User:CognitiveMMA/sandbox/GCI4WebSci2022/bias

Predisposition to type 1 or type 2 reasoning is believed to be a source of cognitive bias in groups, and therefore a source of cognitive bias in social web applications or other applications that facilitate group interactions. What is commonly lost in discussions about what solution is best for a group is that there are two main cognitive biases individuals might have (the predisposition towards (type 1 or intuitive reasoning and the predisposition towards type 2 or rational methodical reasoning), neither of which permits an understanding of the other. Both biases are useful however, but without a platform such as GCI to help select the optimal bias in each context, we cannot as groups switch between these biases in an optimal way, the same way we can switch between these biases within our individual cognition. Because of this inability, it is predicted that better decisions cannot reliably be made regardless of the amount of research performed, or the amount of discussion or information shared. These two decision styles simply come to different conclusions which can't be reconciled since an individual can't understand type 1 reasoning using type 2 reasoning, and no individual can convince a type 1 individual using type 2 logic, and vice versa. GCI on the other hand addresses this where doing so can be shown to be reliably improbable by any other means today.

In conceptual space problems are defined as the lack of a path allowing the cognitive system to transition from one concept to another. Similarly for the system described by some other functional state space, a problem is defined as the lack of a path allowing the system to transition from one functional state to another. Solutions are defined as the paths which accomplish those transitions. All transitions between functional states (all solutions) in any functional state space are one of two types. Type 1 consists of direct transitions from one functional state to another, and type 2 consists of step by step transitions between intermediate functional states. Type 1 transitions are used to solve uncomputable problems, that is problems that are not computable in terms of known path segments. These problems must be solved through pattern recognition, that is, through recognizing patterns identifying cases in which the same solution has been applied to solve the problem in the past. Type 2 transitions are used to solve problems that are computable in terms of such known path segments. Where all possible transition processes can be represented in terms of a combination of some basic set of functions, those functions are said to “span” the functional state space, thereby enabling every possible type 1 or type 2 process and every functional state to be represented so that the optimal (most fit) solution might be chosen.

As an example, this distinction between the two cognitive biases is critical in understanding the assumptions behind what we assume is or is not fake news, and why what is fake news depends on perspective. Depending on whether the individual designing any fake news detection algorithm is conscious of their predisposition towards a given reasoning style for a given problem in a given topic, their assumption of what is fake news might not only be incorrect, but that incorrectness might make any such algorithm a tool to spread their ideology rather than a tool which increases the group’s collective intelligence. In order to be applied more correctly, such algorithms would need to be identified explicitly as operating within one or more of the quadrants in figure 1.



Figure 1: Quadrants defining true and fake news.

The very concept of fake news might be profoundly misguided. Paradoxically, the most intelligent among us who try to inform the public with their opinions in the hopes they can significantly improve decisions on public policy might be the most misled of all. A General Collective Intelligence is a system of collective optimization in that it solves the problem of determining what collective action provides the optimal collective outcome. Without a system capable of collective optimization, civilizations are predicted to act as systems of individual optimization in that they optimize outcomes for some individual or subset of individuals. Regardless of whether all news coverage was perfectly effective in transmitting truth, and regardless of whether the volume of that truth could be increased exponentially, without a system of collective optimization to maximize collective outcomes, all public policy decisions must target the accumulation of benefit for some individual or subset of individuals faster than they can target the accumulation of collective benefit. In natural organisms consisting of groups of cells, using Human-Centric Functional Modeling to describe the behavior of those organisms in terms of motion through collective functional state spaces, collective optimization is predicted to be required to create enough problem-solving ability to coherently perform any collective action targeting any collective outcome. For example, allowing the billions of cells inside a bird to collectively fly, where billions of individual single-celled organisms could never come together to do so. The cells in the bird have some system of collective optimization that allows them to collectively solve the problem of flight faster than any individual cell could direct them to do in a top-down way. What each cell needs to do in cooperating to collectively achieve any collective action are decisions that occur faster than any single-celled organism could compute. GCI describes the behavior of groups of individuals in terms of functional state spaces as well. So by analogy, without a system of collective optimization such as a GCI, public policy decisions must always be made faster than maximum public benefit can be computed. The problem with public policy then isn't misleading news coverage, its the fallacy that better news coverage is the most effective way of improving public policy where understanding what collective actions maximize the public good is not within the capacity of even the most intelligent of us. On the other hand, GCI is predicted to have the potential to exponentially increase collective outcomes so that the biggest challenges facing human civilizations might become reliably solvable. No news coverage or other exchange of information, no matter how perfect, and no matter how large, has this potential. Because we simply aren't collectively intelligent enough without GCI to make significantly better use of the information.