So about 2.5 years ago after Id designed and implemented the second major addition to the Action Oriented Workflow (AOW) paradigm the inclusion of dynamic routing of actions using the Action Delta Assessment (ADA) algorithm. I conducted a lot of research on how some further modification of the algorithm would make it amenable for use as a cognitive algorithm. I reasoned that the similarity in how workflow stages associated with discrete actions and possibly conducted by a sea of potential agents per stage was like the action potential modulations that were enabled by axon and dendrite junctions along neuronal chains. Running with this idea in the next few weeks I wrote a few blog posts on my ideas regarding the importance of emotional and what I called autonomic modulation to driving the cognitive dynamo. I became more and more convinced of this as I read more neuroscience on the internal construction of the brain and as I started the implemenation of building ADA into AgilEntity toward the end of 2011 I wrote down on a sheet of paper the dynamic cognition state diagram that Ive mentioned several times in posts regarding AGI. Later in 2012, I collected my thoughts on the importance of salience in the form of autonomic and emotional drivers to dynamic cognition into a formal theory with a set of testable hypothesis for falsification purposes. The state diagram I designed a year before is made to uphold those hypothesis I called this theory the salience theory of dynamic cognition and as such the state diagram is a salience reliant state diagram for enabling emerged dynamic cognition or AGI. In the last few months Ive been baking up solutions for bootstrapping certain aspects of the autonomic and emotional modulators, I hadnt fleshed out in the first version of the state diagram if the weightings for modulation would be linear or functional, Ive concluded that they must be functional in all cases of salience gradation but one which will be linear. This master drive would set the salience metric of the entire cognitive engine and pace the dynamism of the eventually emerged mind. Also, using functional modulation enables dynamic remodulation to test various cognition responses under given inputs of external experience. Think: Tuner for the mind. Id been hesitant to share my state diagram for obvious reasons but am starting to think maybe I should publish it. If I specify the modular interactions sufficiently and put it out there as a model for intrepid builders to try and implement and if they succeed in building a learning and dynamic cognitive agent, then that proves the model itself. Why wait until I can built an implementation if I can have the crowd of zealous AI researchers built various attempts at implementations at the same time? Any success of a falsifiable theory backed by a state diagram that when implemented actually generates dynamic cognition would be in itself a revolution in the field of AI, bestowing recognition for that achievement as much on the theorist (me) as on the implementor. Still thinking.... sent2null.blogspot/2013/02/on-consciousness-there-is-no-binding.html sent2null.blogspot/2013/11/salience-theory-of-dynamic-cognition.html sent2null.blogspot/search?q=autonomic sent2null.blogspot/search?q=emotion
Posted on: Wed, 19 Mar 2014 18:44:46 +0000
Trending Topics
Recently Viewed Topics
© 2015