“AI In A Nutshell” — The Easiest Explanation Ever

So— AI(Artificial Intelligence), Has 3 Main Parts —

1.ANI — Artificial Narrow Intelligence :

It includes applications like Self-driving cars, Web Search with AI, and AI in Farming etc.

2. Gen AI :

It includes applications like Chat GPT, Bard and other text, audio, video and image generation things.

3. AGI — Artificial General Intelligence :

It is said to have capabilities like us humans have today, from thinking to taking strong decision to emotional intelligence etc.

AI is Driven By ML — Machine Learning —

Machine learning, as the name suggests is just a machine which is learning on data, like we humans learn from the things and store it in our memory to interact with those things.

Machine Learning has supervised learning model, in simple words, like we teach our kids about different things then they learn it and they give answer when something is asked to them.

So in supervised learning we train the machine in A to B format like we give the machine the training data that both contains the A and B means the Input and The Output. So e are training the machine on this data and then after training when we will give a completely new input to the machine, it will try to predict the output based upon it’s past learnings on the data we’ve trained it on.

Why supervised Learning ??

We are training our machines on supervised learning because in the past years, the supervised learning took of because —

From the past 2–3 decades we saw a rise in the internet and a lot of data came over it so as the amount of data increased, and as the data increased we started to train the neural networks and they became Large Neural Networks.

Now WTF is this Neural Network???

So neural networks behave just like the neurons in our brain. whenever we think of something or we learn something new, new neurons in the brain are formed and those neurons connect with each other so we think properly and take decisions accordingly.

So the Neural Networks in Machine Learning also work the same why, while training the machine on the data for supervised learning, we tell it how to think or what to think first then connect things then to derive a conclusion.

How ‘More data’ helps in ML Training ??

So there is a lot of data out there and we are training our ML machine on that so while doing the supervised learning training, we need to give the data of both input and output as I told you earlier in this article. so before giving the data, we label it so that the machine understands it.

For ex- you want to build a NN(Neural Network) that takes a photo as a input and tell you whether it’s a cat or not. so we need to train the machine on labelled data with a label “Cat” on all the cat photos and “Not Cat” on all the photos that are not of cat, to teach the machine about how actually a cat looks like so that the machine gets an understanding of what is a cat or what is not. So next time after the completion of the training, when you give input of a photo of a cat, it will output “Cat”.

So the more the data, the more it will enhance the training and efficiency of the NN. The data can be of any form depending other use-case.

Terminologies of AI —

ML — Machine Learning:

It’s a field of study that gives the computer, the ability to learn without being explicitly programmed. And it often results in a software.

DS- Data Science:

Science of extracting knowledge and insights from the data.

DL — Deep Learning:

A field where we use ANN — artificial Neural Network to make a model that is trained on a vast amount of data.

What AI Can And Cannot DO ?

It is good at : Learning a simple concept when there is a lot of data available.

It works poorly: When it has to learn a complex concept when there is small amount of data available.

That’s it.

Please give your valuable feedback on tis article on [email protected] and suggest me what can I improve also write on which topic you want me to go in-depth and write an article in simplest form.