Cognition vs cognitive computing

In this regard, here are the two actual definitions from Wikipedia:

"COGNITION refers to "the mental action or process of acquiring knowledge and understanding through thought, experience, and the senses". It encompasses many aspects of intellectual functions and processes such as: perception, attention, thought, the formation of knowledge, memory and working memory, judgment and evaluation, reasoning and "computation", problem solving and decision making, comprehension and production of language. Cognitive processes use existing knowledge and discover new knowledge."

"At present, there is no widely agreed upon definition for COGNITIVE COMPUTING in either academia or industry."

Strong AI with cognitive features requires implicitly a body hardware

But what happens when an AI software package merges with a piece of hardware, such as a frame equipped with many and varied sensors, actuators, and organs? Do we then have a cognitive robot? Well, that depends on the software package. So let's take a closer look.

The AI should comprise a collection of connected neural networks only, meaning no conventional processors, operating systems or mathematical functionality are included. And why? Because the biological brain works in the same way. Then what is it made of and how does it work? Well, it should have some cognitive properties, such as:

1. The neural networks can have either a predefined or a user-defined connectome, i.e., the connections between cells that contain the desired functionality; by changing the connectome, we obtain a Programmable Neural Network, PNN.

2. Sensor input signals should be converted into information by some neural networks with this specific goal, similar to a Sensor Input Driver, SID, for video, pressure, gyro, and other sensor types.

3. The specific neural networks for the ac(tua)tors should be able to send commands that are then translated into signals required for their activation - the equivalent of an Action Command Driver, ACD, for limbs or other body parts.

4. The specific neural networks for the internal organs should be able to receive signals from them and transmit them to them through appropriate transformations, similar to an Internal Organ System, IOS. This replaces the external intervention of a user by generating and maintaining its own needs, such as hunger for energy in order to stay alive.

5. The rest of the neural networks should be able to create a knowledge pool from the incoming information as Simple Glue Logic, SGL, in which a specific knowledge area can be achieved by grouping different neural networks, which in turn can create higher forms of knowledge, e.g., identification of objects both indoors and outdoors, orientation and navigation in different landscapes, performance of different tasks, etc.

6. To create a knowledge area, some specific neural networks would be required, such as a processing of information with an own SLM, Self-Learning Mechanism, a LOG as a working memory, and a long-term memory to store and link the information as a SAM, Self-Associating Memory.

7. The neural networks should also be able to create new connections between cells, but with the help of their own SPL, Self-Programming Logic, which replaces the analog equivalent of the biological neuron, namely the reading and decoding of the necessary genes to produce the appropriate proteins needed to build new synapses.

8. The neural networks must all execute simultaneously, creating their own internal time flow as they are processed in real time, continuously and autonomously, similar to the functionality of brain tissue, i.e., the equivalent of an operating system where each neural network must be processed by a neural engine while providing connectivity to the body hardware via a single type of interface, such as bidirectional Ethernet.

9. The sum of all the processors, neural networks and their connectome will form the digital brain with an Artificial Nervous System, ANS, similar to that of a biological brain, where the number of neurons remains the same from the beginning to the end of life; the only thing that changes is the accumulation of knowledge through one's own experience, adding new connections between cells without any outside intervention.

10. The ANS begins like any newborn, namely with a minimum of predefined knowledge known as genetic inheritance that is required to keep the body and brain alive, which may include physical skills such as locomotion as well as mental skills such as teaching others and learning from others.

Can this be cognitive AI? Does it contain enough intellectual features of COGNITION?

Well, if so, that's exactly what our company is developing, a software package for FPGAs called strong AI as a digital brain with an Artificial Nervous System that uses this cognition for various types of hardware, such as buildings, vehicles and robots in different environments, e.g. terrestrial, naval, aeronautics and even in outer space.

#neuralnetworks #artificialintelligence #artificialintelligencetechnology #cognition #cognitivedevelopment #intelligentbuildings #autonomousvehicles #robotics