Technology: Artificial Intelligence (AI)

We get to hear a lot of hype about AI and its applications. AI eases human life or as some believe it is dangerous. This article will give you a brief idea about AI, its application and the impact on the world.   

What is AI?

AI is a contrast to natural intelligence exhibited by humans. It is a machine which is designed to perceive the events and take actions with maximum probability of success. In a simpler word, the machine mimics human minds. AI replicates the cognitive functions of the human brain which is mainly learning and problem-solving. 

As the hype grew through many years, a term came into picture i.e. "AI Effect". It means "whatever hasn't been done yet". Some of the AI capabilities include: 

• Understanding human speech
• Competing in strategic games
• Automated cars
• Routing in content delivery network
• Military simulations

AI was founded as a part of academic discipline in 1956. It is the combination of various fields like computer science, mathematics, psychology, statistics, philosophy and many more. It was claimed that human intelligence can be replicated in machines. This claim has raised philosophical and ethical arguments. Many people consider AI to be dangerous since in near future it might create a risk of mass unemployment. A similar impact is observed during the Industrial Revolutions where humans were replaced by machines. 

Origins 

AI appeared in literature as characters. It was fiction during 1800's. The appearance can be seen in Mary Shelley's  Frankenstein. It is also seen in many other works of literature. The characters and their fate in the fictional stories have created an ethical dilemma in artificial intelligence today.

Later, around mid-1900's there were many discoveries in the field of neurobiology, information theory and cybernetics. This led the researchers to the possibility of building an electronic brain. 

The field of AI research born at a workshop at Dartmouth college in the year 1956. The founders and leaders of this research were Allen Newell, Herbert Simon, John McCarthy, Marvin Minsky and Arthur Samuel. The proposal they created for this research has following aspects:

• Simulation of machines to perform higher functions of the human brain
• Computer programming to use language
• Arranging hypothetical neurons
• Measurement of the complexity of the problems
• Continous improvements on its own
• Forming Abstractions (dealing with ideas rather than events)
• Randomness and creativity

By mid-1960's, the AI research in the US was funded by Department of Defense and many labs were incorporated around the world for this purpose. The progress slowed down in the year 1974 due to some complications.

AI got attention back during the 1980's due to the commercial success of expert systems. In late 1990's and early 2000's, AI was used for logistics, data mining, medical diagnosis and in many other areas. The success was due to:

• Increase in computational power
• Emphasis on problem-solving
• Ties between AI and other fields like statistics, economics, and mathematics
• Commitment by researchers to mathematical methods and scientific standards

In 2011, advances in machine learning and perception are achieved due to faster computers, algorithmic improvements and access to a large amount of data. Deep learning methods dominated around 2012. The current year is the landmark for artificial intelligence. 

Traits

Reasoning and Problem Solving: The Algorithm is designed to develop step by step reasoning while making logical deductions. Logical thinking is necessary to make decisions which are efficient and profitable. 

Knowledge Engineering: This is the central part of any AI system. It gathers explicit knowledge possessed by experts. The machine tries to gather commonsense knowledge of an average person into a database.  

Automated Planning and Scheduling: Most of the time automated systems work as a charm. The requirement of an intelligent system is the ability to set up goals and achieve them. AI has automation capabilities which makes them intelligent.

Machine Learning: Learning by experience is the concept used in Machine Learning.  ML is one of the fundamental concept in AI since the time of inception. It involves 3 types of learning:

• Supervised learning: The outcome is known in this case. It includes classification and regression algorithms
• Unsupervised learning. The outcome is not known. The algorithm tries to find the pattern in the input and then takes a decision 
• Reinforcement learning: The machine learns from the response to events. A good response is rewarded and the bad response is punished. It uses this sequence of rewards and punishments to develop strategies

Natural Language Processing: NLP is the ability of the machine to read and understand human language. It is widely used in the applications of text mining and machine translation. 

Machine Perception and Computer Vision: Machine perception is nothing but using hardware sensors for deduction. The applications include speech recognition, facial recognition and object recognition. Computer Vision is the ability to analyze visual input which can be used for image analytics. 

Robotics: AI is used in robotics for motion planning and manipulation. It is widely used in modern factories as industrial robots which can learn from experience to perform tasks efficiently. 

Affective Computing: These are the systems which can recognize, interpret, process or simulate human effects. In other words, it is social intelligence. It includes textual sentiment analysis and multimodal affect analysis. Understanding motives and emotional states allow the machine to make better decisions. 

Artificial General Intelligence: It is the combination of all the narrow skills mentioned above and at some point even exceeding human abilities in most of all these areas.