Nagabhandanam - The Advanced Encryption Algorithm for Modern Era??

With the advent of supercomputers and quantum computing in the mere future

all the conventional encryption techniques and blockchain hashing data encryption techniques used will become futile.

With the introduction of artificial intelligence and large language models, we have been able to play around with the data.

and create many variations in the data, which has been quite useful in the modern era, but what about bad guys in society?

People who are deepfaking the videos and are hacking the systems to access sensitive information.

Nagabhandhanam, the ancient Indian technique or algorithm, was used to protect the treasures of the kingdoms from unauthorized people, which turned out to be the most powerful thing in ancient Indian literature.

How about using it to encrypt and decrypt the present internet data?

Sounds crazy, isn't it??




 

Let's see the similarities between the existing technology, Mother Nature, and Ancient encryption methods.

1. Artificial Neural Network: It is inspired by the firing of neurons in the brain. 

Which in fact lead to different concepts.

2. IOT (Internet of Things): It is inspired by our five sense organs, where the data is captured periodically and continuously on the same scale, sensors like temperature, etc. can be related.

3. Attention Models: Just like how you focus on a particular thing with more concentration while reading a book not the rest of the text in the paragraph.

4. Long- and short-term memory: Just like how the human brain has long-term and short-term memory and Neural networks have LSTM networks, which are implemented with the help of gates like XOR, etc.

5. Steganography: The process of hiding data within a picture. Yes, you might have flashes of some movies in mind that used or showcased this technique.

 

For Nagabhandhanam, I think introduction

of Sanskrit behind the encryption algorithms can break the shackles.

Why Sanskrit

- has multiple synonyms for a single word.

- Technical and Vocal Magic Language

- large corpus of words

- Ancestor of Many Languages

- More data can be represented in a smaller number of words.

- Patterns with respect to the pronunciation of words

- Through the research of NASA scientists, it was proven that the most useful language for computers and best for the creation of different algorithms.

In the field of cyber security, it is important to have an algorithm. Which has both encryption and decryption at its ease. As we know, generally,

Every lock has to have a key.

With the advent of Google Gemini, where voice is taken as input, this can be the future.

There will certainly be a day where you have to say Khul-ja-Sim-Sim or chant any mantra to decode the things!!

 

spacer

Smart Talk: What if Language Models like Chatgpt Learn and Behave Like Humans

Think of what if Chatgpt can also have some behavioral patterns like the human

Habit

Like Chatgpt doing anything for 15 days, it becomes a habit.

Mirror Neurons

How about Chatgpt is empathizing with the Google Bard like you make a mistake and I penalize, and I have done something great and I am awarded

Assumptions

Anyway, Chatgpt has the problem of hallucinations and would go away in the mere future

Cognitive Dissonance

How about Chatgpt and Google Bard have conflicting beliefs and debate over And finally, one wins and the other compromises.

Conformation bias

How about Chatgpt asking Google Bard when it is not sure about a particular answer.

Serial Position Effect

How about Chatgpt remembers the first piece of information (primary effect)? and last piece (recency effect) of information better than those in the middle.

Overconfidence bias

What if Chatgpt thinks its ability and efficiency is more than bard? and starts comparing with Google Bard for the same question being asked.

Wiseness

How about Chatgpt? Learn from Google Bards mistakes!

---------------------------------------------------------------------------------------------------------------

The main idea behind this blog is that companies are now coming with different Large Language Models (LLMs), i.e., Chatgpt, Google Bard, Claude, Bing Chat, Google Gemini, etc. So the Large Language Models (LLMs) basically are nothing but a huge knowledge base with zero emotions. Though Humans are the smartest creatures in the world, but there are other creatures as well, which may not be as smart as humans but certainly exists, like school of fishes, flock of birds, which certainly do not have as big brains as human, but they do survive well, and surprisingly, we can apply the same way of their thinking to the set of Large Language Models. What is seen behind the scenes is a technique where Imagine that school of fishes are searching for the food. Then each individual fish keeps updating its position based on the nearest fish to the food, Here, they have personal best, global best (best among the school of fishes), and some inertia.

The below diagram illustrates the mechanism



The process goes on till they find food!

So, now mapping it to our use case

fish1: Chatgpt

fish2: Google Bard

fish3: Bing chat

….

fish N - XYZ


Food: Answer of the question or prompt given.

We will certainly end up getting better answers than the existing system.

Is it achievable ?

Yes it is if we add any hyperparameters to the use case certainly possible

spacer