We are pioneering and testing a new methodology for AGI which I have found no evidence has been actually tried. I believe my method overcomes current limits and shows sufficient coverage for understanding with which I will be able to build a basis for human computer interaction.  At the end of the project, from the behavior you might assume what you are seeing to be what appears to be both symbolic and stochastic/connectionistic AGI.. perhaps a hybrid?  But it is neither traditional GOFAI or Machine Learning, it is a new paradigm in AI.  

Phylogenetic Development o/ Behavior

​​​10.1 Methods

Methodologically such investigations are difficult, since one of the most important aids in the study of evolution—fossils—is, for all practical purposes, not available. While there are a few “behavioral fossils,” e.g., petrified footprints, that enable us to conclude something about the method of loco- motion of an extinct species, or petrified stomach contents and feeding tracks in sediment, which permit us to draw some conclusions about the mariner of feeding, their extent and degree of certainty are so limited that it is generally impossible to draw final conclusions.

Statements about extinct species, which are so important for the reconstruction  of  the  course  of  phylogenetic  development,  are  not  available in investigations on the phylogeny of behavior. References to the evolutionary history of behavior patterns are almost exclusively based on the behavior of recent or living species. There are two ways to study the evolution of behavior: the first is to look at the ontogeny of behavior, and the second is the comparison of closely related species, whose position in the zoological classification system and their probable “stage of development” within the particular group are already known from the data compiled in other biological disciplines.

The variety of life forms, their function, and their way of life are the result of a continuously changing process that began with  the origin of  all life. It led to an increased unfolding of all living things, from simple precursors to the development of more highly organized and specialized forms. This process is called Evolution. Its most important driving forces are mutation and selection. Mutations are discrete “jumps”—changes in the genome. They are random events with respect to the direction of evolution and provide merely the raw material upon which natural selection can act.

Selection determines the survival and reproductive chances of an organism: individuals with advantageous characteristics that are best suited genetically to presently existing environmental conditions usually produce more viable and reproductively successful young than individuals with less advantageous adaptations. As a result, the incidence of certain advantageous inherited characteristics increases in a population, and that of less advantageous ones decreases. This results in a gradual change in the population in  the direction of an optimal adaptation to the environment, i.e., it is “im- proved.” In contrast to mutation, selection is always directed.

The behavior of a species is also subject to the laws of evolution. It developed from different precursors and can properly be understood only when its previous “history,” its phylogenetic development, is known.

10.2 Problems of Homology

In phylogenetic investigations of behavior, it is important to know whether the characteristics to be studied and compared are homologous or analogous. Here, ethology again encounters some difficulties not found in other areas, with respect both to methodology and to definitions.

Homogeny is generally understood to be an agreement in structure and other physical characteristics, which depend on the inheritance from common ancestors. In contrast, ANALOGIEs are similarities that evolved in various species independently of any phylogenetic relationships.  Well-known  examples for analogous structures  are the wings  of  birds and  insects and  the lenses of eyes in vertebrates and cephalopods.

The main criterion for a homology is the existence of a common  source of information—normally the genome—that directs the development of the particular characteristics. It is different with behavior, where we know of a second means of storage-memory. In behavior, this has consequences  that can allow the same characteristic to occur in two species that are not closely related, as long as one species has acquired the behavior from the other. This means that carriers of homologous behaviors, in contrast to all others, need not necessarily have descended from a common ancestor. Hence, a homology in and of itself  is not  a criterion for phylogenetic  relationship. For example, if one bird species learns the song or calls of another (see Section 7.4.5), then the vocalizations are homologous with respect to each other, even when both species belong to very different unrelated groups. Hence, we must clearly distinguish in ethology between a phylogeny of characteristics and a phylogeny of a group of animals. The former deals with the phylogenetic development and relatedness of characteristics; the latter deals with the phylogeny of the “cue-bearers,” i.e., the animal species that have the particular character in question.

The way in which information is passed from one generation to the next also differs for both categories of homologous characters: Iron characteristics are passed on by inheritance, and ACQomxo ones by tradition (see Section 7.5). Of special interest for phylogeny are the so-called tradition homologies, where the learning of non-species vocalizations is an actual component of the learning program of a species. This is in instances where the brood-parasitic wydah birds learn the vocalizations of their host species (see Section 7.4.7.2).

Tags: LOGICMOO
     
Copywrite © 2020 LOGICMOO (Unless otherwise credited in page)