https://www.youtube.com/watch?v=CmuYrnOVmfk
He has identified three areas near the brainstem as being the 'upper bound' for the structures necessary to produce rudimentary consciousness a.k.a subjectivity, the internal feeling of 'what it is like' to be something. Namely, he has identified the reticular activating system, the periaqueductal gray, and the superior colliculi as being those structures. He believes that pleasure and pain are the most basic forms of consciousness (It is contradictory for pleasure/pain to exist without a subjective observer). Thus, because these primitive structures generate pain and pleasure, and for other reasons, he believes that these structures are the origin of subjectivity. This is contrary to the mindset of scientists who have mistakenly assumed that subjectivity arises in 'higher order' areas such as the cortex. He also sees pain and pleasure as 'performance standards' (like objective functions) from evolution which guide an organisms (or an agents) decisions which allow it to learn, and believes such a function can be modeled mathematically if we study these primitive brain structures. Furthermore, he thinks that we may be able to get a proto-agi by either modelling the mathematics of these primitive structures, or even faster by simply attaching one of these tiny primitive pain/pleasure generating proto-brains to a brain computer interface and allowing it to control a computer program, which will allow a computer program to use pain as a reward function, and vice versa allowing the minibrain to use the computer program as an actuator to act on it's environment based on the pain/pleasure reward function. Anyone have thoughts on this?
He has identified three areas near the brainstem as being the 'upper bound' for the structures necessary to produce rudimentary consciousness a.k.a subjectivity, the internal feeling of 'what it is like' to be something. Namely, he has identified the reticular activating system, the periaqueductal gray, and the superior colliculi as being those structures. He believes that pleasure and pain are the most basic forms of consciousness (It is contradictory for pleasure/pain to exist without a subjective observer). Thus, because these primitive structures generate pain and pleasure, and for other reasons, he believes that these structures are the origin of subjectivity. This is contrary to the mindset of scientists who have mistakenly assumed that subjectivity arises in 'higher order' areas such as the cortex. He also sees pain and pleasure as 'performance standards' (like objective functions) from evolution which guide an organisms (or an agents) decisions which allow it to learn, and believes such a function can be modeled mathematically if we study these primitive brain structures. Furthermore, he thinks that we may be able to get a proto-agi by either modelling the mathematics of these primitive structures, or even faster by simply attaching one of these tiny primitive pain/pleasure generating proto-brains to a brain computer interface and allowing it to control a computer program, which will allow a computer program to use pain as a reward function, and vice versa allowing the minibrain to use the computer program as an actuator to act on it's environment based on the pain/pleasure reward function. Anyone have thoughts on this?
