Types of Artificial Intelligence

Types of artificial intelligence

The following article provides an overview of the most important types of artificial intelligence. The main purpose of artificial intelligence is to allow machines to function like humans. Thus, the primary means of classifying AI is based on its ability to reproduce human functions. AI can be broadly classified into two types, both based on its ability to reproduce the human brain. A type of classification that is "functional" classifies AI according to its resemblance to the human brain and its ability to think and feel like humans. The second most prevalent style in the technology industry is the AI-based "capabilities" of human intelligence.

Different types of artificial intelligence

There are two types of AI based on functionality and capabilities.

Type 1 - Performance-based

These are the most basic and oldest forms of artificial intelligence. They reproduce the ability of man to react to different abilities. This type of AI does not have the power of memory, so they do not have the ability to use the information / experience gained to get better results. No capacity.

Example: Deep Blue, IBM Chess Supercomputer, is the best example of this type of machine. He is best known for defeating International Grandmaster Gary Kasparov in the late 1990s. Dark blue can identify different pieces of the board and how each one moves. He can point out all possible legal action for himself and his opponents. Depending on the option, it chooses the best possible move. However, he does not have the opportunity to learn from his past movements because these machines do not have their own memory.

2. Limited theory

This type of AI, as well as reaction machines, has a capacity for memory so that they can use past information / experiences to make better decisions in the future. Most of the general applications around us fall into this category. These AE applications can be driven by a large amount of training data which they store in their memory in the form of reference models.

For example: Many autonomous cars use limited memory technology. They store data such as the position of the GPS, the speed of the surrounding cars, the shape / nature of hundreds of other obstacles to human-like driving.

3.Theory of the mind

Theory theory is the next level of AI, which has little or no presence in our daily lives. These types of AI are mostly in the "development at work" stage and are usually limited to research laboratories. This type of AI, once developed, will have a much deeper understanding of human minds, including their needs, tastes, emotions, thought processes, etc., about human minds and their desires. Will be supported, and will be able to edit his own answer.

Example: Researcher Winston showed in his research a prototype robot that can walk with other robots in opposite directions in a small corridor, predict other movements of the AI ​​robot and move left, right, left or otherwise. Can change to avoid. Possible collision with incoming robot. According to Wilson, the robot determines its action based on its "intelligence" of how other robots will move.

4. Self-aware AI

This is the last stage of AI. Its current existence is only fictional and can only be found in science fiction movies. These types of AI can not only understand and ignite human emotions, but can also have their own emotions. This type of AI has been around for decades, if not centuries. This is the type of AI that suspects like Elon Musk are wary of. In fact, once it becomes self-aware, AI can go into self-defense, perceive humanity as a potential threat, and continue its direct or indirect efforts to eradicate humanity.

Type 2 - capacity-based

1. Narrow artificial intelligence (ANI)

All of our existing AI applications fall into this category. ANI includes an AI system that can perform specific tasks defined as humans. However, these machines cannot perform tasks that were not previously planned, so they cannot perform unprecedented tasks. Based on the above classification, the system is a combination of all responsive and limited memory AI. The AI ​​algorithm we use to perform the most complex prediction modeling in the world today falls into this category.

2. General artificial intelligence (ANI)

AGI has the ability to train, learn, understand and perform functions just like normal human beings. These systems will have the capacity to reduce multi-purpose capabilities in different areas. The system will be more agile and will react like humans when faced with unprecedented scenarios. There is no real example of this type of AI, but significant progress has been made in this area.

3. Super Artificial Intelligence (ASI)

Super artificial intelligence will be at the forefront of AI development. ASI will so far be the most powerful form of intelligence on the planet. This will allow humans to perform better due to their excellent data processing, memory and decision making capabilities. Some researchers fear that the advent of ASI will eventually lead to "technological isolation". It is a hypothetical situation in which the advancement of technology will reach an uncontrollable stage which will bring about an unimaginable change in human civilization.

Currently, it is very difficult to predict what our future will look like when our more civilized form takes the form of AI. However, we can say with certainty that we still have a long way to go because we are at a very early stage in the development of advanced AI. For supporters of AI, we can say that we are scratching the surface to discover the true capabilities of AI and for expert breakers of AI, it is necessary to cool down on the technological singularity.