If you are working in an IT industry, you might have already heard about AI taking over the jobs.

Before even going there, let us understand what exactly is Artificial Intelligence. Most people confuse Automation with AI, since both these terms are often used together, not interchangeably.

Well, automation is software program which mimics manual + repetitive + sequential activities without requiring intervention or assistance. These are rule-based systems which have to be pre-programmed to complete series of steps.

On the other hand, AI is judgement-based system which mimics learning process of Human Operator. It gains knowledge from data as “experience” and adapts itself to execute processes more efficiently next time.

There are four major areas in which typical AI system replicates human brain.


It is how you perceive environment and collect relevant data. It uses computer vision/audio/sensors for the same.


Analyze and understand the information. It involves NLP (Natural Language Processing, Knowledge trees).


Act based on the information available. It uses inference engines to formulate proper response.


Learn based on the feedback/outcome. That is where machine learning comes into play.

Now, let us look at leading systems in AI space.

IBM Watson

It is one of the prominent system which helps apply human-like behaviour to computing problems. It uses NLP(Natural Language Processing) to sense/comprehend and analytics to generate and evaluate hypothesis. It also has capability to learn dynamically based on outcomes.

So, how do you really ‘train’ Watson ?

It starts with data procurement (gathering), then preparing it in Watson format and then uploading it to the server. Then you prepare training modules within Watson and verify. You can also re-train at frequent intervals to match real-world behaviour.


It is open source library which mimics neural networks and is used in speech recognition, computer vision and other language related tasks.

Microsoft Azure Machine Learning Studio

It is drag & drop graphic-based tool to  build,test predictive analytics solution and deploy it as web service.

Python NLTK & Stanford Parser

It is machine learning open source library for NLP and data mining.

Amazon Alexa

It is voice-based NLP and ML (Machine Learning) algorithm tool. It can hear voice request from customer and provides appropriate response e.g. Weather forecast for today.

While, such systems are also evolving, we should understand that AI is still way off from completely replacing human operator’s cognitive tasks. While, it can take burden off by assisting level 1/level 2 tasks (e.g. providing probable solutions in support ticket), it still needs to ‘learn’ a lot to handle cognitive tasks in entirety.

We should consider AI as enabler (personal assistant) who takes care of entry-level cognitive tasks and leave us enough time to handle more complex ones. If you are up for tasks requiring higher cognitive capabilities, then you should not be as scared with stories of AI taking over your job.