What is Artificial Intelligence/ Machine Learning?

2 mins read
Artificial Intelligence

A computer is a machine that can only perform those tasks which it is programmed to do. It cannot think on its own. Artificial Intelligence (AI) is a branch of computer science that aims to bridge this gap and give a computer this ability to “think” and mimic human brain. This includes reasoning and making decisions so as to achieve a specific goal. This field holds the belief that human intelligence can be quantified and thus simulated. It was Alan Turing who asked the question first time whether machines can think or not. Since then, machines and AI have come a long way.

In the most basic terms, AI is a set of algorithms (i.e. a series of steps) that are used by a machine to arrive upon a decision. These decisions could then be validated based upon the inputs and outputs given to the machine before the machine is used for actual decision making. AI is considered to be of two types – weak and strong. Weak AI is used to perform a specific task whereas Strong AI is generic in nature and is able to perform complicated tasks like a human being.

Machine Learning (ML) is an application of AI where a machine or a computer program learns to do a task without the user giving it explicit instructions for the same. This is done by way of learning patterns in the data provided to it. In order for such computer program to derive meaningful patterns, they need to be provided with a large amount of data records. Deriving patterns from small data sample is bound to fail in a large number of cases as the patterns will not be truly reflective of the underlying data.

ML is basically of 2 types – supervised and unsupervised. Supervised learning relates to use of labeled data-sets where both inputs and outputs are known. Thus relation between input and output can be derived in a relatively easier way. In unsupervised learning, unlabeled data sets are there where only inputs are known and patterns and structure are found in the outputs.

Deep Learning is a type of AI where working of the human brain is imitated for making decisions by processing data and creating patterns. Deep learning can learn from data that is both unstructured and unlabeled.

Some of the applications of AI are:

  • Self driving cars
  • Image Recognition
  • Personal assistants like Siri, Alexa, etc
  • Natural Language Processing
  • Spam filters in email
  • Robots doing specific tasks

As we see that AI/ML can be used to perform a variety of tasks, but there are a few downsides of AI/ML which are as follows:

  • AI/ML can be used to harm humanity
  • AI/ML can lead to loss of jobs as machines will replace humans in taking decisions or doing work
  • AI/ML may not be able to take appropriate decisions in new situations that are complex
  • AI/ML may take biased decisions based on the data used by it for learning
  • AI/ML may develop its own goals apart from the goals given to it

This post is written by Amandeep.

For Wikipedia entry on Artificial Intelligence, click here.

For Wikipedia entry on Machine Learning, click here.

For Wikipedia entry on Deep Learning, click here.

For more posts on Artificial Intelligence, click here.

For more posts in The Cyber Cops project, click here.