Introduction to Artificial Intelligence

How do you define thinking? In a way machines already think don’t they? If someone or something thinks differently from you, does it mean that it’s not thinking? We define intelligence and thinking on the basis of decision making ability. Artificial Intelligence is the creation of computers and machines that are capable of making logical decisions on their own.

Father of Artificial Intelligence

The mathematical logic and the thoughts of the philosophers led to Alan Turing developing the theory of computation. He is known as the father of Artificial Intelligence. His theorization and logical reasoning directly led John McCarthy, Allen Newell and company to research the prospect of AI and it’s feasibility. It is important to mention that the official idea and definition of Artificial Intelligence was, in fact, coined by McCarthy himself.

“Every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.”

John McCarthy

When a program or a machine passes the Turing test, i.e. it is indistinguishable from a human being to a blind tester i.e. a person who does not know who he’s talking to, it can be known as an AI. In essence, AI is a machine with the ability to solve problems the way of, and with the thought process of, a human being. A computer will be considered a form of intelligence when it learns how to improve itself at solving problems.

The Progress of AI

It is very simple to program a computer to do a simple task, the difficulty increases as the complexity of the task increases. Programming a computer to make decisions is on a whole different level altogether. If programmed correctly, AI can make progress worth of billions of years in days. To make progress, you need to provide the AI data. The more data it has, the more it can improve. What you basically do is program the AI to learn from the data and make improvements on what was lagging.

Generally, the best way to make progress is to program the AI to do a task or run a simulation and then restart when it fails while passing the acquired information of the events leading up to the failure and moving forward with the approach which yielded the best results.

To work, AI needs access to components which it can operate. And to do that, it needs to know how to operate those components freely as well. Giving knowledge is the main part of engineering an AI program or network. This includes mainly the knowledge of determining intelligently how to differentiate between what to do and what not to do.

A correctly developed AI can render humans obsolete given time. Primitive forms of AI are used from maintaining your house temperatures to advertising as per your need.

An AI needs to be able to do the following to be successful:

• Generalize: Be able to act logically in a previously unencountered situation.
• Reason: To draw conclusions based on the situation at hand.
• Solve Problems: Find a solution to a given a problem.
• Perceive: Analyzing an environment and drawing relations between objects, and outcomes of an action.
• Language: Understand language, grammar, syntax, words and rules of a language while being able to distinguish between tones.