Utilizing better solutions with collective intelligence
Rolf K. Baltzersen
We are all familiar with the jelly bean guessing game. Let every kid in the birthday party make their best guess on the number of jelly beans inside the jar. Whoever is closest to the correct number, get all the candy in the jar. However, did you know that the most likely winner is the aggregated average guess from the whole group? This is the classic example of collective intelligence where the diversity benefits of averaging independent contributions from a large group of individuals, provide the most correct answer.
Unlike guessing games, collective intelligence today seeks to address complex problems where the best answer is not known in advance. For example, a study shows that the most accurate probabilistic forecasts of cases and deaths during the COVID-19 pandemic is based on averaging multiple models, rather than one single model.
This book examines a wide range of examples that aim to utilize diversity benefits in an attempt to make better decisions. Case stories cover both the political and scientific domain, underlining the importance of citizen expertise and citizen empowerment:
In Reykjavik in Iceland, the residents elected 111 projects to improve city conditions in 2022. In cities all over the world, citizens are invited to share ideas and communicating with the municipality through digital ideation platforms. The municipality gain access to many more innovative ideas on how to improve societal conditions. As a part of this process, participatory budgeting is also used as a strategy to empower citizens (Better Reykjavik, My Neighbourhood and Decide Madrid).
Representing the “peoples voice” in a more direct way, Citizen Assemblies are becoming a new part of democractic systems in countries like Ireland and Scotland. Here, a representative group of 100 citizens is invited to learn and discuss about urgent political issues such as Climate Change. Discussions may last a year before the “mini-nation” eventually vote on a policy recommendation. The national parliament is required to follow up and respond to the recommendations.
A citizen science projects like eBird involve 100 million bird sightings contributed annually by eBirders around the world with an average participation growth rate of 20%.
In disaster management, it is important that everyone who is affected contribute with data. In the Haiti Earthquake in 2010, nearly 40,000 independent reports were analyzed in a volunteer-driven effort to produce a crisis map after the earthquake.
The COVID Symptom Study app, being the largest community monitoring of COVID in the world. Millions voluntarily shared personal information and answered questions related to any underlying chronic conditions. The app helped identify the problems of Long Covid.
Some scientific journals have started with crowd peer review. Instead of letting two scientists do a separate review, a crowd of 20 individuals will give detailed feedback on the same manuscript. In this way, the peer review process is reduced to a couple of days, the quality improves and a more diverse group of reviewers are also invited to be part of the process.
Data from social media platforms can now also be used to inform policy decision-making. During the early stages of the pandemic, geolocated Tweet Intensity on COVID-19 could strikingly predict the extent of mortality across Italian, Spanish, and United States regions a month later.
Although many of these examples involve the use of digital technology, collective intelligence is not a new phenomenon. The main goal of the book is to explore how nature and humans throughout history and biology have use three distinctly different types of collective problem solving; stigmergic, swarm and collaborative. The author identifies mechanisms that can support humans in their strive to develop new knowledge and leverage diversity.