Artificial+Intelligence

toc

History of Artificial Intelligence
Work on intelligent machines began after World War II. The English mathematician Alan Turing may have been the first. In 1947 he gave the first lecture pertaining to this topic. ‘He also may have been the first to decide that AI was best researched by programming computers rather than by building machines. By the late 1950s, there were many researchers on AI, and most of them were basing their work on programming computers.’ (McCarthy 2004)

**What is artificial intelligence?**
Artificial Intelligence (AI) is the scientific and engineering process of making intelligent machines with an emphasis on the engineering of intelligent computer programs. ‘It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable.’ (McCarthy 2004)

In order to begin this process one must define intelligence. Although there is no one universally accepted definition of intelligence, it is found that many people have an intuitive notion of what it is. (Detterman 1006) Experts and scientist tend to use tests to measure intelligence but still fail to encompass all factors that contribute to intelligence as a whole. For example things such as ‘cleverness, common sense, practical problem solving ability, verbal ability, and interest in learning. In addition, many people think social competence is an important component of intelligence’ and not just ‘the ability to learn from experience and the ability to adapt to one’s environment.’ (Detterman 2006)

There are scholars that also argue that intelligence is defined in relation to one’s culture and whatever abilities are valued by one’s culture is what should be considered intelligent. One widely accepted definition is that intelligence is the ‘computational part of the ability to achieve goals in the world.’ (McCarthy 2004) This is a very general definition and many tend to add sub-definitions and facts to this statement.

Artificial Intelligence is usually tested by a series or a group of different tests, such as the Turing test. One dilemma is that ‘we cannot yet characterize in general what kinds of computational procedures we want to call intelligent. We understand some of the mechanisms of intelligence and not others.’ (McCarthy 2004) There are multiple mechanisms that are involved in artificial intelligence, some computers can carry out some and others not making it hard to define whether a machine is intelligent or not.

By observing people or methods tested AI scientists learn how to make machines solve problems but do not base their entire study on simulating human intelligence. Research is mostly dedicated to ‘studying the problems the world presents to intelligence rather than studying people or animals. AI researchers are free to use methods that are not observed in people or that involve much more computing than people can do.’ (McCarthy 2004)

Artificial Intelligent Systems
Artificial intelligence systems that are used most can be classified into the following major categories:

>
 * ===**Expert Systems**===
 * Also known as the knowledge-based system, this form of artificial intelligence applies reasoning to reach a recommendation/solution. This intelligence works best in scenarios where there is a need to diagnose a problem and prescriptive problems. The main focus of the expert system is obtaining knowledge from a //domain expert// and taking their knowledge and placing it into the system. By capturing the expertise of a domain expert the system can reason through their understanding (weights) to evaluate the situation and come to a clear conclusion.
 * Expert systems are wonderful when you need a helping hand but keeping in mind they only work best with tasks that can be conducted with clear and precise rules given to it. It can process an extensive amount of information, minimize human error and time thus increasing revenue (Haag). It can also assist in decision making based on the well collected information. Although there are set backs to this, information given by the system cannot defend its action thus you don’t know how the results came about. Experts also lack judgment and overall street smarts (Haag).
 * ===**Neural Networks**===
 * Is an artificial intelligence system that has the capabilities of recognizing patterns and differentiating between them. In the realm of biometrics neural networks have the ability to assist in voice, facial, body, and palm recognition. Differentiating between different people on physiological characteristics is a promising use for neural networks. Neural networks learn by example and adapt to new concepts and knowledge. Financial institutions have also used neural networks to detect fraud and find market irregularities in the stock market.
 * Neural networks can either be //self-organizing// - have the authority to find patterns and correlations on their own without the intervention of a human component. Or, they can be //self-propagation// neural networks - systems which are taught through example.
 * ===**Genetic Algorithms**===
 * Works on the principles of finding the best solution from a set of derived solutions. Genetic Algorithms attempt use three methods to find the top outcome: (1) //Selection// - is similar to survival of the fittest, only the best solution gets picked. (2) //Cross-Over// - combine aspects of two good outcomes to possibly create an even better outcome. (3) //Mutation// - the process of randomly selecting different outcomes and combining them in the hopes of creating something successful.
 * This type of intelligence is very helpful in dealing with thousands of possibilities. For example, when deciding what would be the best possible way to lay down the Wifi network across Toronto a genetic algorithm could prove useful to maximize coverage throughout the downtown core. Genetic algorithms succeed in these scenarios since they use selection, cross-over, and mutation to constantly explore endless solutions to a problem with no definite answer. They have indeed made an otherwise daunting task faster and more efficient.
 * ===Intelligent Agents===
 * "Is software that assists you, or acts on your behalf, in performing repeititve computer-related tasks." Specifically these agents fall into the following categories; //buyer, user, monitor and surveillance, and data mining//.
 * Buyer agents typically help web surfers find products and services of interest. Consider Amazon.com, the site uses shopping bot which recommends possible purchases on previous bought items. This is also known as filtering, bots tend to place customers into specific categories through different filtering techniques to better determine what you might be interested into to. The following are four buyer agents use to anticipate user wants: (1) //Collaborative filtering// - placing you into a group of people of similar interest, (2) //Profile filtering// - based on terms and keywords you have entered - paints a personal picture, (3) //Psychographic Filtering// - determines preferences based on a questionnaire you may have filled out, (4) //Adaptive Filtering// - preference is determined by rating products or taste and monitors actions over time.
 * User agents perform medial tasks on your behalf. This can anything from checking your email, notifying you of events, computer game bots, self-assembling home pages, and etc. Beyond personal use, the user agent has also expanded into the business realm as well. Companies such as Movex will use user agents to search the company intranet to find the best suppliers and distributors to go with.
 * Monitor and surveillance agents monitor and report on the status of equipment. Their main function lies with observing and reporting on system hardware, making sure crashes do not occur, and if they do a warning period prior to can be sent out.
 * Further use can be towards internal aspects of business, such as monitoring inventory levels, spending, and resource planning to keep costs down.
 * Lastly, data mining agents work in data warehouses to derive useful information. Data warehouses contain vast amounts of information from all aspects of the organization; it is the data mining agent that has the ability to extract through the vast data useful information from which employees can use.

Artificial Intelligence in popular culture
Artificial Intelligence has been seen in many cultural artifacts, such as television and film. Artificial is often represented as a super-computer, operating at higher capacities than normal computers, or as a sentient intelligence that is on its way to, or to have already achieved conscious thought.

Examples of this include:

- **HAL**, in 2001: A Space Odyssey. HAL is a computer capable of speech/face recognition, emotional response and reasoning. //HAL 9000

-// **The Terminator** (Arnold Schwarzenegger), in //The Terminator//, and //Terminator 2: Judgment Day//. The Terminator is part of a larger, sentient system known as SkyNet. He is appropriated by the humans that SkyNet seeks to destroy, and is sent back in time to protect the future general of the human resistance. He is capable of reasoning, conscious thought, a slew of military applications (including nightvision, battlefield medicine, and being really strong), and also has the ability to ride a Chopper at night while wearing sunglasses. //The Terminator//

//-// **The** **Universal Soldier**, in //Universal Soldier//, and the lesser known (and less well-received) //Universal Soldier 2 & 3//. The Universal Soldiers are a group of soldiers that were previously dead. The US government resurrects the soldiers with erased memories and controls them via radio command. The Universal Soldiers possess superior calculation abilities, strategic decision making, and physical strength in comparison to their human counterparts. In the end, the Universal Soldier group attains sentient knowledge, and the ability to express emotion once again, and this spells certain danger for the humans who control them. (An interesting sidenote: Jean Claude Van Damme and Dolph Lundgren both have more emotion when they are non-sentient Universal Soldiers. As soon as they become conscious again, they go right back to having terrible acting skills.) JC Van Damme sporting a headset as GR44