The Thousand Brains Theory of Intelligence
Image source: Google

Recent progress in artificial intelligence, i.e. deep learning, has borrowed concepts from the human brain. The architecture of the deepest learning model is based on the layers of processing - an artificial neural network which is inspired by biological brain neurons.

Yet Neuroscientists does not agree with what the intelligence is, and how it has been created in the human brain - this is an incident that remains chaotic. 

Jeff Hawkins, co-founder, Scientist, and Technologist at Numenta presented the innovative concept to understand deeply how the human neocortex functions and he named theory as "The Thousand Brains Theory of Intelligence". He with his team presented this concept at the Human Brain Project Conference in the Maastricht of the Netherlands in October 2018.

The Neocortex is the part of the human brain that is involved in high-order functions such as conscious thoughts, spatial logic, language, generation of motor orders, and sensory perception. 

Researchers from Numenta believe that every part of human neocortex memorizes the complete model of objects and concepts. The team hypothesizes that neurons are present in every column of grid cells.

The research team also proposes a new type of neuron called the displacement chamber, which acts as a supplement to the grid cells, and is also located in the entire neocortex. 

Grid cells are modulated neurons that enable understanding of the condition. Researchers believe that each cortical column grids learns the model of the whole object by combining the input with the cell-derived location, then integrates the movements.

In order to illustrate this concept, researchers used a coffee cup as an example. When we see and touch a coffee cup, many columns in the visual and somatosensory hierarchy parts of the cup together. In each field, each column learns the complete model of the cup based on the sensory input (in this example, sight, and touch) with the object-oriented location of that input, and then integrates the activities of the sensor. 

The Thousand Brains Theory of Intelligence
Image source: Google

Cup models are not the same because each model of the cup is memorized from a different subset of the sensory arrays the commonly held view, where sensory input is processed in the hierarchy of the cortical regions, this theory states that the connections are not hierarchical in nature. 

Instead, non-hierarchical connections may be combined between brain hemispheres, and methods and hierarchical levels. Due to non-hierarchical connections, there may be estimation with the movements of the sensor.

According to the researchers, Neocortex has thousands of models of each object in the world at all the levels of the hierarchy. At all the levels of the hierarchy, the integration of the facilities observed in each column, not only on the top of the hierarchy - therefore name is given, "The Thousand Brains Theory of Intelligence". The structure again defines how human neocortex works.

According to the researchers, thousands of models not only work in the hierarchy but also work in parallel in Neocortex. It is an innovative theory that challenges traditional ideas, and in the future can affect both Artificial Intelligence and Neuroscience.

For more Trending News on Artificial Intelligence and Neuroscience, Stay Tuned with us on FoxSplit.