Difference and Similarities between Artificial Intelligence , Machine Learning and Data Science.

Original article was published by Henil Vedant on Artificial Intelligence on Medium


Difference and Similarities between Artificial Intelligence , Machine Learning and Data Science.

We all are well acquainted with these jargons of the modern computing world and the hype around them , but let us dive a little bit deeper to actually understand what these words mean and how often can we use them interchangeably and when should we use them explicitly. The blog aims to explain the distinction without difference between the terms and also the fundamentally different concepts about the domains.

1) Artificial Intelligence :

Artificial Intelligence is the name given to the domain of study that aims at teaching computers the ability to have a cognitive reasoning and problem solving abilities that can help drive business solutions and innovation and obtain a particular task. There are various ways through which a computer is trained and the expected result from the computer is obtained. The training phase can be classified into : supervised, unsupervised and reinforcement learning. In simple words supervised is trained , unsupervised is not trained and reinforcement learns over the course of time based on the feedback it gets.

But AI, is often considered as a threat to society, the idea of “computer’s taking over the world”. It might seem as a overstatement but some of the greatest contributors to the world of computers share the same idea..

Bill Gates said “First the machines will do a lot of jobs for us and not be super intelligent. That should be positive if we manage it well. A few decades after that though the intelligence is strong enough to be a concern”.

Elon Musk : “I think we should be very careful about artificial intelligence. If I had to guess at what our biggest existential threat is, it’s probably that.”

Now these are some heavy words but to understand what they really mean you first need to understand what AI is all about, the threat that these great personalities are talking about is not the same AI that you and I might be acquainted with currently.

AI can broadly be classified in various ways depending from people to people , however for our understanding we will divide AI in 2 parts :

1) ANI — Artificial Narrow Intelligence.

2) AGI — Artificial General Intelligence.

Let us now look into each of them individually.

ANI — Artificial narrow intelligence is almost everything that you see today and know about today , it literally constitutes about 95% of all the Artificial Intelligence that there is in the world.

ANI refers to a sepcific goal driven output for the computer. It does not need to anything else except to achieve the goal , it has specific approach to reach that goal and that is the main idea behind ANI. So for example, a self driving car is an example of ANI , it has just one goal.. avoid accidents and drive safely from one destination to another.. It can have alot of sub-goals for instance in this example, the subgoals would be to avoid accidents,take the least dense road and to reach the destination in most fuel efficient manner.

ANI can also be called as narrow AI or weak AI.

This is the type of AI which is beneficial to the society and makes us move forward and take the next step in the history of evolution of mankind.

ANI is not the type of AI that is feared by the forementioned personalities , however the rate at which AI is catching up , it won’t be long before ANI evolves itself and catches up , we shall learn about that when we talk about AGI.

Some of the examples of ANI include : Face recognition tech, Self driving cars , IBM watson etc.

AGI– AGI refers to Artificial General Intelligence , it is when a computer can do everything that a human can do and provide more efficient results in less time with minimal complexity or computation. To develop such a type of machine we need alot of research and development in the field of neural networks as they should be able to implicate the complex decision making and cognitive thought process that a human mind performs at really superfast speeds. Now for a computer to be called an Artificial Intelligent being it not only has to be able to perform these tasks, it also should be able to perform them at faster speeds and more complexity should be handled, this can be called as STRONG AI.

This is the type of AI that can be a threat to mankind , if used maliciously..however, the concept of AGI is still a hypothetical concept with no real groundwork to give us a idea of how things will be under the AGI controlled system.

2) Machine Learning :

Machine Learning is the domain where computer is trained and taught to give expected resutls and have the consistency to produce same results again and again, each and every time. Machine learning is an application of ARTIFICIAL INTELLIGENCE.

It is the ability of a computer to learn on its own based on various feedbacks and the system behaviour without being having to explicitly code everytime. This saves a lot of time,effort and more importantly the computer learns better and understands the small details that humans can not. At the same time accuracy and the consistency over repeated trials is what makes ML a very important part of modern day computer world.

The ML learns in the same categories as mentioned above.. Supervised , unsupervised or reinforcment learning.

As you now know, ML is a subset of AI. They are almost used extensively and interchanged as there is not a lot of difference between them for a layman. However, AI is a very broad domain that includes alot of other things.

Machine learning is essential for implementing AI , however it is not the only thing. ML is a ever developing field with new developments occuring daily.

3) DATA SCIENCE :

Data science is the field of study which uses both.. ML and artificial intelligence but in a limited approach. Data science is all about finding patterns , knowledge from a collection of data, to make better decisions. For this purpose you can train the machine using ML for pattern recognition or even use AI to create composite dynamic models that can predict output of the likely events.

Data science is almost everywhere from stock market to sport, Data science is a buzz word that almost every other person in the industry uses without actually doing much in the field of DS.

Data science includes Data Cleaning , Data Pre-processing , Data sorting and then processing data to get results.

Alot of tools are available for you to visualise data and represent it into a form that a common person can understand.

AI is a Superset with ML being the subset.

Data science uses both AI and ML but it can not be called as a superset for AI and ML as its practical approaches and other functionalities are really different.

Nor can AI be called as a superset of Data Science because there are alot of things that Data science has but AI does not, They are 2 non disjoint sets with some similar elements.

But except for those elements , they can easily be disjoint sets, that’s how wide the difference is.