Understanding Artificial Intelligence and Machine Learning

Artificial Intelligence (AI) and Machine Learning (ML) have become mainstream nowadays. These two are hot buzzwords right now in technology market. They are not the same thing, the difference in both the terms can be confusing sometimes.

In simple words, Artificial Intelligence (AI) is the ability of a program or machine to think ,learn and solve problems like humans. To create intelligent computer programs/machines that can think logically, solve the problems and react like humans. AI can be categorized into two categories:

a) Applied AI: When the machines are designed to perform some specific tasks then it is known as Applied AI. For example: the program designed to recognize an image of Cat, so in this case we can say that program is intelligent enough to identify Cat. But this program is intelligent in that specific area only because it is programmed only for that specific task. If same program will see an image of Dog then it can't recognize which animal is that because it only knows the pattern of Cat not Dog. So we can say Applied AI is intelligence, but in very limited/specific field. Some real time examples of Applied AI are face recognizing software (Facebook), stock trading system....etc.

b) Generalized AI: It can be referred to the machines/devices which can evolve, learn and are intelligent enough to handle any tasks like humans. It is also known as Artificial General Intelligence, it is the intelligence of machines/programs that could perform any intellectual task that a human being can. It is an emerging area of AI and it also bring us the concept of Machine Learning.

Machine Learning: In simple words, Machine Learning (ML) is one of the approach to achieve AI, but it's not the only approach. For example: self driving cars are using machine learning and also rule based systems (other approach) to achieve AI. So all Machine Learning is AI but not all AI is Machine Learning.

ML provides program the ability to automatically learn instead of hard coding the program with specific sets of instructions to perform the task. ML focuses on the development of computer programs/software's that can access the data so that programs can use that data to learn for themselves. It involves feeding large amount of data to the algorithm and allowing the algorithm to learn itself from data, and improve by its own without human intervention. The technique to learn from data by a program is also known as Deep Learning. So, Deep Learning is itself a part (subset) of Machine Learning.

Some real time examples of ML are:

1) Remember before when you upload any picture on Facebook, it will prompt you to tag your friends in the picture. But now, it will recognize familiar faces from your friend's list and will tag them on its own. That's really impressive machine learning algorithm.

2) Amazon uses ML to offer personalized recommendations to the customers based on the customer's previous purchase or other activities.

3) Even, LinkedIn also uses ML algorithms to give you the message suggestions to reply to particular message. For example: if someone messaged you Hi on LinkedIn, then it will give you message suggestions like Hey, Hello, Hi so that you can select any of that suggestion to reply to that message. It's really an impressive user experience feature.