10+1 attributes of ideal challenges for Collective Intelligence

Desafios-1

Not all problems are equally suited to a collective approach. In this post I propose a way of typifying problems most likely to be successfully treated with CI. Here is a list of 11 attributes of a task or challenge that give reason to believe it is particularly suited for the use of Collective Intelligence. The greater the number of these attributes presents in a certain problem, the greater the chance it is wise to go for a collective stand:

1.- Geographically highly disperse data that is costly to collect: Situations in which collecting and aggregating large amounts of data can significantly improve our analysis but in which this data is so highly dispersed that it is expensive and cannot viably be gathered by a small group of agents.

2.- Vastly varying views when interpreting the problem: When a problem, or its interpretation, can be seen in different lights, depending on the interests, roles and experience of different agents in relation to the challenge, it would seem a good idea to create a collective space in which these differing perspectives can meet. CI is favorable if diversity is a factor that affects the quality of the final results.

3.- Multidisciplinary nature: Situations that may coincide with previous attribute, but in this case refer to cognitive diversity (neither roles nor interests, differing paradigms) that requires the solution of a complex problem with inputs from different fields of knowledge. As we shall see, the greater the mutidisciplinarity of a problem, the more can be gained with CI because participating agents will self-select and no point of view, that can add value to the analysis, will be lost.

4.- Mistakes, negligence or infringement causing serious collective damage: When a collective system or area is susceptible to failure, excess or abuse that can damage the group’s results. The greater the negative impact of these mistakes and the harder they are to detect by a control unit, the greater the justification for creating mechanisms for collective and distributed vigilance that help to create transparency.

5.- The possibility of splitting a challenge into multiple micro-tasks: If a problem can be broken into numerous “micro-tasks” that can be quickly executed with little effort, then it can be approached by a large group of people working in parallel. This will lead to a noticeable increase in productivity.

6.- Uncertainty relating to necessary knowledge and resources: The resolution of the task or challenge demands resources, and talent, that are unknown to all. In this case there is uncertainty relating to the types of resource and knowledge that are needed, and we don’t even know where to find them. This scenario is so uncertain that it is not possible to define a search strategy and, therefore, the best we can do is to invite the collective to self-select and help to identify the pieces needed to reach a solution.

7.- Activities in which the participants passion and enthusiasm is decisive: Those tasks where the enthusiasm and involvement of end users makes a significant difference to the final result. When this is the situation, collective empowerment, handing over control, contributes in achieving an increase in quality that is hard to estimate, but facilitates achieving the critical mass of active participants needed for the sake of viability (the higher the threshold, the greater the convenience of promoting a participatory logic).

8.- High inter-dependence in the use of shared resources: Situations in which there are dilemmas and opportunity costs related to the use of scarce public or collective resources. In such cases, the interdependence among the users of these resources requires collective negotiation. In this scenario, the collective option is indeed the only option that leads to a legitimate solution.

9.- Problems that are very sensitive to biases: Challenges that are susceptible to partial or incomplete analysis, due to implicit interests or objective reasons. The adoption of open participatory models, that correct individual biases by increasing the number of participants and, above all, their diversity/variance, can for example help to cut the “expert bias” and enable a more impartial judgement of the problem. Concrete examples are initiatives that predict and estimate the probability of future events, because large numbers of participants help to cancel biases.

10.- Activities with, up to now, wasted socialization potential: Areas in which there is a potential sense of community, or where there are things to share because people want to create an experience in common. Emotions and sense of membership is a definitive of this case. If the project touches areas (products, processes, objectives) that arouse social instincts (sharing, debating, demonstrating, teaching, exchanging, etc.), it is a good candidate for applying CI.

11.- Advantageous access to information leading to asymmetries of power: Situations in which the excesses of some agents that make use of information asymmetry must be corrected. When these agents enjoy, thanks to their resources, authority or other advantages, privileged (and exclusive) access to relevant data that affects collective interest, the lack of balance must be counteracted. To do so collective data collection and processing mechanisms can be activated increasing the number of observers/antennae and diversifying the access strategies to that information, thus enabling it to be free and equal opportunity in its access. A concrete example is collective actions to generate open data, in contraposition proprietary data in private hands, that anyone can use and share with independence of social hierarchy or purchasing power.

Notes: This post was translated into English with the help of Peter Hodgson. The image of the post belongs to the album of Benjamin Lehman in Flickr. Read this post in Spanish (Lee este post en Español)

Leave a Reply

Your email address will not be published. Required fields are marked *

*