User:CognitiveMMA/sandbox/General Collective Intelligence

Research suggests that groups of individuals have an innate degree of general problem ability that has been described as the "general intelligence factor (c)" characteristic of each group. Similar to the g factor (g) for general individual intelligence some collective intelligence researchers have attempted to extract this general collective intelligence factor (c factor) for groups as an indication of a group's ability to perform a wide range of tasks. As g is highly interrelated with the concept of IQ, as a measurement of collective intelligence this c factor has been interpreted as the intelligence quotient for groups (Group-IQ).

A General Collective Intelligence or GCI platform is a hypothetical platform that combines  groups into a virtual collective cognition with general problem solving ability that might be exponentially greater than that of any individual, thereby enabling it to potentially solve "wicked problems" like poverty or climate change where they can't reliably be solved otherwise. While a complete GCI has not yet been implemented, some functionality of GCI has already been approximated, and case studies based on this subset of functionality have confirmed this potential to exponentially increase problem-solving ability and therefore impact. By analogy, where an Artificial General Intelligence (AGI) is an Artificial Intelligence (AI) with general problem-solving ability (intelligence), a General Collective Intelligence platform is a Collective Intelligence platform with general problem-solving ability at the group level. Where individual general problem-solving ability or intelligence is represented as solving the problem of optimizing outcomes for some individual entity, general problem-solving ability at the group level (collective intelligence) is represented as solving the problem of collective optimizing outcomes for those participating in the group. In addition to the innate general collective factor of each group, collective intelligence platforms potentially provide groups with increased problem-solving ability in some narrower range of problems, such as medical diagnosis, or management of agricultural resources. A model of the mechanisms by which an innate general collective intelligence factor might arise has been proposed. However, since the introduction of the concept of a "collective superintelligence" a number of methodologies by which that general collective intelligence factor might be increased into a "collective superintelligence" have also been proposed. Thomas Malone of the MIT Center for Collective Intelligence (CCI) has associated the concept of an extremely high group general intelligence factor with that of a supermind. Others, have proposed the concept of a Global brain. However, a General Collective Intelligence or GCI is a model of collaborative computing that might be implemented by a software platform and applied to any problem in general, as opposed to collective intelligence from the perspective of the supermind concept which describes a problem-solving methodology that must be manually applied to each problem, or as opposed to collective intelligence from the perspective of the "global brain theory" which describes an evolving concept concerning the net impact of current technologies connecting individuals in a way that improves general problem solving ability.

Empirical Evidence on Problem-Solving Methodologies in Different Domains

Research across various disciplines indicates that problem-solving strategies are influenced by factors like signal to noise ratio, bandwidth, and relevance and engagement. These factors determine the most effective methodologies, ranging from individual expert solutions to collective consensus. In just the case of signal to noise ratio:

Signal to Noise Ratio:


 * Low: Problems with a low signal to noise ratio often benefit from the input of a few experts. These individuals can discern the optimal solution amidst the 'noise'.
 * Medium: In these scenarios, a combination of consensus and expert knowledge is effective. Collective discussion aids in navigating partial signals.
 * High: High signal to noise ratio problems require a collective consensus approach, as the solution is evenly distributed among the group members.

Implications for Collective Intelligence:

This evidence challenges the one-size-fits-all approach of consensus. It underscores the importance of contextually adapting problem-solving strategies to effectively leverage collective intelligence in diverse situations. The collective intelligence field, while often leaning towards prioritizing consensus, is of course diverse. Researchers and practitioners adapt their methodologies based on problem specifics, highlighting the field's dynamism and adaptability.

However, it’s also critically important to distinguish the concept of a General Collective Intelligence platform able to orchestrate cooperation to select whichever problem-solving method is most “fit” in terms of optimizing collective outcomes, as opposed to a narrower collective intelligence method in which decision-makers effectively “compete” to impose their decision-making strategies. Theoretically, networks of interventions can cooperate to vastly or even exponentially increase their impact as compared to interventions that compete for impact on their own. Consensus fails in its ability to select the networks of interventions that are potentially capable of this increase in collective impact, since such strategies are currently too new and potentially too disruptive for consensus to be possible.