Does Creating an Artificial General Intelligence Require General Collective Intelligence in Order to be Reliably Achievable?

Abstract

General Collective Intelligence or GCI has been predicted to create the potential for an exponential increase in the problem-solving capacity of the group, as compared to the problem-solving capacity of any individual in the group. A functional model of cognition proposed to represent the complete set of human cognitive functions, and therefore to have the capacity for human-like general problem-solving ability has recently been developed. This functional model suggests a methodical path by which implementing a working Artificial General Intelligence (AGI) or a working General Collective Intelligence might reliably be achievable. This paper explores the claim that there are no other reliable paths to AGI currently known, and explores why this one known path might require an exponential increase in the general problem-solving ability of any group of individuals to be reliably implementable. And why therefore, AGI might require GCI to be reliably achievable.

The Role of Cognitive Computing and Collective Cognitive Computing in the Future Internet

“…The takeaway is that this model suggests that the functionality required to maximize the ability to implement the future Internet is the functionality required for both adaptive problem-solving and for massively increasing that problem-solving ability, such as through facilitating massively decentralized cooperation. Since GCI might be required to achieve AGI [14], GCI provides what is believed to be the only comprehensive model for both. By extension, this implies that without GCI hidden bugs (or intentional features) in any implementation must prevent optimal success in the stated goals of the future Internet from being reliably achievable [13].…”

General Collective Intelligence vs the Innate Collective Intelligence Factor

“…Therefore any problem that has been defined or solved at any time in history is by definition within the capacity of individuals or groups to solve without GCI. And by the same token, any problem that has not yet been solved, whether global poverty or other sustainable development goals [8], [9] implementing an AGI [10], universal global access to health care, smart sustainable cities, or even achieving a higher level of convergence in science [1], perhaps as represented by the so-far unattainable goal of a unified field theory, may not be reliably solvable without GCI.…”

Applications for General Collective Intelligence

“…Semantic modeling of concepts has been suggested to required to achieve human-like general problem-solving ability. Because semantic modeling of concepts is required in order for an AGI to understand and reason with such concepts, because in the absence of AGI that modeling may have to be done manually, and because of the sheer magnitude of this modeling, GCI has been predicted to be required to actually implement the proposed model of AGI [10]. For this reason, and because GCI is expected to be required in order to make AGI safe, despite the ground-breaking importance of AGI, GCI has been suggested to be far more important [11] Conclusions GCI has been suggested to have the potential to drive an exponential increase in general problemsolving ability.…”