Artificial Intelligence

Home Benefits Dangers Development Contact Us Go back to main site

What is A.I (Artificial Intelligence)?


The formal definition of Artificial intelligence is that
"A.I is the field that studies the synthesis and analysis of computational agents that act intelligently".
Taken from book "ARTIFICIAL INTELLEGENCE" BY "DAVID L.POOLE AND ALAN K.MACKSWORTH".
Now in order to understand this definition we need to look at the individual parts of it.
we can break it down into few parts as follows:

  • Agent

  • Computational Agents


Agent means anything that work in a field.

This means it can be a human, a car, an animal and the same way a software or a hardware in general.
However, it is considered intelligent if it performs tasks that are appropriate to its objectives and circumstances, learns from its experience through interaction with the task, its flexibility towards changing environments and objectives, and makes suitable choices based on its interaction with the task.

An agent whose judgement about its behaviour can be described in terms of calculation is called a Computational Agent.

This means that its behaviour can be fragmented into basic processes which can be used in a system.
The main goal of A.I is to understand the foundation that create astute actions possible in common and man-made systems,
but it should not be confused with thought because thought affects behaviour and this is not efficient.
Now we can assume that if something acts intelligently it is intelligent.

This is where external factors that define intelligence come in.
So in order to explain intelligence from External Behaviour Alan Turing created a test called the Turning Test in 1950.
The test comprises of an Interrogator and a witness, Both interact through a text interface.
The interrogator may ask the witnss any questions, If the interrogator cannot distinguish that that the witness is human or not then the witness is intelligent.

The Turing Test unfortunatly can show how to percieve intellegence but it does not show how to become intellegent.
In general terms Artificial Intelligence is the replication of human brain, which is also called the Neural network or Nervous System .

Picture of a circiut giving birth to intelligence

Types of Artificial intelligence


There are four types of Artificial intelligence.
  1. Reactive Machines
  2. Limited Memory
  3. Theory Of Mind
  4. Self-Awareness
    5. Taken from Article on >>Website<<
    Author "AREND HINTZE, MICHIGAN STATE UNIVERSITY / NOVEMBER 14, 2016"
Reactive Machines as the name suggests are based on reaction only.
As it only takes into account curent reactions from interactions with tasks, This is because it can only precieve the world directly.
It does not rely on interanl concept of the world and makes decision on the go.
Limited Memory means limited concept of the world Which is Preprogrammed.
This type of A.I is made for a specific task like self-driving cars.
Here the A.I makes observation in its environment and add it to its preprogrammed concept of its environement.
Theory Of Mind is a Psychological term used to describe " the understanding that people, creatures and objects in the world can have thoughts and emotions that affect their own behavior".
Theory of mind is the step that we are trying to achieve this is when ARTIFICIAL intelligence learns that everybody has feeling and thoughts, that there is right and wrong.
If and when we Achieve this step then we can go and make Type 4 a reality.
Self-Awareness is the last and final step in cognitive Artificial Intelligence.
This is where humans will be able to create neural networks like that of humans.
This is considered as dangerous by some and obsession by others, but overall this is the pinaccle of human intelligence, because after this we wont need innovation it will be done for us.