Md. Sadat Hussain Rafsanjani :
Artificial intelligence means any sort of intelligent device or a computer that can mimic the functions of a human brain. It is also attributed sometimes that any expert systems which can make decisions on its own by analyzing various criteria like a human brain have a self-awareness or singularity. Therefore, a true artificial intelligence having machine can behave like a human! Aristotle, the famous Greek philosopher, first invented the concept of logical reasoning 2400 years ago. Before Karel Chapek, Leonardo Da Vinci sketched a mechanical man in 1495 although we don’t if they had intelligence or not.
Initially the term was first coined by John McCarthy in 1956 in MIT. Although some games have hard coded intelligence but no computer is able to exhibit full intelligence till now. The biggest challenge man have ever faced in 1997 when Chess grandmaster Gary Kasparov was defeated by a IBM super computer called Deep Blue, an IBM RS/6000 SP2 model. Perhaps it is the only sector computer has gained more intelligence than human brain, but not the total. For instance, Gary Kasparov can evaluate three chess positions at a time where Deep Blue is capable to evaluate 200000000 positions per second. Gary can learn from his mistake and apply the acquired knowledge but Deep Blue is unable to do that. Deep Blue has little knowledge of chess but enormous calculation power while Gary has huge experience of chess but limited calculation ability.
Alan Turning, a renowned computer scientist believed it is hard to define intelligence. He devised a test called Turning test that can formalized a machine as an intelligent one. The test says if a machine can carry on a conversation without making an impression to the human that it is a machine, than it would be ruled as an intelligence possessing machine.
There are many types of behaviors we encounter or make in our day to day life. But not all of them are considered intelligent. There are specific tasks that an intelligent human brain can simulate. Behaviors like learning from experience and apply that learning, determining importance, applying heuristic (commence sense rule) and ability to solve problems when adequate information are still missing, doing creative and imaginative works, processing visual signal and understand them are some of the instances of intelligent acts we do on regular basis in regular life.
There are several types of artificial intelligence. Some of them are visional that can capture and store images to process them. A perceptive system approximates the way in which a human can see, hear or react, robotic device that can mimic the works of human but relentlessly and precisely. Although robots are far stronger than human they cannot identify an object efficiently and does normal works very clumsily. Neural networks is the most advanced implementation of computer network that can function like human brain, expert systems that can store data to interfere in due time, natural language processor that can understand the auditory or written natural language like English to interact with the user. Natural language processors are hard to build and they are not that much reactive. Even auditory and voice recognition systems are not that much expert. They need translating software to translate human voice to electronic signal still they don’t know what is being said or commanded to them. They use dictation methodology. The neural network is inspired from the study of human brain. It is a network that is massively parallel, consists of many small units and highly connected. These can be of three types: perceptron, multi-layer perceptron and back propagation net. An intelligent agent is a system that perceives its environment, learns from it and interacts with it intelligently. The agent can be physical such as a robot or logical, a software. Autonomous agent has autonomy power, social activity and reactivity.
Voice recognition can be a chief way to command a system. But as mentioned earlier, it is hard to build. Each person’s voice is unique. The shape of mouth, movement of tongue, nasal cavity and throat creates resonance when we make sound. Also, human talks in a continuous manner so that it is hard to distinguish sound alike phrases for a computer like “ice cream” and “I scream”. It is difficult to difference between “eye” and “I”.
The biggest challenge while working on machine intelligence is computer has no thinking power or emotion thus it cannot justify or evaluate the morality, ethics or legality of a job that is being assigned to it. Previously it was thought that expert systems would be the future of artificial intelligence but for the last twenty years it has been examined in every possible aspect, from normal to special. But it is proved wrong as expert systems are efficient in only special fields like scientific or medicine, they are very hard to develop as well as maintain and on the other hand quite expensive. They are very much prone to error, cannot readily deal with mixed knowledge and often raise moral or ethical concerns. Indeed, expert systems are used and implemented many scientific works. Expert systems are useful because if given certain data they can explain it by performing a self analysis over the data and draw a valid conclusion. The first expert system was developed in Stanford University in mid 70s. The research was called Heuristic Programming Project led by Edward F.
AI technologies were heavily developed keeping one thing in mind from the early stage. And that is deploying the technology in defense industry and the military use. It is a tradition for the US military to embrace new technologies from time to time. Initially DARPA and US Naval Research first approached to employ AI technologies in the force. Another example is DART; a planning tool with AI was used during the operation Desert Storm in Iraq, also known as Persian Gulf War, to control 50,000 vehicle, cargo and people. Mostly, AI technologies are used in non-combat role due to its nature of controversies and lack of ethics. It is often disputed that a robot cannot identify the friend or foe in the battle field. There was technical demise in 1980s. DARPA financed to develop a smart truck but due to technical setbacks Pentagon cut down the funding.
Currently three sectors of military advancement have been introduced with AI technologies. These includes: intelligent pilot associate, autonomous land vehicle and naval battle manager. The IPA or intelligent pilot associate project mainly projects the idea of a phantom pilot crew. The abstract concatenates four basic operations namely: situation assessment manager, a tactical planning manager, a mission planning manager and system status manager. Extra care is taken for pilot vehicle manager. This sort of technology is managed with voice command and speech recognition. The autonomous vehicle deploys computer vision and image recognition system. They need AI reasoning technologies so that they can sense and react according to the results. The most advanced one is the naval battle control manager where the controller is AI capable. In time the AI should be able to function as a human. The responsibilities includes capabilities assessment, force requirements, simulation of campaign, operation planning and strategy assessment. These sorts of jobs are very laborious and expensive to handle by manpower. Thus expert system can be built to tackle them. Such an example is SAR or Synthetic Aperture RADAR, usually found in Predator Drones. A SAR can collect data for automated target discrimination; create a height map of the surface, useable in both day and night whatever the weather is sunny of cloudy. Still now USA is funding 100 billions of dollars to develop smart soldier or robots those can serve in military and replace human soldiers on the front line. Even intelligent flight simulator is developed to train a pilot.
Conceivably the most applied field of AI is game industry. Games like HALO can use AI to deploy vehicle and team tactics. Left 4 Dead uses new AI engine called The Director that is so dynamic that it can rearrange and adjust each new level difficulties with the help of player performance.
AI technologies have been a great inspiration on the Hollywood movies. Mostly from the beginning AI has been portrayed negatively on the silver screen. The protagonists are either android or sentient computer. These films often deal with some common subjects like machine going out of control, domination over humanity or replacing humanity. Movies like Terminator exploit the theme of cybernetic organism dominating the future where human fighting for their life to survive the cut. I, Robot deployed a malfunctioned AI computer VIKI that tries to submit humanity to the robots. The story was inspired from the robot universe of famous science-fiction writer Isaac Asimov. Eagle Eye, another movie intensify the theme of AI when AI enabled supercomputer ARIIA tries to escalate a false implication of terrorism act. In Prometheus, David serves as the maintenance man on the board. It is a humanoid android possessing high level of artificial intelligence. In Arthur C. Clerk’s famous science fiction Odyssey series, the antagonist is a sentinel computer called HAL 9000. In the Matrix film series, Jewish deities are depicted in a sci-fi manner where the human are used as power source and they are given a false expression of the worldly living using simulated reality called Matrix. n
