The Affects of Competition and Cooperation on The Probability Paths used in Determining Success in Goal Seeking Systems
Based on Human Learning as a Modeled Goal Seeking System Derived from MemBrain ANN
by Michael Ryan
It had dawned on this Goal Seeking System, (Me), that the Concept of Cooperation and Competition was not taken into account in ANN software when designing best course, least impedance (cost) solutions to problems. So I thought, why not get those Axon’s a pumping and write an overview of the potential affect(s) of Cooperation and Competition on the paths created by Goal Seeking Systems, both Organic and Synthetic.
The Affects of Competition and Cooperation on The Probability Paths used in Determining Success in Goal Seeking Systems
What are the lowest common denominator and lowest impedance factors and their roles in a goal seeking function machine? Does Competition lower the likelihood of success in otherwise high probability decision paths? Does Cooperation increase likelihood of Goal achievement?
Initial New Paths to New Goals often do not consider competition, merely the attainment of the goal. In the initial stage of Learning, the question that generally begins the program for the Goal Seeking Machine is: “what do I need to do to achieve this Goal?” The question of “who else” is attempting this path, is usually not considered.
Hardwired path choices to goals based on “low impedance” or least cost assessments may provide starting points, however initial successful outcomes are often achieved without considering the affect of competing function machines. That is, First Decisions to follow a Path to a New Goal are driven by a least cost assessment of all the possible paths to that desired goal where the optimum path being the shortest possible line between “start” and “finish.”
High Impedance Factors most likely come from competing goal seeking machines whereas Low Impedance Paths are created by Cooperating systems.
Best Guess, Feedback and Feed-Forward are nodal functions that determine the future paths of action based on probability of success, however do not innately take into account the reduction of the likelihood of success due to competition.
It is possible that popular paths to popular goals present lower probabilities of success. It is when the Goals become popular that the least cost assessment must include competing factors.
Experience is a combination of least cost trial and error attempts at achieving a goal…a goal that is desired usually to minimize pain and maximize pleasure. When many seek the same things, competition multiplies the risk, cooperation minimizes impedance. See Hobbes, Bentham and Peirce.
Present favorable probabilities of considered pathways to a goal are determined by past tests of Trial and Error for the most part without regard to the reduction of success likelihood due to competition.
Failure to achieve goals both in the past and the present are often not the fault of following low probability paths but of the competition for high probability paths. It is possible therefore to choose a “proven” high likelihood path to a highly desired goal based on experience and still fail to achieve goal because of “hidden” intermediary impedances within the path caused by competition.
When creating current decision paths and assigning probabilities, likelihood of path competition should be considered.
When the competition rate increases and the number of goals decreases, the likelihood of success is subsequently decreased by a multiplying factor…the more people desiring the same things, or the fewer number of desired things will create reduced likelihood paths therefore the lower likelihood of achieving a goal.
By extension therefore, the more machines there are seeking the best path to a goal each running a program based on optimum least impedance solutions, the more likely the goal will not be reached by any of them. However, if a “distributed” sharing of likelihood discovery paths were used, the sum of total of each should create the absolute least resistant solution path to the goal.
Is it possible that Cooperation as the term implies thus: Co-Operation actually reduces path impedance and increases likelihood of success?
Applying these arguments to a biologic goal seeking machine, i.e. the human brain, a question should arise: do neurons compete or cooperate when creating a complex connection? Is it that the “best” set of cells creates the thought first and remains the ruler of that concept? Or are a myriad of cells set to the task “cooperatively” and actually end up supporting one another in completing the connections? Perhaps a combination: many cells (machines) are set to the task of creating best paths to a goal (creation of an Idea), collectively, then the many cells create complex routes or paths based on ion strength (likelihood of successful conduction), and once the goal (Ideation) is achieved, each path is still used based on changes or requirements of re-accessing the Idea.
It is likely that the Abstract creation of a bit of Computer Software that mimics in gross terms the workings of the Human Brain can at best give us a starting point for a discussion of the dynamics of Learning and the success or failure of achieving popular sought after goals.
Copyright © 2007 Michael Ryan. All Rights Reserved.
