1/2
I’m reading about #AI in the broader sense and how it’s evolved. I worked in experimental perceptual psychology in the early 80’s and even have a paper with #StevenPinker! My career ended in data and information science, so I find #GroundedCognition very interesting - how our brains work as an intersection between #sensation and #cognition. It’s also the goal of #AI.
2/2
“According to #Barsalou (2008), #GroundedCognition refers to the belief that simulations within specific sensory systems, bodily states, and situated action mediate all cognitive processing.”
You might find the work of Rodney Brooks (professor emeritus MIT) interesting. One of his contributions to robotics was identifying the significance of sensorimotor interactions (Behavior-based robotics) and the need for distributed local control (subsumption architecture).
The "Grounded Cognition" model does not explain abstract thinking ("intelligence"). Abstract thinking would include language. In later writing Barsalou conceded that higher cognition is likely "amodal". Even so, the process of transformation from sensorimotor interaction into amodal symbols remains a mystery.
GC hasn't caught on in cognitive robotics. Generally I have found that the research area of "embodied cognition" ("embodied robotics") implicitly includes "situational cognition" and is thus very similar to GC.
I prefer the GC model as it is more explicit and precise in its terminology. But GC doesn't appear to have caught on, even in Cognitive Science (not sure about psychology).
(last one) This thread gives some insight into the research into sensorimotor interactions and cognitive robotics. Covers a lot of topics with some interesting references.
[side note: There are many rabbit holes to dive into in cognitive science/robotics/AI. Easy to get lost. The relationship between cognitive science and cognitive robotics (AI) is largely conceptual as the physiological mechanisms of human cognition, particularly for higher level cognition, are not understood].
