions. Many of these ideas are, in their words, “wrong, silly, or useless”. However, using this in collaboration with humans can result in a collective intelligence superior to humans alone.
AI and humans are both intelligent in their own way, but they can be even more intelligent when they are combined together.
When Has Butterfly Data Used Collective Intelligence?
On a recent project for cross government departments we utilised collective intelligence in the development of a single source of truth platform on the UK's Net Zero supply chain.
The work stemmed from the following requirement: ‘The UK Government needs a clearer understanding of the opportunities and challenges faced by the Net Zero supply chain in achieving theirour 2035 target […] so that they it can make informed decisions about support, expectations, and optimal routes to international advantage or collaboration.’
Butterfly Data developed an Alpha bringing together open-source data to identify challenges and opportunities in the net zero supply chain. For the Alpha we utilised our in-house solution Collide, developed a tailored data platform for cross government usage, and conducted data gathering. This helped to enable an improved collective intelligence approach because all members had an increased and central understanding of the current net zero supply chain.
We worked with intandem and Bays Consulting who focused on discovery and worked together discussing and drawing the insights together. We also brought in a diverse expert team from within Butterfly Data including 3 PhDs and experts across Design and User Research, Materials Research, Physics, Machine Learning and Software Engineering.
The work brought diverse and innovative expertise together to help tackle complex challenges. The collective intelligence involvement included workshops, face-to-face presentations and discussions, panel and Q&A sessions, weekly sessions discussing the challenge and data, and then utilising research and data as part of these discussions. These sessions included academia, industry, and over seven government departments. This totalled over 150 involved contributing to the collective intelligence and development. This also included a number of experts in or around collective intelligence, Peter Baeck (Director of the Centre for Collective Intelligence Design Nesta), Prateek Buch (PolicyLab), and Floriane Fidegnon-Edoh (BMNT).
We demonstrated to the client the importance of a single source of truth shared across Government departments in meeting the UK Gov 2035 and 2050 net-zero goals. We utilised our data analytics platform, Collide to provide a solid and reliable infrastructure to host, share and visualize data.
We had the opportunity to conduct early user testing by multiple users across seven government departments. The overall feedback has been very positive, including the utilisation of collective intelligence.