Research & Development Focus

Some of the scientists of semantic system ag are doing research since over 20 years in various areas of neural and mathematic disciplines. The broad focus is needed in order to get groundbreaking results. Since 4 years the scientists focus in semantic system ag for the big goal to crack the neural code and to understand how the biologic brain processes and stores information. Enclosed a non binding list in which areas scientists of semantic system ag are doing research.

Research of the Brain Functions

The goal of our research was to emulate as realisticly as possible the ability of the brain to think, for creating the same ability in a computer chip. By understanding the functions of the elements and pathways in the neocortex the chip now has the capability to process similar complex analyses and problems as humans. For generating the correct results it is essential that the chip has access to a vast knowledge base. The intriguing point is that we learn to understand, how much knowledge is required for the chip to come to the correct decision or to gain new conclusions. This research is just at the beginning.

Cracking the Neural Code (the genetic Algorithm)

The neural code refers to the thought process in the brain. To date, science has endeavored to understand this process by analyzing the reactions of the brain and the given neural structure, and then recording the communication stimuli. We believe that this method of research will not culminate in a complete understanding of brain function. In practice, the brain structure does not represent a fixed information pattern. Similar information patterns, in different brains, may be represented in totally different neural structures. These structures can virtually change with every input of data or even previously known data can be reorganized. For this reason, it is imperative we understand the neural code (otherwise known as the Genetic Algorithm).

To correctly understand the underlying mathematics of the thought process, we have taken a different approach. The secret is to find and understand the genetic algorithm. This algorithm not only describes why and how the cells (neurons) fire, but also successfully interprets the actual meaning of the stimuli.

Building a Biologically Compatible Holosemantic Meshwork (HSMW)

The holosemantic meshwork (HSMW) is the key to building a compatible neural net. However, there is more to take into account than constructing the neurons (cells) that are similar to the biologic ideal. The inherent ability to be self-organizing and self-learning is critical. The HSMW is not intrinsically intelligent, as it simply represents storing and processing information. The real intelligence appears in the interaction between the algorithm and the stimuli (data), via the communication (dialog) with the HSMW.

The Thought Process Enables Intelligence in Data Storage and Data Processing

The combination of the HSMW with the genetic algorithm allows the simulation of the biological thought process. In addition, the simulation of phantasm (imagery, imagination) and extrapolation (estimation, drawing a conclusion) are enabled. Science and commerce are now able to use our new processors to emulate the actions of billions of biological cells and form a model of the biologic brain. Intelligence can only be enabled; it is not a given static structure or ability. Correctly stated, intelligence is the product of the interaction between the brain (neural net and algorithm) and the environment (stimuli or data input). In other words, real intelligence is the ability to correctly interpret the meaning of the data that is inputted, stored, and processed.

Understanding Neocortical Information Processing

Neurons (cells) follow a specific set of rules as to why and how they fire. The genetic algorithm, in combination with the HSMW, is the key to building intelligence in computing. The semantic system ag chip, with the simulation of the biological thought process, is now reality.

Application fields for the processors

The ai-one™ chip should be understood as a complementary addition to current computer technology and not as a competing replacement for existing processors. In addition to the processor features already described above, they can be programmed with mathematics and algorithms. The chip allows processing of all Boolean or semantic commands of the developer.

Features & Benefits

Intrinsic information detection (pattern concepts)
Recognition of semiotic (study of signs) relations and meaning in any data format
Facsimile information storage
Automatic knowledge andd information capturing

Applicable Branches:

Biometrics (pattern and concept recognition)
Forensics (profiler)
Archiving (semiotic and associative data bases)
DMS/Workflow (automatic classifications)
Information matcher (comparison of content)
Data mining/search (semantic search, associative brainstorming, semiotic conclusions)
Knowledge management with autonomic knowledge formation
Intelligent computing (robotics, dialog systems, chat robots, intelligent events)
Many other branches

Technical Features and Benefits:

Low power consumption
Almost no heat development (no cooling units required for the high speed ai-one™ Chip)
PIM (process in memory): The whole process is dynamic and self-controlled, storing and processing happens directly in the ai-one™ chip processor