The world is known to have experienced four major revolutions from the past. The first revolution being the steam engine invented in 1784 which was one of the most important inventions of the Industrial Revolution. The second one was the invention of electricity in 1870. The third revolution came into existence when the world had experienced the taste of Information Technology in 1969. Nonetheless, the fourth one is the dynamic one, the introduction of Artificial Intelligence which has completely transformed the entire world to a different extent. And the impact of the same will be seen shaping our next generation in a fully automated way.
Before proceeding further, let us have a look at the definition of Artificial Intelligence.
‘Artificial Intelligence is the set of various tools and programs that can simulate human intelligence through the process of supervised learning, reasoning, and self-correction.’
Generally, AI is classified as weak and strong depending upon its function and limitations. AI has three phases- ANI (Artificial Nominal Intelligence), AGI (Artificial General Intelligence) and ASI (Artificial Super Intelligence). ANI is in the form of a virtual person that provide assistance like that of Apple’s Siri is considered as weak because they are designed and trained for limited task. While AGI is considered as the strong one that can find a solution without human involvement because it is designed with generalized human cognitive abilities. Whereas ASI goes a step beyond other AI defining a time when the capability of computers will surpass the human making it more superior to humans.
Though the concept of AI technology has been into existence for centuries it wasn’t until the late 1950s. The concept was brought up into concern at a workshop held at Dartmouth College in 1956. Scientists, Philosophers, and Mathematicians around the world had this idea from generations who had attempted to explain the human thinking process as the mechanical manipulation of symbols. But unavailability of computers and hardware resources became the biggest halt for this project to carry forward. And funding was also a major issue to increase resources. By the year 1974, with the flourishment of computers that was fast, affordable with advanced features and was capable of storing more commands and data. By 1980, AI Research was accelerated at a great pace with an expansion of funds and various algorithmic tools. The use of deep learning techniques helped computers to learn through experience and act and take actions like humans.
Types of AI
AI has been generally classified into 4 different categories including existing and non-existing both because there are certain high level of AI that has been predicted but are yet to be designed-
These are known as the most basic type of AI that can neither form memories nor can retain past experiences to perform present action. In a simple sentence, these are programmed for here and now not for before and after. They are limited to perform only to current situations. The reactive machines have no understanding of the world hence it cannot go beyond and conceptualize like a human brain. These are programmed in such a way that it can be used only for activities where repetition is required. ‘Deep Blue’- a chess player computer is a perfect example of a reactive machine created by IBM in the year 1980. It was created to play chess with the intent to defeat a human competitor. It could make predictions regarding its competitor moves based on what should be its move and was programmed and designed with an ability to predict, select, determine and win. Deep Blue is the first computerized program to defeat a Russian chess player Garry Kasparov in a series played in the 1990s. Deep Blue and Google’s AlphaGo were designed for narrow purposes and because of their limitations they cannot be applied to another situation.
Unlike reactive one, these AI can retain memories and use past experiences to perform actions in the future. It comprises machine learning techniques that can learn through observance, derive knowledge and information from the past and store them to function in accordance. Some of the best examples of this type are Self-working robots or Autonomous vehicle/ Driver-less cars. Driver-less cars have the capability that can sense the environment, observe other vehicle’s movements and move safely with or without any human effort. It does so by reading their environment, detect patterns or changes in external factors and accommodate accordingly. AI-based on limited memory builds on observational data in addition with earlier programmed data contained by them. The limited memory AI is mostly used in the work of automation and work on static data such as traffic lights, lane traffic, customer service kiosks at the airport, etc.
Theory of Mind
The concept ‘theory of mind’ constitutes the decision making abilities by a robot just like a human. With the development of science and technologies, scientists have already been successful in making such robots, but the only thing lacking behind is the incapability of holding emotions and feelings equal to the extent of a human. It is a highly advanced form of artificial intelligence & machine learning that is subjected to transform the entire world. This concept has been designed by keeping in mind to create a robot/ machine embedded with feelings, emotions, understanding similar to human that can affect behavioral output. In order to act and react like a human, machines would have to be able to retain the emotional output, remember the information derived by interacting with everyone and adapt it into their learning areas to understand and respond with different situations accordingly. Some of the existing examples of theory of mind are the robots Kismet created by Prof. Cynthia Breazeal in 2000 and Sophia designed by Hanson Robotics in 2016.
Self Aware is the advancement and extension of the theory of mind. These are known to be characterized by their sense of self with a human-level consciousness. This type of AI is considered as the most advanced form of artificial intelligence and currently does not exist. The astonishing feature of self-awareness machines is that it can not only recognize and replicate actions like humans but also to have a human-like thought process, desires, understandings, and emotions.