Understanding The Classification Of Artificial Intelligence

If I ask you one simple question, “Who would you want to be friends with in the future?” what would your answer be? While you think over it, let me tell you who I want to befriend. My future friend will be able to perform tasks better than any other human and will be able to understand and simulate my emotions. Furthermore, it would have the capability to address any situation at hand with an uninfluenced strong decision-making power with an unshaken sense of purpose to fulfill its roles. Looking as impeccable as any other of my human friends, my future friend would indeed be an artificially intelligent robot! However, just as good friends are hard to find, amidst a crowd of different people, things might be the same with robots. Branches of artificial intelligence span out in every direction, some of them grasping the role of imitating humans better than others. Thus, just like people, even robots come from different backgrounds and are divided into different categories. Artificial intelligence can be classified based on its level of innovation and its functionalities.

Also read: Artificial Intelligence: The Birth and The Growth

According to its level of innovation, artificial intelligence can be categorized into three broad categories:

• Artificial Narrow Intelligence (ANI)
• Artificial General Intelligence (AGI) 
• Artificial Super Intelligence (ASI)

1. Artificial Narrow Intelligence

Artificial narrow intelligence, as the name suggests, focuses on a narrow range of tasks that the machines are designed to do. They are not instilled with the level of intelligence that allows them to transgress beyond the pre-programmed tasks, assigned to them. Though these machines can use their human-like capabilities of conversing, mobilizing, navigating, and comprehending emotions, they cannot make and follow their own decisions to do anything they are not required to do. As surprising as it is, with so many advancements in the field of technology and robots, every kind of nearly intelligent machine ever made to imitate humans, falls under this category. However, there might be examples where these machines can do their narrow-defined tasks better than humans. One such example is of the chess playing software, which was able to defeat the reigning chess world champion, Gay Kasparov. The computer software was fed with many different combination of responses to a wide variety of inputs, and outshone its component by using those different combinations to generate new moves, never before seen in history. Kids learning to play chess, are now taught these moves. Artificial narrow intelligence has integrated itself to higher levels of creativity with its obsession to be more like humans. One might be left surprised if he has had the pleasure of interacting with Apples’ Siri or Google Assistant. A person might even doubt if these machines fall in the most basic of the artificial intelligence categories, as they show off their ability to answer just any question asked. While, chatbots like Apples’ Siri, may give you accurate answers regarding the weather situation or by soaking up data from Google, at WeblineIndia, you can get your chatbots developed for various platforms, and get them tailored to suit your different needs. The experts at WeblineIndia, know how to expand the scope of artificial intelligence and inculcate it in the daily lives. Artificial Narrow Intelligence has equipped these experts to develop knowledge-based systems, which utilize the knowledge of genius human minds stored in it to make informed decisions.

2. Artificial General Intelligence (AGI)

Artificial General Intelligence aims to dissolve all gaps between the cognitive abilities of computerized software and an intelligent, sentient human being so that the two react to any circumstances similarly. It is a dream, which the researchers and developers are working towards achieving, however, it would take a lot of time before this dream takes the form of something material. Artificial General Intelligence refers to a class of machines, which could combine human-like abilities of flexible reasoning and logic, with their computational gifts of processing, summarizing and presenting information in a matter of seconds. Fanatics of growth in artificial technology, like us, at WeblineIndia are working and waiting impatiently for this dream to come true.

3. Artificial Super Intelligence (ASI)

Artificial Super Intelligence is the talk of the future. A product of the imagination that runs wild. It refers to an era of technological evolution, after artificial general intelligence, where machines have a higher level of intellect than humans themselves. A renowned philosopher from the University of Oxford, identifies artificial super intelligence as “any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest”. Surpassing the most genius of human minds, super intelligent machines will have the advantage of perfect recalling, making informed independent decisions based on facts and figures and multi task in ways that are beyond the potential of biological entities. However, a school of thought caters to the idea, that this super intelligence would be achieved when humans themselves will greatly modify their own biology.

Though talking about artificial general intelligence and artificial super intelligence may seem like a fairy tale narration, there was a time when even artificial narrow intelligence was thought to be out of human reach too. Anyhow, the White House seemingly skeptical about the practicality of the latter types of artificial intelligence, suggested that the world won’t be able to see machines “exhibit broadly-applicable intelligence, comparable to or exceeding that of humans” in the next twenty years.

Related: What's Possible With Artificial Intelligence for Your Business

Apart from differentiating the types of artificial intelligence on the basis of the level of intelligence instilled in the machines, the classification can also be done depending on their functionalities. There are basically four types of artificially intelligent machines, based on how they in particular they are able to imitate humans. These four types include: “REACTIVE MACHINES”, “LIMITED MEMORY”, “THEORY OF MIND” and “SELF AWARENESS”.

• Reactive Machines

Reactive machines are the most primal form of artificial intelligence. They encompass the ability to respond to different stimuli. However, this ability is not enriched with being able to modify their responses through experience. They do not have the memory to “learn from past interactions” and implement that knowledge to deal with the ever-changing situations. These machines are only essential to react to a narrow range of input combinations. Doing what they do best, these machines cannot stray from their pre-programmed responsibilities, or even do these allocated tasks better, as they do not have the flexibility to improve through practice. The best example of a reactive machine is IBM’s Deep Blue chess player. Programmed to defeat its human competent, Deep Blue outshone its rival through being able to identify the game and its essential features, understand the moves from his opponents, predict their next moves, and then astonish the other player with its own unexpected move. Not knowing the world beyond the chess board, it plays with respect to the present situation, and does not cater to what happened before or will happen after the game of chess. At the heart of every software built by WeblineIndia, has been a feature of “reactiveness”, however, the technology has improved and so have we.

• Limited Memory

Limited Memory is concerned with the machines possessing the power of retaining a past event or previously learned data and information, in order to influence better decisions. They are a more advanced version of the reactive machines and since their invention, the reactive time of the reactive machines has dropped, significantly. Most of the existing inventions under the banner of artificial intelligence belong to this category. WeblineIndia has invested much of its talents and efforts to contribute to this branch of artificial intelligence. Pattern and Image Recognition applications are designed to produce images and identify patterns in it. It is also able to detect objects whether living or non-living with its Object Detection Feature.

• Theory Of Mind

Theory of mind is a concept derived from empathy, which is an abundant trait in humans. Theory of mind, as a result, refers to the decision-making ability exhibited by humans and is proposed to be inculcated in machines, in the near future. However. The decisions made by these machines, would depend on understanding, retaining and remembering emotional stimuli. The information would help the machines to respond to these stimuli. Having the fast-learning capabilities, theory of mind, would allow the machines to adjust, and appropriately respond to the fast-changing human behavior.

• Self Awareness

These now hypothetical machines would be as intelligent and as conscious as the human mind. They would have the ability to think for themselves, understand their own emotions and feelings as well as of others. They would be aware enough to have its own desires, recognize their inclination towards these desires and pursue them. Machines having this ability would almost be the next of kin to man. They would be an equivalent race to human kind, created by the humans themselves! 

Looking for AI development services? Look no further. Our AI services that can help implement and scale the impact of AI to reinvent your organization’s workflow.