top of page
  • Pranisha

Generative AI

Generative AI has evolved at an unimaginable pace, it’s changing the way the mankind thinks and behaves - Tends to change people’s approach towards life in all aspects.

This is doing what computer, iPhone, internet did to the mankind.

It’s like a behemoth brain anyone from anywhere may apply (prompt) in the same way like a human brain. It’s not absolute yet, but it’s not a overstatement to say that evolving and catchup at the speed of a lightyear.

This is going to be the new world order and will take hold of every human and business under the sky.

Artificial Intelligence and Generative AI


Artificial Intelligence is a field of computer science that develops intelligent software models that learn, train, reason, and work independently.


Generative AI is a type of AI that produces various types of content including text, imagery, audio, video and synthetic data based on the human or software input.


Machine Learning and Deep Learning


Machine learning is a subfield of AI and a program system that trains a model from the input data, the trained model makes useful predictions from new and never seen data drawn from the same data that was used to train the model.

The Different ML models are:

  1. Supervised models are the ones that produces output based on the labelled data like name, number etc..

2. Unsupervised models generates output based on the raw data (unlabelled data).

3. Semi supervised models are a mixture of labeled and unlabelled data.

Deep Learning is a type of Machine Learning, it’s based on artificial neural networks that are connected through nodes, knowns artificial neurons.

Artificial neural networks are similar to that of the of the human brain. Artificial neural networks processes more complex patterns of data and information. These neurons take input, process, predict data, and produce output through the neurons. It’s multi layered neurons  that learn more complex patterns from both labelled and unlabelled data.


Labelled data helps in learning and unlabelled data helps to generalise to new examples.


Discriminative AI and Generative AI 

These are the types of Deep Learning.

  1. Discriminative model classifies or predicts labels for data points that are trained on the data set of labelled data points and learns the features of data points and the labels. Once the discriminative model is trained, it can predict label for the new data points.

2. Generative AI  is the subset of deep learning which uses artificial neural networks that processes labelled and unlabelled data using supervised and unsupervised and semi supervised methods. Natural Language Processing (NLP) and Large Learning Model (LLM) are used in the generative AI for training, generating a processing the input data and generating the output.

Studies mention that the Generative AI applications may run up to around $5 Trillion of the global economy per year. Any business model across globe or a person if not connected to Generative AI will become redundant and may turn out to be incompetent. The one who is not using AI may definitely be out of the radar in the current and future era.


It’s not that the mankind coming in touch with the AI, it’s already known to the mankind and is widely used in the form Siri on Apple products and Alexa from Amazon, YouTube recommendations, chatbots that show up to help the surfers of websites.


For instance, let’s take the example of a computer trying to learn what is a cat - In a traditional programming model, one has to define or hardcode the dog with features like legs, fur, etc.. Whereas in Generative AI, it’s not just perceives and but also classifies the image of a dog and creates an image or text description of a dog or video even.


Generative model generates new data instances. The output of a Generative AI is Natural Language, Image, Audio, Video or a combination of all. A model is build through labeled data and training code. The developed model predicts or classifies or a cluster output.

 For getting the best out of the Generative AI for self or company, one must have the imagination (what can I do and How can I do), this skill is known as Prompt Engineering. Back propagation and Reinforcement Learning with Human Feedback (RLHF) help the generative AI on getting better with the output. 


Will discuss in detail about Large Language Model (LLM), Prompt Engineering Skills, Natural Language Processing (NLP).


Generative AI tools


Gemini from Google, Microsoft’s Copilot, Copilot Corporate and Copilot Studio and Generative AI from Open AI. Shall talk about all these in our next blogs.





bottom of page