Wondering what AI really means?

By Dinesh Sharma

AI Artificial Intelligence as a concept has taken the world by storm, but has also left millions confused about its true definition. With over 400+ million hits on Google, it’s a natural assumption that there is a lot of curiosity about AI and all the technologies it comprises. Through this blog post, you can discover and understand AI effectively, thanks to a few real life use cases that elaborate on its implementation across industries. And hopefully, you will understand it well enough to spread the knowledge among your peers!

AI Definition – Breaking it down.

Artificial Intelligence has been defined in several ways, but the one we find most precise is – “Intelligence that’s been constructed by humans through the usage of machines to mimic their behavior’. Before we delve into how data scientists have reconstructed these behaviors, let’s take a look at the different aspects of the human mind.

  • Observe-observe-observe!
    On a daily basis, humans are exposed to heaps of data, making observation a key trait. A study even says that the human brain processes 50 to 70 thousand thoughts a day, which it then processes to create a memory. Examples of this include something as simplistic as observing a no parking signboard, to something as complex as taking in data from a teacher on a complex topic and processing it.
  • It’s all about the learning
    Observations provide humans with a plethora of signals, direct or derived, that allow us to build appropriate actions or responses to these signals in the future. Now, a direct signal allows a person to gauge a response based on a previous experience, such as getting an electric shock prepares you for being careful the next time you are around an exposed wire or outlet. On the other hand, a derived signal gives the human mind cues based on an observed second hand experience, like seeing someone get electrocuted, also makes a person more careful when presented with similar situations in the future.
  • The power of prediction: Now when an event takes place that the human mind has been exposed to before, it brings up the memory of the processed observations, correlates it to the learning gained from that experience and helps predict the smartest course of action. Going back to our earlier example, human cognition predicts danger when it sees an exposed electrical wire.
  • Getting into action mode
    The final outcome of the previous behavioral stages, action mode is when the human reacts in tandem with the prediction signal that’s being relayed. Meaning, the human stays away from the electrical wire, based on the previous observations, learning and prediction.


So far, we’ve gone through the different human behavioral aspects that AI draws from. In AI, what happens is data scientists train machines to artificially reconstruct intelligence that can draw on the human responses to an event. Let’s take a look at how the human intelligence building process takes place, step by step.


The process begins with feeding as much relevant data, direct or derived, as possible into machines, more the better in fact. Bigger the data pool fed in, more the observations


AI responds to different types of stimuli, depending on what you train it to do, these can be both direct and derived. In the case of direct signals supplied by data scientists, machines are able to create connections and learn the meaning immediately. This is also known as supervised learning and is the first step in discovering the world of AI. For example, if the data provided says 6=2 X 3, and 16=4 X 4, the machine learns that N=A X B.

On the other hand, when the data supplied consists of derived signals, the machine does everything it can to derive learning – also known as unsupervised learning. Meaning if the machine is provided visual data of different types of shirts, it eventually learns to identify a shirt, which constitutes the second step in understanding AI.


AI is built around the paradigm of a technology’s capability to predict outcomes based on data used to build its intelligence quotient. So when a machine has undergone training with either the supervised or unsupervised methods, it has essentially learnt to predict outcomes for both known and unknown data. Meaning if shown an unseen image of different garments, the machine would be able to predict if it consists of a shirt or no shirt.


Once AI has been used to generate predictions, the same are used to configure and implement probable actions based on the scenario. For instance, on command the machine would be able to pick out shirts based on various parameters like color, pattern, fit, etc.


Now that we have a grasp on what makes up AI tech, let’s move on to its innumerable implementations in the digital age. AI tech can essentially augment or replace any type of human behavior. However, the complexities in implementation would vary scenario to scenario. One thing we can say for sure is that the sky is the limit when it comes to making machines human-like in behavior and response, but it sure is a long road to get there.  

Owing to its immense popularity and acceptance in tech and business, AI has grown to become all pervasive. From our phones, smart home devices, to healthcare and human-like chatbots, the reality we are witnessing is most exciting.

Popular use cases:

  1. Trend prediction: Supply chain, sales, etc.
  2. AI for Chatbots and conversational apps: Customer service automation, product search, product questions and answers, recommendations etc.
  3. Voice interactions: Automated IVRs, Alexa, Google Assistant, new automobile experiences like ‘Hello Mercedes’.
  4. Computer vision: Automated traffic violation tickets, face detection and recognition etc.
  5. Healthcare: Correlating diagnosis reports and predicting probable health risks, patient care management and predicting risks etc.


We’ve learnt how AI constructs human intelligence through machines to replicate human behavior, allowing it to automate and repeat structured tasks. A lot of work is still required in reducing the challenges that arise with human cognitive involvement in making tasks happen accurately. To take it to the next level, AI will have to grow and morph to automate human-like reasoning, analyze sensitive and complex data and diagnose problems in real-time.

To know more on the differences between AI and ML, you can also read the article ‘AI vs machine learning’.     Curious much? Drop us an email at or visit our website to start a conversation!

Is AI and Machine Learning the same?

By Tejas Venugopal

We have all heard or read about how AI is the next big thing. It seems like every 3rd article or blog is about AI and how...Continue

Can AI think?

By Nitish Deshpande

What AI can do today AI today is helping to improve efficiency, accelerating growth, and making better policies in areas such as...Continue

AI without data

By Kamlesh K

Artificial intelligence or AI refers to the simulation of human intelligence in machines. (Refer the article Are AI and...Continue