Cybernetic investigation of novelty

The novelty research is primary a cybernetic investigation. Learning what is inherently unknown is in the first place cybernetic paradigm of regulation or control. The key concepts as bootstrapping, complexity, mediation, actors, the artificial, etc. help to build a world view where learning novelty is not a paradox. The solution is a system approach to learning. The central model is the novelty-action model (NA-model), or the the act of discovering something new. By our own Artificial Intelligence research a design was created that looked similar to two other analyzers, so the NA-model was discovered. Three independent investigations in different domains came up with a similar control system, there is no coincident. All three where investigating the novelty-action. Our own design by simulations, a neurological design by brain analyzes and a anthropological design describing the emerging of science (see EMCR paper 2008).

For the NA-model two learning processes play a major role: reflection and mastering. Reflection denotes the internal learning taking place in a learning system. Connections are made based on the connections that already exist within the system. As reflection is an internal process, the internal learning does not necessarily result in a knowledge model that is consistent with the external world. During mastery, the internal knowledge model of the learning system is applied to the external world. The external world produces feedback which in turn causes learning, bringing the knowledge model in line with the world outside the learning system. The NA-model is composed of four parallel processes. By acting on the cognitive network of the system, each of these processes contributes to the production of discontinuity in the reasoning. This happens as the concurrent processes share a working memory that entwines the processes: the spreading over the network by one of the processes can get altered by the other processes. This can create sudden jumps in the cognitive network that can bring concepts together that would be unlikely to get together by only one process. Thereby the aggregations of the four parallel processes result in a whole that is more than the sum of its parts.