The Ai Function

A relationship between objects can always be written, that is, everything that exists has some relationship with everything else. Whether that relationship is definable or not is only limited by other parameters. That being said, the key to any artificial intellect is to design the system with the ability to take any set of objects in any space and in short order write their true mathematical relationship. 

Most of the modern scientific community approaches Ai concepts using forms of logical schemas such as regular expression. The draw back to this approach is that it is not truly object-oriented logic and generally falls short of defining a system. It is cumbersome and much like trying to teach a language to someone without using pictures or sound.

We tend to think of intellect as overwhelmingly difficult since we do not understand the foundation. Any intellect starts with two or more objects and, depending on its depth, writes their logical relationship by observing their place in the same space. In animals, this involves a set of parameters such as sensory perception ( taste, sound, smell...), weighted importance, and synthetic creation. The later being unique to humans at this point in time.  Feelings and emotion are not a consideration in the primary design of Ai.  Once a true Ai system is built, adding advanced concepts are "icing on the cake", so to speak.  A fundamental Ai system involves memory, process and synthetic projection.  

Most scientists believe synthesizing is a much more advanced concept when developing Ai, but it may surprise even the best designers that the most basic Ai can predict events far more accurately than the average human.  This is because we think in linear fashion, where, there isn't probably ten people on earth that can permute past the second degree in their minds.  However, Ai can permute a virtual unlimited set of data with no error, in milliseconds.  There is a key to this process that did not exist until the advent of advanced processors.  Coupling computer software drivers with advanced mathematical concepts allows the Ai system to virtually formulate independent thought.

It is a fundamental process to take objects and continually update their relationship in the human intellect. This is done without scripting ( equations ) in humans by weighted pointers. For example, a child does not need to define grammatical structure in order to speak his native language. Why? Because they acquire their object oriented relationship through weighted pointers. These logical pointers, if viewed, would look like a maze of interconnected weighted links. Each link would point to an object and relevance would be their key.  

As an example, if you are asked "Why is the sky blue?" you're mind first associates the word "Why" as an introduction to a question, perhaps with little relevance since you were not told the importance of the question. Your ability to know that "Why" is an introduction to a question is also precluded by an array of pointers.  Next the word "is" has been pointed to from "Why" giving much more relevance to the statement since the statement "Why is..." is extremely weighted and your mind anticipates the outcome - got your attention so to speak. You are somewhat unaware of the mental processes taking place while your brain is analyzing this simple question.  Of course, it is a loose analogy to use the word "weighted" since in a true Ai system "weighted" will mean crunching data objects using advanced mathematics with supportive programs. 

Looking at an Ai process at this point shows a rather huge collection of pointers moving in all directions. These pointers are all weighted and span everything in your mental acuity. Your brain is in standby, waiting on the next word to reduce the possible solution path. With each successive word, the conclusion path is reducing and given what you do know about the statement, it may point to a single set of objects satisfying the outcome.

The point to this exercise is that you do not need to quantitatively solve this statement using logical schema in order to respond and neither does any Ai system. If the fundamental Ai touring machine where placed in a child's environment it would be only as smart as the child. However, the same machine placed in an engineering environment would expand to the level of that environment and be at least as rote smart as any engineer, have virtually unlimited memory and be able to assimilate extremely complicated outcomes in seconds.

What are the advantages to the Ai system?  Imagine a system capable of taking observed empirical data, reducing that data ( while throwing out non - related objects ) and deriving the highest possible solution set.  All the while, Ai shows the user what percent correlation has taken place in the process.  This is done in seconds and with no errors.  As an example, a cancer researcher collects pages of tested data and through exhaustive trial tries to find a correlation between events and objects.  The researcher is incapable of reducing relations beyond P2, since he or she would generally use some predefined schema or transformation technique to try and find the answer.  Ai would take the same data in its entirety and write the true equation of the system in seconds, while discarding unusable data.  The result is that cancer would become a forgotten disease in short order.  In space travel the Ai system would be able to derive conclusions of data in seconds and be only limited by its sensors.  In the more difficult Mathematical problems Ai, though not showing a discreet proof of the problem, could show the equation that satisfies the solution set.

The applications are as large as mankind for the Ai function.  The algorithm for Ai is one of my pet projects and it quite involved.  However, I have seen outstanding results in the ability of this system to predict Newtonian and Euler outcomes even at an incomplete stage.  I envision the Ai processor as a new environment that will facilitate every walk of life.