The BrainChild technology is a mathematical framework that explains and defines the functioning of the human in respect to cognitive functions and reasoning.
This technology is a mathematical concept, which, when implemented on a computer, shares many features of a Neural Network, yet it is distinctly different in many other respects. Although research in Neural Networks and other AI concept has been ongoing for many years and astonishing advances have been made, the BrainChild technology is a uniquely different concept.
While the human brain has extraordinary abilities to learn, reason, conclude and deduct, it has also many shortcomings in that it can be distracted, tired or even defective. A mathematical concept such as the BrainChild, allows one to utilize the capabilities of the human brain and to combine it with the power of the computer, while avoiding the effects of human nature.
The main drawback of this technology is that, like in the human brain, large amounts of data must be stored and processed in a short time. While the human brain has the ability to process data simultaneously, modern PCs have only one or two processors, which are generally very fast and can perform many million operations each second.
Thus, a large number of slower, less powerful processors would be more suited to execute the functions required to process data in accordance with the BrainChild concept. This is both a drawback as well as an opportunity. While it represents challenges in implementing the BrainChild technology in a conventional PC, once this technology has been fully developed it offers the opportunity to collaborate with a manufacturer of microprocessors to design a chip that consists of many smaller, less powerful processors dedicated to processing data using this technology.
The BrainChild technology differs fundamentally from Neural Networks as it "explains" how cognitive learning and reasoning takes place in the human brain and unlike Neural Networks it does not require any specific implementation. The BrainChild technology can be implemented in a neural structure, although this structure will be noticeably different from a conventional Neural Network. Although the neural structure may be an efficient implementation of this technology, it is not a requirement to utilize a neural structure. In fact, our current research uses a conventional PC and implements the algorithms in C++ language. Unlike Neural Networks, the BrainChild technology makes it possible to implement portions of the program using BrainChild technology and integrate it with conventional software to best utilize the computer's ability to perform mathematical calculations and to interface with the user.
At this time, the SmartPlace has developed the theory underlying the BrainChild technology. The full framework explains how sensory inputs can be combined to associate, combine, classify, relate, and react and how to learn relationships, rules and behavior. Only a small subset of these capabilities has actually been tested, verified and demonstrated. The development and implementation is currently a level that mimics the capabilities of a 2-3 year old child. Some more study and research is required to fully implement complex functioning and strategies, such as the capability to learn and play chess or to analyze and perform simple analytical tasks.
Once the concept has been fully verified, the BrainChild technology can be used to interact with computer users and to adjust the software to understand what the user wants to accomplish and how the user wants to accomplish a task. It is envisioned that a word processor implemented with the BrainChild technology would first be given a number of past documents generated by a specific user as well as other i.e. generated by the company or general related literature. The user will then, verbally or via e-mail, instruct the computer to generate a letter concerning a specific subject, together with all relevant information, whereupon the software using the BrainChild technology will generate this letter in a style compatible with the user in a format acceptable to and compatible with the employer's or other applicable standards (legal briefs, references, etc.).
Similarly, the task of generating computer software is greatly simplified since only the desired function and the desired graphical user interface needs to be defined. In a development environment powered by BrainChild technology this definition can be in a free format as long as all the relevant information is provided in a logical fashion. Thus, developers can focus on the task of developing applications, rather than to be distracted by the task of programming. Consequently, the process becomes more efficient, more productive and more fun, allowing for the development for specialized, user friendly, cost-effective applications even in markets where few users exist.
