If one presupposes (and that is what it is: presumption/belief, and not well established or well-founded assumption) that the development of each and all of the hierarchical levels of thinking and learning (aka "abstract" thinking) do NOT involve innate guidance of basic sorts of perception/attention for the development of new ways of learning and
thinking, then you are left with 2 possiblities:
(1) Simple learning processes (alone, and unaffected in their nature) are behind all learning (this includes both the classic simple types of learning and the vaguer ideas/notions of "social learning").
The result here: each individual is VERY individual and learning is quite arbitrary -- prohibiting any standardization of the understanding of behavior needed for AI.
OR
(2) development continues to change in character due just to the expansion and elaboration of sensori-motor responses (and elaboration of Piaget's finding that sensori-motor developments are behind basic object knowledge in infancy) (this is what is now known as "embodied cognition"). This simply-by analogy 'conceptualization' of the development of thinking and learning provides no insights, has no real good basis whatsoever, and has no promise of any real contribution to understanding or clarity -- (see "The Poverty of Embodied Cognition", cited below; full text available at https://link.springer.com/content/pdf/10.3758%2Fs13423-015-0860-1.pdf ).
Thus, here again we have a situation where the modeling of human thought is impossible because the [unlikely] "evidence" is always way too indirect to be clear (and thus clearly replicated).
BOTH of these outlooks are based on the presumption that all significant innate factors are present in infancy AND that "higher levels of learning" involve LESS (to nothing) that is innate. (Both of these presumptions has a long history in Western thought, along with notions of dominion and other atrocities.) Both of these outlooks completely doom any real true AI to non-existence.
----------------------------------------
There is a more reasonable alternative to each of these "views" with Ethogram Theory taking opposite, though more likely, positions on each of these issues (and ethology's alternate assumptions are more biologically congruent and likely). Here: Important innate guidance determines perceptual/attentional shifts which alter the course of learning and thinking [ the innate factors being essentially (and, in actual effect) simultaneous with the learning -- or mixed in, if you like ]. These perceptual/attentional shifts hypothesized are _NOW_ TESTIBLE AND VERIFIABLE. With the new eye-tracking technology, etc. we now have, if these exist (and we know how to look), we shall find these in replicable studies.
This will also make human behavior quite possibly replicable in true real artificial intelligence. This is not "a way" to AI, but among all current alternatives, is THE WAY for true full artificial intelligence (take it or leave it). For the beginning of the basic perspective, see the "Human Ethogram" reference, cited below. (For more guidance on the development of true artificial intelligence, see: https://www.researchgate.net/project/Developing-a-Usable-Empirically-Based-Outline-of-Human-Behavior-for-FULL-Artificial-Intelligence-and-for-Psychology )
Article A Human Ethogram: Its Scientific Acceptability and Importanc...
Article The poverty of embodied cognition
- Badges
- Science topic
- Similar topics
- Psychology
- Mental Processes
- Learning