Is there a model of the brain that does not use pure computation?
Why? Because the computer paradigm, especially when applied to natural language understanding (NLU) and brains, isn’t effective when needed in artificial intelligence (AI) applications.
Patom theory (PT), a pattern-matching paradigm, is a different approach based on the analysis of brain damage, scans and capability. PT was designed to explain normal brain function and the effects of brain damage. A brain theory should also have evolutionary plausibility since a human brain is very similar to many other animal’s brains.
Instead of data driving a brain with some kind of construction or process, it is stored patterns driving everything: sensory patterns stored with their originating sensors, and matched patterns continuing away from the senses into the brain for further consolidation, specialisation and pattern composition.
In short, the brain is a distributed and connected model starting with senses and ending with muscles, and the reverse. And its role is to store, match and use the resulting hierarchical, bidirectional patterns. Deficits from brain damage are the consequence of lost patterns or their connections to other patterns.
Perhaps the most important take-away today is how a brain represents multisensory…
