Why Artificial Intelligence (AI) is a Fallacy

0
1837

Man’s concept is believing that he can create awareness in a hundred years when God took billions of years through natural order to create human awareness. The theological discussion around evolution and creation is an ongoing affair. There are two sides to the argument, those that believe God literally created the universe and world in seven days and those that understand that billions of years of natural creation are the work of God, for a billion years is like a micro-second to infinity.

So to the issue with artificial intelligence, or AI. The concept that machines are intelligent, or will develop awareness is so far from the truth that it is about as ridiculous as believing that the world was literally created in seven earth days. The core of the misconception is our understanding of the hardware and software combination that is now used to develop computing systems or logic circuits.

The reality of so called “AI” is far from the fictions of a fevered mind; software developers and academic research can produce the fastest logic machines. They can process phenomenal amounts of data to complete complex algorithms. No matter how much programming you negotiate and how many tests, combinations and permutations of data you allow the computer to process, it is not learning. It is only extrapolating data based on a query set. It might even be taught to ask questions based on pre-defined question basics, but you have to remember that the computer does not “understand” what the letters and words are any more than it understands what a tree or an emotion is.

Language as we know it is only a set of binary code to a computer. All programming languages translate the data created by programmers into a series of binary notes, or bits, and through programming selection, using pre-defined logic circuits in the hardware, can a computer distribute the data in a way that can calculate an equation. For a computer, the process of processing data for 1+1 is the same as asking a complex question about the weather. It is based on a series of pre-defined data sets, hardware logic gates and software algorithms used to process the data from combinations of input to binary and then back to combinations of output.

Chess is a classic example of machine logic, not intelligence. Any move can lead to several alternative moves; the limitations are based on the size of the board and the rules of movement. A computer does not analyze the game; it calculates all the probable moves, basically playing a game against itself, and then moves the piece to what it considers is the best logical solution. It might even analyze the movements of the opponent to gauge where the opponent is “planning” on moving next, since the target of the game is to “check” the king, its logic is based on how to attain this target. There is no intelligence involved, only logic circuits manipulated by software to speed up the analysis and statistical probability process as well as computed combinations and permutations based on competing moves for continued play.

Computers that ask questions do not understand what they are asking or even why they have no comprehension. They are programmed to deliver specific sets of data based on pre-defined inputs. No matter how advanced we become, a software/hardware combination will not be able to process comprehension; it will not be able to understand the “meaning” of a tree. This is not the concept of a tree, but the meaning of a tree. When a human sees a tree, it has a meaning based on years of development, environmental influences, and comprehension through emotional negotiations with trees and the personal interaction that includes trees in the life of the individual. Since there are billions of individuals, trees have billions of “meanings” and as such humans comprehend trees and understand the concept as well as the meaning.

When applying software programming to language questions, all the programmers have done is input a data set of the language, the alphabet, the words, and the thesaurus as well as a dictionary. They added a data set of combinations using grammatical clauses and logic. The hardware is a set of logic circuits connected together and with added memory management circuits and data storage medium. In other words, data is computed through the manipulation of bits (electrons) in an electronic circuit, stores and moved around to produce a result. Or simply put, the abacus of the modern age is a little bit more complex in design, but the mechanics are identical. The computer does not “understand” what it is “reading” or what “language,” what it does is it can calculate the different combinations of code within the data set based on the pre-defined commands and rules that are set by the programmer. The computer can then logically assess the correct answer to a question based on these inputs. The real test is to ask the computer to create new sets of questions based on the data set, which can lead to some very interesting results, especially when you have to associate abstract concepts with physical representations. Such as “why is a leaf green?” Instead of “why is time red?”

When it comes to mathematics, physics, and other exact sciences, computers can provide better self-learning or questioning results. Where the concepts are in fact rule-based, such as the laws of dynamics or the equations that define the way electricity is manipulated through matter. These are pure mathematical equations where physical aspects can be reduced to rules. A computer can use the take into account all the other rules and observations that can affect an open system. Such as computing the effects of UV rays penetrating the earth electromagnetic field, it does not need to know what the sun is or what the EMF is, all it needs to know is the speed and variance of UV and the mathematical properties of the EMF. In order for a reliable result to be reached, the computer must factor in other effects such as the earth magnetic field, cosmic rays, and physical bodies. These would be added to the equation to create alternative results. A computer cannot make up the information; it has to be inputted into its data banks, only from these databanks can a computer calculate the data through a pre-set list of equations and algorithms.

The autonomous car will use a computer system that processes navigation together with outside sources of traffic and proximity of objects, with moving and standing to define a safe path from point a to point b. There is no AI in driverless cars; there are only logic circuits assessing pre-defined sources of information including those that it receives through various sensors. The fact that the computer can “see” an oncoming car does not make it “comprehend” the car, it only assesses the presence of the physical object as a source of danger and must be avoided when the “car” shows signs of proximity to the computer based on a matrix of data.

Computers can be programmed to assess patterns, this is done through observation, so if a dog barks at every squirrel, a computer will record this data into a pre-defined set of data that includes a squirrel, a dog and the sound a dog makes every time a dog see’s a squirrel. The way a dog see’s a squirrel is through direct eye contact, so whenever the dogs head is turned to allow a squirrel into its field of vision, the computer will observe and record. The algorithm will allow the computer to assess if there is a pattern, how the pattern occurs, and it can then provide a logical assessment of a pattern as observed and recorded. It will not comprehend the dog, the squirrel or the bark. These are just bits of information that are defined in the logic circuits.

A quick word on neural networks. These are complex circuits and software designed to process information the same way a brain does. However close they might come to the real thing, will not make logic circuits intelligent. They only engage the system in the mechanical process of how a biological brain organ works. Creating complex architecture will not bring logic circuits closer to intelligence, it is like building a house to look like a tree, but the house will remain a house and never be a tree.

Bottom Line: Maybe in the future humanity will develop a synthetic or evolutionary awareness, if it does it is not artificial. Humanity is a product of nature, and anything a human does is natural. The unnatural is something that is not natural, and since everything in this universe is a product of its existence, everything a human does is by default (logically) natural, and is a part of the evolutionary process.