Artificial intelligence is a science, and like all examples, it has its subdivisions. See below, are the types of artificial intelligence according to their capabilities and functionalities within the spectrum of approximation between the functioning of machines and the human brain.
Artificial intelligence has long been perceived by many people as a fait accompli. Today, this term is understood in several directions:
- Teaching with the help of electronic computers.
- Computer networks are based on neurons.
- Speech artificial intelligence (synthesis and recognition of languages (speech).
- Creative processes with the use of computers.
Types of Artificial intelligence
Table of Contents
Since AI research aims to make machines “emulate” human-like functioning, the degree to which an AI system can replicate human capabilities is used as the criterion for determining existing types.
Depending on how a machine compares to humans in terms of versatility and performance, artificial intelligence can be classified into one or several types of AI.
The greater the ability to perform more human-like functions with equivalent levels of proficiency will be considered a more evolved type of artificial intelligence. In contrast, those with limited functionality and performance are considered a simpler and less evolved type.
Artificial Intelligence: What it is, how it works, and examples
AI capability
1. Reactive machines
They are the oldest forms of AI systems with limited capacity. They mimic the human mind’s ability to respond to different types of stimuli. The machines do not have memory-based functionality.
In short, it means that they cannot use previously acquired experiences to inform their present actions, that is, these machines cannot “learn”. Its usability boils down to automatically responding to a limited set or combination of inputs. His classic example is IBM‘s Deep Blue which beat Garry Kasparov in a chess duel.
it is a fundamental type of artificial intelligence. This type reacts to this second situational situation. He does not have informational data beyond the scope of this situation. He does not have a memory of past events to form a new decision based on them. That is, it is designed to solve one specific problem.
A good example of this type of artificial intelligence is the IBM Deep Blue chess electronic computer, which beat the world chess champion, grandmaster Garry Kasparov, with a large advantage back in the late nineties of the last century. She can distinguish between chess pieces and knows how to move each piece, the rules of the game of chess, can make a probable prediction of the opponent’s next move, and also choose her best response move from all possible options. But she has no memory of the events that happened before. Except, of course, the rule of three repetitions of the move, which by the standards of chess is a draw. That is the program does not take into account all previous actions, but only analyzes the current position of the pieces on the board and selects the most optimal next move.
2. Limited memory
Machines with limited memory are those that, in addition to having the resources of purely reactive machines, are also able to learn from historical data to make decisions. Almost all existing apps that we know of fall into this category of artificial intelligence.
All current systems, such as those using deep learning, are trained by large volumes of training data that are stored in their memory to form a reference model for solving future problems. From chatbots and virtual assistants to autonomous vehicles, all are powered by limited-memory AI.
This is the next-level type of artificial intelligence. He remembers a certain amount of data from the past, and they can influence the analysis of the current situation. A car’s autopilot is an example of this kind of artificial intelligence. For example, it monitors the speed characteristics and direction of movement of neighboring cars. These procedures cannot be done at once; for their implementation, it is necessary to determine the necessary objects and monitor changes in the parameters of their movement over a certain period. The results of the observations are summarized with the images of the external environment specified by the program, which already contains traffic lanes, traffic signals, and similar basic parameters, such as the turn of the roadway. They are used to decide when to change lanes without disturbing other road users. But these little bits of information from past experiences are not permanent. They are not stored as episodes of driving experience in a database that would be able to collect the skills of a chauffeur working as a driver for many years.
3. Theory of mind
For now, it works as a concept or a work in progress. Theory-of-mind-based artificial intelligence is the next-level type of artificial intelligence system that researchers are committed to innovating.
A mental level theory will be able to better understand the beings it is interacting with, discerning their needs, emotions, beliefs, and thought processes.
While artificial emotional intelligence is already a developing industry and an area of interest for leading computer science researchers, reaching this level will also require the development of other branches of AI.
To truly understand human needs, AI machines will have to perceive humans as individuals whose minds can be shaped by various factors, in fact, “understanding” humans.
He can perceive the emotional feelings and train of thought of people when performing certain actions. In addition, he captures the motives of behavior, and intended intentions, and can even show social community with a person.
Of course, this type of artificial intelligence does not yet exist in reality, but its examples are given in the Star Wars movie. One could end here and define this type as the main difference between the machine intelligence of the present and the future. But it is possible and concrete to suggest what type of concepts machines will form. In the future, machines will become more perfect and will be able to form not only the image of the external environment (the world) but also images of other living and non-living beings. In psychological science, there is a term “theory of mind” about this, that is, you need to understand that human beings, animals, and objects in the world around them have emotional feelings and thoughts that affect their behavior and actions taken. If you do not take into account (or do not understand) the motives and intentions of other people,
4. “Self-aware”
If the theory of mind is a work-in-progress concept, self-aware AI is a hypothetical formulation. This type of AI will not only be able to understand and evoke emotions in those it interacts with, but it will also have emotions, needs, beliefs, and, potentially, desires of its own.
It’s the kind of artificial intelligence that tech naysayers are wary of. While the development of self-awareness can potentially propel our progress as a civilization by leaps and bounds, it can also lead to catastrophe — the movie “The Matrix” and the domination of machines.
Once self-aware, AI would be able to come up with ideas like self-preservation, which could directly or indirectly spell the end of humanity, as such an entity could easily surpass the intellect of any human being and devise elaborate schemes to tame or enslave humanity.
This is the last stage in the progress of artificial intelligence when the system itself can form a concept of itself. As a result, scientists who deal with the problems of artificial intelligence will be able not only to understand the structure and operation of consciousness but also to design and build machines that have it. Individuals with consciousness understand themselves, understand their current state, and will be able to predict the emotions and feelings of another person.
Technical and functional ratings
5. Narrow artificial intelligence (ANI)
This type of artificial intelligence represents all the AI out there, including even the most complicated and capable AI that has ever been created. ANI refers to AI systems that can only perform a specific task autonomously, using human-like resources.
These machines cannot do anything other than what they were programmed to do and therefore have a very limited or narrow range of competencies. Even the most complex AI that uses machine learning and deep learning to teach itself falls under ANI.
6. Artificial General Intelligence (AGI)
General artificial intelligence is the AI agent’s ability to learn, perceive, understand and function completely like a human being. These systems will be able to independently build multiple competencies and form connections and generalizations across domains, greatly reducing the time required for training.
This will make AI systems as capable as humans by replicating our cross-functional capabilities.
7. Artificial Superintelligence (ASI)
As far as we can imagine, it would be the limit of AI development. The development of artificial superintelligence will likely mark the pinnacle of AI research, as AGI will become by far the most capable form of intelligence on the planet.
ASI, in addition to replicating the multifaceted intelligence of humans, will be vastly better at everything it does because of the overwhelmingly larger memory, faster data processing and analysis, and decision-making capabilities.
The development of AGI and ASI will lead to a scenario known as singularity. And while the potential of having such powerful machines at our disposal sounds appealing, these machines could also threaten our existence, or at least our way of life.
These are the 7 ratings thought to determine the level of artificial intelligence, if we think about capability, we are halfway through and evolving, but when taking into account the technical rating, we are in the first of the 3 stages and it will take some time to (and if) we get to the next.
With information: Javat point, Gov tech, Forbes
I was looking for this particular info for a long time. Thank you and good luck.