Types of Artificial Intelligence: Introduction
This article will provide an overview of the most important types of Artificial Intelligence. Artificial Intelligence is designed to allow machines to perform human-like functions. The primary criteria for determining whether AI can replicate human-like actions is what is used to classify it. There are two main types of AI classifications. Both can be based on the ability of AI to reproduce the human brain. The first, “Based on Functionality”, categorizes AI based on their similarity to human minds and ability to think and feel just like humans. This second method of classification, “Based on Capabilities” of AI in relation to Human Intelligence, is more common in the tech industry.
Different types of Artificial Intelligence
The two main types of AI that are based on Functionality and Capabilities are
Type 1 – Based On Functionality
Reactive Machine
These are the simplest and most fundamental type of Artificial Intelligence. They are able to respond to various stimuli in the same way as a human. This type of AI has no memory power, so they lack the capability to use previously gained information/experience to obtain better results. These AI are not capable of learning like the ones we see today.
Example: Deep Blue is an IBM chess-playing computer that’s the best example of these types of machines. It is known for defeating Garry Kasparov, an international grandmaster, in the late 1990s. Deep Blue is able to identify the different pieces on the chessboard as well as how they move. It can identify all legal moves possible for both itself and its opponents. It chooses the best move based on each option. It can’t learn from past moves because these machines lack any memories.
Theory is limited
This type of AI, along with the ability of Reactive Machines, have memory capabilities so they can use past information/experience to make better future decisions. This category includes most of the applications we use every day. These AI applications can be trained using large amounts of training data stored in their memory as a reference model.
Example: Limited memory technology is used in many self driving cars. These devices store data such as GPS location, speed of other cars, and size/nature of obstructions. They allow you to drive like a human.
Theory of Mind
Theory of Mind, the next level of AI is very rare or absent from our daily lives. These types of AI are usually in the “Work in Progress” stage and are often restricted to research labs. Once developed, these types of AI will be able to understand the human mind in depth. This includes their thoughts, needs, and emotions. The AI will be able alter its response based on their understanding of Human minds.
Example: Winston, a researcher, demonstrated a prototype for a robot that could walk down a small corridor with robots coming in opposite directions. The AI can see other robots’ movements and can adjust its direction to avoid collisions with them. Wilson states that this Robot’s actions are determined by its “common sense” about how other robots move.
Self-Aware AI
This is the last stage of AI. Its existence at the moment is only a hypothetical possibility and can only be found in Science fiction movies. These AIs can sense and evoke emotions in humans and create their own emotions. These kinds of AI are still decades away, if they ever do materialize. This type of AI is what AI skeptics such as Elon Musk are concerned about. Because once the AI is self-aware, it can go into Self-Preservation mode. It might view humanity as a threat and attempt to end it.
Type 2 – Based upon Capabilities
Artificial Narrow Intelligence
This category includes all the AI applications that we see around us. ANI is an AI system that performs narrowly defined tasks, just like humans. These machines are unable to perform tasks that it has not been programmed for, so they cannot complete an unimaginable task. This system, based on the above classification, is a mixture of all reactive AI and limited memory AI. This category includes the AI algorithms we use today to perform complex Prediction Modelling.
Artificial General Intelligence
AGI can perform the same functions as a human being. It can learn, train, understand, and perform functions like a human. These systems will be able to perform multiple functions across many domains. These systems will be agile, and they will respond and improvise like humans in unprecedented situations. Although there isn’t an actual example of such AI in the real world, significant progress has been made.
Artificial Super Intelligence
Artificial Super Intelligence (ASI) will be the pinnacle of AI development. ASI will be the most powerful form of intelligence ever to exist on this planet. Because of its superior data processing, memory and decision-making abilities, ASI will be able perform all tasks better than humans. Many researchers worry that ASI’s arrival will lead to a “Technological Singularity.” This is a hypothetical scenario in which technology’s growth will cause uncontrollable levels of technological advancement, leading to a drastic change in Human Civilization.
It is difficult to predict what the future of AI will look like. We are far from reaching that stage, however, as we are only in the very early stages of developing advanced AI. AI advocates can tell us that we are only scratching the surface of AI’s true potential. AI skeptics should be aware that it is still too early to start thinking about Technological Singularity.
Additional Resource:
https://www.techtarget.com/searchenterpriseai/definition/AI-Artificial-Intelligence
https://www.investopedia.com/terms/a/artificial-intelligence-ai.asp
https://www.ibm.com/cloud/learn/what-is-artificial-intelligence