Artificial Intelligence

Human beings are different from all other animals in our beautiful Earth. It’s definitely because we can think, we can understand things, apply knowledge to improve our skills. These all points that we are intelligent.

Well, I wish to repeat the question asked by the father of modern computer science, Mr. Alan M Turing, “Can machines think?”.

Alan Turing is an eminent English Mathematician and Computer Scientist. He proposed the question in his paper ‘Computing Machinery & Intelligence’. At present, we will first say “sure, they can think!”. But even today no machines are intelligent we will discuss this in later in this article.

Alan Turing

So, what does the term Artificial Intelligence really mean? The term AI was first coined by John McCarthy in 1956 when he held his first academic conference on the subject and is defined as the science and engineering of making intelligent machines, especially intelligent computer programs. Even though the term was coined and was defined in 1956, the journey to understand if the machines can truly think began much before that. In Vannevar Bush’s seminal work ‘As We May Think’ (1945), he proposed a system which amplifies people’s own knowledge and understandings. Five years later, Alan Turing wrote a paper on the notation of machines being able to simulate human being and the ability to do intelligent things.

No one can refute a computer’s ability to possess logic. But to many, it is unknown if a machine can think. The precise definition of think is important because there has been some strong opposition as to whether or not this notation is even possible. For example, there is the so-called ‘Chinese room’ argument. Imagine someone is locked in a room, where they were passed notes in Chinese. Using an entire library of rules and look-up tables they would be able to produce valid responses in Chinese, but would they really ‘understand’ that language? The argument is that since computers would always ve applying rote fact lookup they could never ‘understand’ a subject.

This argument has been refuted in numerous ways by researchers, but it does undermine people’s faith in machines and so-called expert systems in life-critical applications.

The Turing Test

Do you know how a machine is that a machine is said to be intelligent? A single sentence answer is ‘The Turing Test’. Well, what is that?

Alan Turing went on to propose a method for evaluating whether machines can think, which came to be known as The Turing Test. The test, also called the ‘Imitation Game’ as it was called in Turing’s paper Computing Machinery & Intelligence, was put forth as a simple test that could be used to prove that the machines could think. Turing Test takes a simple pragmatic approach, assuming that a computer that is indistinguishable from an intelligent human actually has shown that machines can think. The Turing Test is a central, long-term goal for AI research -will we ever be able to build a computer that can sufficiently imitate a human to point where a suspicious judge cannot tell the difference between human and machine? From its inception, it has followed a path similar to much of AI research. Initially, it looked to be difficult but possible (once hardware technology reached a certain point), only to reveal itself to be far more complicated than initially thought with progress slowing to the point that some wonder if it will ever be reached. Despite decades of research and great technological advances the Turing Test still sets a goal that AI researchers strive toward while finding along the way how much further we are from realizing it.

The Turing Test involves a human interrogator and two contestants — a computer and a human. The interrogator converses with these contestants via computer terminals, without knowing the identity of the contestants. After a sufficiently long period of conversation, if the interrogator is unable to identify the computer, then the computer said to have passed Turing Test and must have considered intelligent. Turing predicted that by 200 computers would pass the test. There have been various programs that have demonstrated some amount of human-like behaviour, but no computers have thus far passed Turing Test! I have got a youtube video for you to understand Turing Test more easily. https://www.youtube.com/watch?v=sXx-PpEBR7k

AI has two main goals. One is to create Expert Systems, the systems which exhibit intelligent behaviour, learn, demonstrate, explain, and advice users. And second is to implement human intelligence in machines. Creating systems that can understand, think and behave like humans.

Most of the AI systems in today are weak AI, which was designed to solve specific problems. Even AlphaGo which was able to beat the human champion in a board game, Go is considered to be a narrow AI.Go is a more complex game than Chess. It is played on 19x19 board. What makes Go complex is the fact that the game is played purely by instructions and patterns from past experience and not by strict rules, that an expert can explain. Also in Go, there are endless possibilities for each given move.

Machine Learning

Well, then how is a computer system learns things from experience? That is what so-called machine learning. ML is the field of computer science that gives computers the ability to learn without being explicitly programmed.

The term ML was first coined by Arthur Samuel, an American pioneer in the field of computer gaming and AI in 1959, while he was at IBM. Let us consider a sample program.

You have a website or a program in which your users can upload images and wanted to tag whether it contains a cat or dog. It is quite difficult to do by traditional programming or, say it is impossible for traditional programming. There is the use case of ML. We first create our own model that is we let the program to remember some photos of cats and dogs. And we train our model again and again. Feed it more and more data. After these steps, our program will able to identify whether the user uploaded media contains a cat or dog.

Machine Learning itself can be classified into 3 based on the nature of the learning.

1. Supervised Learning: In this type input and output are specified for training.
2. Unsupervised Learning: Only inputs are given to recognize the pattern.
3. Reinforcement Learning: Real world feedback is provided to the system on the go.

There are a bunch of libraries which can help you to create an intelligent program. TensorFlow is the most popular and open source library developed by Google for creating neural networks and computation using dataflows and graphs.

The hard problem in the field of AI is finding a way to teach a machine to think, but in order to articulate ‘thought’ in a way current computers can understand, we must first understand thinking and intelligence ourselves. Until then, we will go on creating chess programs which rely on brute force, expert systems which fail to notice the obvious and converse with programs which just aren’t that interesting to talk to.

Let’s conclude this article here. We will talk about Expert systems and the story of Sophia and lot more in another article.

Have great day friends :)

Related Topics For You To Refer:

• Computing Machinery & Intelligence — Alan Turing
• As We May Think — Vannevar Bush
• Expert Systems
• To learn More about TensorFlow: www.tensorflow.org

I’ve created a presentation for you to see. I highly recommend you to learn more from it. Click Here.