What is Artificial Intelligence?

Source: Deep Learning on Medium

What is Artificial Intelligence?


We propose an answer to the “What is AI?” question and in which areas of our lives artificial intelligence is used. Along the way, for a clear understanding of the subject we explain differences between Artificial Intelligence, Machine Learning and Deep Learning. Then, briefly describe Artificial Neural Networks, their applications and where it come from.


Artificial intelligence (AI) is a science that targets intelligent programs. AI can solve complex problems by imitating human thinking and improve their expertise by learning. Machine Learning (ML) is a type of artificial intelligence. The key difference between ML and AI is that ML is a type of AI that gives the ability for a computer to learn without being explicitly programmed and AI is the theory and development of computer systems able to perform tasks intelligently similar to a human. Deep learning is a subset of ML. DL designed to analyse data with a logic structure similar to human brain. To achive this, DL applications use a layered structure of algorithms called an Artificial Neural Networks (ANN). ANN is a deep learning method that simulate the neural system.

1.1. What is Artificial Intelligence?

Definitions of AI are both controversial and elusive. These definitions can be divided into four categories;

1. Systems that think like humans

2. Systems that act like humans

3. Systems that think like rationally

4. Systems that act like rationally [2]

To sum up we can say Artificial Intelligence is roughly defined as the ability of a computer-controlled machine to perform tasks related to higher mental processes such as reasoning, meaning, generalization, and learning from past experience, which are generally assumed to be human-specific qualities [1]. Thus, AI is neither a pure science nor solely a novel engineering discipline. This is proof of the cross-disciplinary nature of AI, involving cognitive scientists, computer scientists, engineers, and mathematicians.

The methodology and terminology of AI is still developing. AI is decomposing into and finding related subfields: logic, neural networks, object-oriented programming, formal languages, and so forth. This explains why the study of AI is not confined to mathematics, computer science, engineering, or rather, each of these disciplines is a potential contributor [2].

Another confusing issue is difference between Artificial Intelligence, Machine Learning, and Deep Learning. You may have recently been hearing about other terms like “Machine Learning” and “Deep Learning,” sometimes used interchangeably with artificial intelligence. Machine learning is a way of “training” an algorithm so that it can learn how will do their mission. “Training” involves feeding large data to the algorithm and allowing the algorithm to adjust itself and improve. So, program has learned from data. Deep learning is one of many approaches to machine learning. Deep learning was inspired by the structure and function of the brain. Artificial Neural Networks (ANNs) are algorithms that imitate the biological structure of the brain [8].

Difference Between Artificial Intelligence, Machine Learning, and Deep Learning[8]

1.2. The Various Areas of Artificial Intelligence

Artificial Intelligence has become an important part of our daily lives, we can be sure that AI will infiltrate our lives more and more in coming years without us even knowing. Our life is changed by AI because this technology is used in a wide area of day to day services. These technologies reduce human effort. Nearly every tech company is heavily investing in AI research and development. For instance; Google, Amazon, Microsoft and IBM are in a heavyweight fight investing over $20 billion in AI in 2016 [7] which clearly demonstrates us the importance that AI holds for businesses in general.

Artificial intelligence has produced many spectacular products even at this early stage in its development. In general, we can say that AI has some features like, reasoning, learning, language communication, planning, decision making and common-sense. Thanks to these properties, various problems were solved on financial systems, medical, heavy industries, and transportation systems. We can also talk about these subfields where AI is dominant such as such as perception and logical reasoning, proving mathematical theorems, diagnosing faults in aircraft, making autonomous decisions on space missions, diagnose diseases, vision systems, handwriting recognition and also enabling human-machine speech interfaces, intelligent robots, to specific tasks in gaming zone such as making the characters in a video game behave in a more human-like way or developing strategies in chess, poker or tic-tac-toe, etc.


2.1. Mechanism of Brain

The human brain is the most complex structure known in the universe. Artificial intelligence research is closely related to the study of the functions of the human brain. Computers are faster and more reliable than human in computing problems. In applications such as real-time natural language and image processing, human has an overwhelming advantage over the computer. For example, understanding a question is a big deal for a computer but in aspect of human, most difficult thing is find an answer because of limited memory. So, what makes man superior to the computer? Although neurons, which are the building blocks of the human brain, are slower than transistors [1]. The operation of brain is believed to be based on simple basic elements called neurons, which are connected to each other with transmission lines called axons and receptive lines called dendrites. The learning may be based on two mechanisms; the creation of new connections, and the modification of connections. Each neuron has an activation level which, in contrast to boolean logic, ranges between some minimum and maximum value [4]. Neural Networks makes an attempt to simulate human brain. If computers can be modeled to match one to one, then computers with the capabilities of the human brain are made.

Neural Networks

2.2. Artificial Neural Networks

Artificial Neural Networks is a functional unit of deep learning. Deep Learning is part of Machine Learning which is under the large umbrella of AI. Neural Networks learns from large data sets of known correct and incorrect examples. ANN has simple computational units that work together to solve complex problems. Each unit has a basic task outputting true or false based on several inputs. These units are grouped into three categories input layer, hidden layers and output layer. We are going to give an example about face recognition in ANN which is surprisingly difficult to identify someone for a machine. Firstly, input layer reads an image of someones faces. Hidden layers recombine the input layers outputs. In each hidden layer draw from the previous layers units to create new outputs. Finally, the last hidden layers gather the outputs and translates them into understandable result.

We can use ANN in some areas like marketing, healthcare, speech recognition, translating between languages, face recognition, identifying objects and images. [6] There are some examples from current applications such as google translate which can instantly translate between more than hundred different languages or we can base on automated self-driven cars being perfected with the help of neural networks.[5]

Artificial Neural Networks


Artificial Intelligence is a part of our lives. Even if we do not want these improvements, it will find its place in daily life. AI provides convenience in many areas. ANN shows us how people find solutions using natural systems and how machines learn from mistakes and datas. On the one hand, we are still a long way from building a humanoid and we do not know wheather it will be possible. On the other hand, if technology will move too fast like now, anything can be happen thanks to artificial intelligence.


[1] Doç. Dr. Vasif V. Nabiyev, Yapay Zeka, 2nd edition, Ankara: Seçkin,2005

[2] Robert J. Schalkoff, Artificial Intelligence: An Engineering Approach, 1st edition, New York: McGraw-Hill Education (ISE Editions), 1990

[3] James Allen, Thomas Dean, Yiannis Aldimonds, Artificial Intelligence: Theory and Practice, 1st edition, Boston: Addison-Wesley,1995

[4] Chennakesava R. Alavala, Fuzzy Logic and Neural Networks: Basic Concepts and Applications, 1st edition, Delhi: New Age International Pvt Ltd, 2008

[5] Poramate Manoonpong, Neural Preprocessing and Control of Reactive Walking Machines: Towards Versatile Artificial Perception-Action Systems (Cognitive Technologies), 2007th edition, New York: Springer, 2007

[6] Yu Hen Hu, Jenq-Neng Hwang, Handbook of Neural Network Signal Processing (Electrical Engineering & Applied Signal Processing Series), 1st Edition, Florida: CRC Press, 2001

[7] Simon Greenman. (2018, March. 06). Who Is Going To Make Money In AI? Part I [Online]. Available: https://towardsdatascience.com/who-is-going-to-make-money-in-ai-part-i-77a2f30b8cef

[8] Michael Copeland. (2016, July. 29). What’s the Difference Between Artificial Intelligence, Machine Learning, and Deep Learning? [Online]. Available: https://blogs.nvidia.com/blog/2016/07/29/whats-difference-artificial-intelligence-machine-learning-deep-learning-ai/