Is the thousand brains theory a completely theory for the neocortex or is it just a big peace of an incomplete puzzle? If so what’s more that Numenta has to figure out for a full theory?
Until we hear what exactly is “reference” frame in terms of data structure realized by real neurons (axons/dendrites/synapses, APs and cell polarization) - not yet complete.
General direction is right : multiple (about gazillion) sequences from all senses forming an idea of an object (ah, that coffee cup…).
The question is how a sequence is encoded in the brain ?
Whenever trying to solve a any complex problem, first thing we need to do is come up with a idea (more like a formula) that is reasonable or valid. Then we can form a complete theory out of it. Numenta’s Thousand brain theory is an idea or formula.
Is the thousand brains theory a completely theory for the neocortex or is it just a big peace of an incomplete puzzle? If so what’s more that Numenta has to figure out for a full theory?
From my perspective, Thousand Brains Theory (TBT) seems to explain the neocortex at a high level in its apparent features. And it does so very well. However, TBT does not claim to describe all the details here. For example, the role of areas of the brain that interact with the neocortex. Furthermore, I could also imagine that there are other high-level functions that have not yet been discovered, but are absolutely needed for the functioning of the neocortex. How exactly does the thalamus work on its own and what role does it play from the perspective of the neocortex? At least the pathways leading away from and to the individual layers of the isocortex are known, but not what semantic meaning a certain signal now has when it arrives somewhere.