Many people have heard the term artificial intelligence. While this term is commonly used many people are still unsure of exactly what this term means.
Table of Contents
- What is Artificial Intelligence?
- What is it used for?
- Different Classifications of Artificial Intelligence
- Elements of Learning
- Unsupervised Learning
- Reinforced learning
What is Artificial Intelligence?
Artificial intelligence often goes by the term AI. Back in the 1950s two men Minsky and McCarthy were the first people to use the term artificial intelligence.
They described this term as a task being performed by a program or any type of machine that if the same task was going to be performed by a human, the human would need to apply some level of intelligence in order to complete the task.
While this definition is broad it does cover the basics of this form of intelligence. Some people say that basic pulley machines can be considered artificial intelligence while other say that only advanced technology can count as artificial intelligence.
Many modern experts state that in order for something to be considered artificial intelligence it must meet on of the following criteria in order to compare with human intelligence.
The device has to show some sort of planning, perception, motion, and manipulation and to a less extend some form of creativity as compared to human intelligence.
Today many forms of artificial intelligence include machines and robots that can process data and increase the speed and accuracy of a task.
In some cases the artificial intelligence can identify the problem faster than a human can and many are finding this form of intelligence more useful than humans.
What is it used for?
There are many uses for artificial intelligence and the use will vary based on the type of intelligence. This intelligence is used to make a task easier.
It can be used to help a person in their daily life. This form of intelligence can also be used to help people in their place of work.
Different Classifications of Artificial Intelligence
There are two main classifications of artificial intelligence. There is what is known as the weak or the strong AI.
Strong AI is considered to be artificial general intelligence. This technology is able to compare with the average human cognitive functioning. When this AI device is given an unknown task they have as much intelligence as the average human as to where they can find a solution to this task.
Some of the ways to determine if something is strong AI is to use the Turing Test. This test was developed by Alan Turing in the 1950s. This method was developed to determine if a computer is able to think in the same manner as a human. Many still think this method is rather controversial.
There is another way to classify artificial intelligence. Arend Hintze who works in the integrative biology and computer science department as well as the engineering department at Michigan State University further classified artificial intelligence into four different types.
Type 1 is the Reactive Machine. An example of this type of intelligence is a computer program developed by IBM that was able to beat a world class chess player.
This computer was able to identify the different pieces on the board and make predictions. The computer is not able to remember information and cannot use past experiences to help with future issues.
Type 2 of this intelligence is known as Limited Memory. This system is able to use experiences form the past to help with future decisions. Some cars are being designed with this technology. They will use this information as well as current observations to make informed decisions.
Type 3 of Artificial Intelligence is uses more psychology and is the Theory of Mind. This type of intelligence is said to have its own belief system, desires, and intentions to help it make decisions.
While this type of intelligence does not current exist many people do not think that is it too far off in the future.
Type 4 of artificial intelligence is said to have self-awareness. This intelligence will have a sense of self and its own personality. Machine and devices that will be self-aware will be able to express their feelings to other.
While this type of technology currently does not exist there are some that are trying to develop it. Others find this technology to be a little too advanced.
Elements of Learning
There are different elements when it comes to machine learning and programming machines to be able to learn.
Supervised Learning is common for most forms of AI. This learning gives the machine data and is able to program it to use this information to relate it to other data.
The system needs to be trained. Once the system is trained it can take in new information and apply it to labels and related information. The training has to be facilitated by humans since the machine will be given a great deal of information to take in and process.
This system teaches the machine in new ways. They are programmed in order to group items together by any similar features that they may have.
The machine is designed to look for a specific type of data or traits. For example Google groups its new stories by related content each day.
The computers are given a reward based on its performance. This is a trial and error process until the machine finds the best possible outcome with the information that it was given.
For example the Google DeepMind network is able to use human performance to develop the best type of video game and the game that a person would find the most interesting based on their playing habits.
Artificial intelligence has been around for a long time and there are different levels and degrees of this artificial intelligence. This technology is designed to make life easier for people and will be able to complete tasks that will make daily activities easier for people.