Differences Between People and Cognitive Substrates in Understanding Language


The great AI debate is an argument that argues that machines or computers are capable of true intelligence, with those arguing that they are and others arguing that they are not. Therefore AI stands for artificial intelligence. This debate has never been resolved and it is doubtful if it will. Those that believe that AI is possible believe that the brain and the computer have similar computing powers and that one day it may be possible to create a computer that functions like a brain and exhibits intelligence. These supporters believe that the brain can be recreated using formal mathematics, though it is quite difficult for many to believe that brain function and behavior can be reduced into several algorithms.

Those that support the weak AI theory, in that machine intelligence cannot occur argue that machines that appear intelligent are simply following orders. They do not have any knowledge or intent about the action that the task is accomplishing. This theory was proposed by the AI scholar John R. Searle and is extremely influential for the weak AI stand. The other argument that comes into play is the existence of the soul and that machines cannot possess a soul. Though most AI supports for both the strong and weak side believe that intelligence and brain function is purely physical.

Computer “Deep Blue” Defeats Human

Intro to Artificial Intelligence

Much of the AI debate is considered with the algorithmic computing and if currently the right algorithm exists for intelligence. The current issue with the algorithms is that the structure of the majority of the world’s computers is based on what is known as the Von Neumann “stored-program” structure. This structure is very different from the brains neural structure. Therefore in order for a computer to be constructed with intelligence the structure needs to be similar to that of the neural structure of the brain. This is why currently the prevailing views are for the weak AI theory in that artificial intelligence cannot exist.

Human cognition and the way the brain functions and makes decisions have been studied for many years, resulting in several different types of mental processes. The main processes include making decisions, solving problems, understanding languages, producing languages, remembering and attention. Computer science, linguistics, philosophy and psychology are all used to investigate and analyze cognition. One aspect of cognition is the social aspect as children that are exposed to languages will learn them while those that are not will not learn languages, known as emergent behavior. Therefore the human brain is set up so that it needs input from the surrounding world in order to learn. This is why education is very important to human development.

A very basic structure of the human brain starts with the individual experiencing the surrounding environment. From this environment the individual stores visual, touch, movement and auditory information. This sensorial information then moves into the perception and this perception can then go into working memory and long term memory. The different processes occur in recognition which includes problem solving, recognition, decision making and reasoning. Anything in working memory can go into and out of long-term memory. Also working memory can lead to the activation of a response and then the execution of the response which results in an output. This output is like feedback and can then circle around to influence the environment. Besides this large loop with branches emotion and attention can influence perception, working memory, long-term memory and response. As can be seen by that explanation brain functions are very complex. The main storage facility of the brain is the long-term memory and information can move from working memory to long-term memory by rehearsal or repetition of tasks and can be recalled when needed or be forgotten.

Currently the AI model is not nearly as complex as the human brain. In fact AI can best be described as a reflex agent. The AI perceives the environments through a range of sensors. This then tells the AI what the world is like now and then moves into determining the appropriate action. This action is chosen based on a set of conditions or rules set in place that is something along the lines; if X happens then do Y. The action is then performed by actuators, which then changes the environment. So in order to have an AI that works like a human brain several different loops are required in one computer, with each loop interacting with all the other loops and influencing the loops. Additionally using a simple, if-then rule, does not necessarily apply as the possible actions from perception does not have only one or two outcomes, instead something like if X happens then A, B, F, H, T, and K may occur.

It is quite interesting that the AI debate is so controversial and well debated because the world does not have the technology or computing power to come even close to creating computer that functions like a brain. Therefore many scientists simply do not worry about it as it cannot be proved, so why worry about AI existence at the present. Wait until the AI can actually be proved or disproved.

Use of AI for Legal Evidence

Applied Artificial Intelligence

Artificial Intelligence Hasn’t Peaked