Abstract
This paper seeks to examine what constitutes artificial intelligence, the principles behind it, its application, and the cases whereby artificial intelligence has proved to be a success. On a broad scale, this paper finds that artificial intelligence is a relatively new field, but its outcomes are promising. Creators of artificially intelligent systems aim to make machines capable of making independent decisions, and artificially intelligent objects heavily rely on agents to be able to make the independent decisions. Artificial intelligence is widely applied in various fields such as medicine, banking and the video gaming industry among others. Future forecasts prove that artificial intelligence will be important, and this calls for researchers to come up with artificially intelligent systems capable of addressing the problems posed.
Introduction
Artificial intelligence (AI) is a branch of computer science which deals with the study and design of intelligent systems; to a great extent, AI confines itself to the design and production of intelligent computer programs. The history of AI dates back to the period preceding the middle ages in Europe. During this period, people gave myths about some artificial beings endowed with intelligence; according to the myths, gifted craftsmen created artificial beings which could perform some limited tasks to the same perfection as humans. As time went by, classical philosophers gave scientists something to think about by suggesting that the human thinking process involves manipulation of mechanical symbols. Buoyed by this assertion, scientists tried to produce programmable digital machines, and this culminated into the production of the first digital computer in the 1940s.
In the summer of 1956, several scientists met at a conference in Dartmouth College and resolved to make the field of AI research operational. The conference pioneers were made into the first AI leaders, and given the task of charting AI research forward. The AI research leaders promised to make machines as intelligent as human beings in a few years, and this encouraged the US and British governments to pour their money into AI research. However, by 1973, the AI research had not made much progress; this prompted the US and British governments to stop funding artificial intelligence research. Despite these challenges, AI has made some significant accomplishments in the recent years and some of the successful projects have been made into commercial products.
Principles of Artificial Intelligence
Artificial intelligence integrates several fields of study such as linguistics, computer engineering, philosophy, mathematics, economics, neuroscience, psychology and cybernetics. All the fields are essential in designing intelligent systems. For example, knowledge of cybernetics is useful in designing machines which can operate on their own, while the knowledge of mathematics helps to create machines that follow certain rules and come up with solutions even with uncertain information. However, the backbone of artificial intelligence lies in the use of agents. Agents are computers systems which are capable of making independent action within their environment in order to meet the design objectives. There are five categories of agents which are classified according to their complexity. These are: simple reflex agents, goal-based agents, learning agents and utility agents.
Some well-known agents include robots and thermostats. The agents can only make independent action(s) if they can perform without the intervention of humans or other systems. The agent receives sensory input from the environment, and works on the information received to produce an output/action that meets the design objectives. In most cases, the agent is equipped with sensors for perceiving the environment, and actuators for taking action on the environment.
The action taken by an agent depends on its perceptual history, knowledge of its environment, and the actions available for deployment. All these factors determine the performance of an agent, and its degree of success.
Application of Artificial Intelligence
Artificial intelligence is useful in the fields of finance, medical diagnosis, manufacture of robots, trading in the stock exchange and production of toys among other fields. For instance, financial institutions use artificial intelligence to organize their operations; some of the systems used in financial institutions have the capacity to detect fraudulent activities and flag them for human investigation. It is essential to note that banks such as Lloyds Banking Group use an automated AI customer support system to assist their customers in checking account balances and retrieving forgotten passwords. In hospitals, AI is used to help in generating digital images. AI systems make digital images taken using computed tomography conspicuous, and this highlights possible diseases such as tumors.
Artificial intelligence has also been successfully applied in robot technology, whereby robots are programmed to perform tasks which humans consider too risky. When appropriately programmed, robots can be useful in performing tasks which humans can cause accidents due to a lapse in concentration. The use of AI in the production of toys and games either for leisure or for education is well-documented. The artificially intelligent toys and games help to introduce children to the world of artificial intelligence. In most cases, the user-machine interface is quite simple, and this makes it easy for the young generation to use AI systems.
Advantages of Artificial Intelligence
- AI systems are useful in undertaking several types of jobs. Some AI machines can perform complex and stressful tasks within a short period of time, something which may take humans a long time to complete. Undoubtedly, as the level of AI increases in the future, so is the capacity of AI systems. Consequently, AI systems would be more efficient in performing their tasks.
- With AI systems, there is no risk of harm. AI systems are emotionless. When they an AI system breaks down, there is no risk of harm. Therefore, AI machines can be used in dangerous expeditions which would be too risky for humans to venture. This property of AI machines enables them to be used in performing some tasks with less danger.
- AI systems are useful aids for people with disabilities, the elderly, and the children. For example, AI systems have been useful as teaching tools to children. Also, people with hearing and visual disabilities have found these systems useful. AI systems are also useful in alerting people about a crime, or in alerting people of a fire break out.
- AI systems have been helpful to mankind in discovering the unexplored world. Robotic technology has been deployed to explore the outer space and other unexplored landscapes, something which would have not been possible a few decades ago. For example, recently, NASA sent an AI system to mars. This would have been a dangerous mission for humans because nobody has ever set foot in mars.
Disadvantages of Artificial Intelligence
- Over reliance of AI systems may make humans less useful. The use of AI systems to perform even the simplest tasks can ruin peoples’ lives if something goes wrong with the AI system. Also, because machines can perform some functions better than humans, this will make people lose their jobs, and render many people jobless.
- AI systems lack the human feel and, therefore, cannot replace human beings entirely. AI machines cannot offer emotional understanding, or the feeling of togetherness. Due to the lack of empathy, AI machines can act irrationally because they do not sympathize with your situation.
- AI systems are subject to misuse. If they fall into the wrong hands, this may result into cases of mass destruction.
- Lastly, the use of AI systems faces resistance from some people who argue that the use of AI is ethically wrong. Opponents of AI technology contend that attempts to replicate human intelligence are akin to “playing God”.
The best examples explicating the application of artificial intelligence can be found in video games. In the recent past, most of the video games coming up incorporate artificial intelligence into their technology. Therefore, video gaming technology provides a platform for examining the theoretical principles of AI technology, something which can not only be beneficial to beginners, but also to those interested in AI research.
A development team at the Digital Media Collaboratory in the University of Texas developed a video production technique which utilizes artificial intelligence. The process was successful, and yielded results which demonstrate some means of combining the two domains together. The project, known as NeuroEvolving Robotic Operatives (NERO), brought together students and programmers who designed a prototype game based a novel neuroeveolution method. One thing which the team could be proud of is the fact that the project is replicable, and the tools and ideas applied in the original project can be implemented in similar projects.
The initiators of the NERO project, pressured to produce tangible results, were motivated to create an application that would demonstrate the use of state of art AI in the video gaming industry. The team managed to use standard software development models, with each model used achieving some degree of success. The standard software development methods used were: waterfall method, spiral method and the incremental method.
Before initiation of the NERO project, many stakeholders in the video gaming industry had identified the need to use machines with an improved processing power as the next phase of development. Video players were also demanding improved artificial in video games. This motivated a Ph. D student, Kenneth Stanley, to propose the creation of a video game which uses the concept of NeuroEvolution of Augmenting Topologies (NEAT). After Stanley introduced the idea to a workshop held at the University of Texas, it was agreed that a prototype based on the NEAT concept would be developed.
After encountering some challenges when trying to use the waterfall method, the team adapted to the spiral method. Before implementing the NEAT AI technique, the team created some experiments to validate NEAT capabilities, and this process went on successfully. However, several modifications had to be made on the 3D engine in order to accommodate the NERO game design. In the end, the NERO project team created a playable video game prototype, which made the project a success.
The NERO game requires beginners to train a group of ignorant robot soldiers in the initial stages of the game. The player sets the objectives for the group, and the robot soldiers learn how to achieve the goal. The game allows the player to customize the training area, and this exposes the soldiers to sophisticated challenges with time. After attaining the level of sophistication required, the player then saves the team, and battles can be organized between the player’s team with a human or computer opponent team. Artificial intelligence was incorporated when sensors had to be applied to enable agents to understand their environment such as walls and enemies. The sensors equipped were called egocentric sensors. The NERO video game received favorable ratings and opened doors for further research into video technology.
Most importantly, the NERO project provides an example of the existing technologies which incorporate the use of AI in their systems and the potential avenues that exist to bridge commercialization of artificial intelligence technologies. Other potential applications of artificial intelligence exist outside the realm of video technology, and researchers and investors can take advantage of these opportunities to implement other technologies which make use AI.
Conclusion
Ultimately, more innovative AI techniques will be implemented even outside the video gaming technology, and this will benefit AI researchers as well as consumers who use artificial intelligence products. There are several opportunities which exist to create artificially intelligent systems, and more efforts should be made to realize this. In the near future, humans will need to use machines which maximize their effectiveness and perform their tasks to near perfection. Artificial intelligence will be useful in this phase, since past experiences show that AI systems have been effective in addressing some of these prevailing challenges. Above all, the future of artificial intelligence looks great and the expectations are also bound to be great.
Bibliography
Gold, Aliza. "Academic AI and Video Games: A Case Study of Incorporating Innovative
Academic Research into a Video Game Prototype." Symposium on Computational
Intelligence and Games. Piscataway, NJ: IEEE, 2005. 2-8.
Harris, Michael C. Artificial Intelligence. London: Marshall Cavendish, 2005.
Kumar, Ela. Artificial Intelligence. New Delhi: I. K. International Pvt Ltd, 2009.
Negnevitsky, Michael. Artificial Intelligence: A Guide To Intelligent Systems. Boston : Addison
Wesley, 2005.
