AI And Sanskrit

By Dakshmitha G

Artificial Intelligence has pervaded many fields today.

Did you know that there's a new artificially intelligent oreo? It was one smart cookie!

The quest for advancement in AI is a never ending one. There are new technologies being invented and developed as we speak. But the one challenge that AI has always been facing is Ethics.

Throughout the process of developing AI, the number of ethical questions that arose against it about unemployment, equality and fatal mistakes that might happen, among others have been countless. Hence a method to weed out these flaws and make AI more compatible with humanity has become essential. While there are numerous ways to go about this, one of the more revolutionary techniques is to use Vēdās ( literally knowledge in Sanskrit ) to teach AI ethics.

The Vēdās are four in number : Rig Vēdā , Yajur Vēdā, Sāma Vēdā and Atharvana Vēdā. They've been orally transmitted since the 2nd millennium BCE in Vedic Sanskrit. They are considered to be not of a man, superhuman (aparuseya) and impersonal, authorless and are revelations of sacred sounds and texts heard by ancient sages after intense meditation.

Hinduism has numerous schools of thought that interpret the Vēdās. One such school is the Mimamsa. This system is one of the earliest of the schools and has deeply influenced the formulation of Hindu Laws. It literally means Reflection or Critical Thinking in Sanskrit and uses an extremely rigorous approach to analyse the prohibitions and obligations mentioned in the Vēdās to support their arguments. It is this method of argument that got researchers interested in using the Vēdās to teach ethics to AI.

In the 7th century CE, a Mimamsa scholar named Prabhakara applied an intricate set of logical rules to support his argument. Other scholars refused to believe his reasoning and the matter was left unresolved. However, when researchers at Vienna translated Prabhakara’s Mimamsa arguments into mathematical rules, they discovered that he was right all along. Prabhakara’s logic was flawless.

“We want to understand how to express with mathematical precision beneficial logics dealing with prohibitions and obligations,” says Agata Ciabattoni, supervisor of the research project.

Classical logic only deals with “true” or “false” statements. However, this is useless in real-world situations. What is required for building an ethical system in an AI is something called “deontic logic” that deals with a set of obligations and prohibitions that have to be followed when certain events are triggered.

Several people have tried creating a system of perfect deontic logic, only to eventually fail. But with Vedas and Mimamsa, Agata believes that his team can create a workable deontic logic system.

Not just that but Sanskrit is considered to be a language most well-built and extremely well suited for computers that are communicating via Artificial Intelligence.

Whenever we speak any language, we must get out our ideas by understanding the context of the speech and then formulate a sentence by using the rules of the grammar of the language. It is in this aspect of the grammatical construction of the sentence that Sanskrit outshines any other language.

We should understand that computers, unlike humans, do not actually understand language. They are computational devices or devices that are primarily governed by the rules of Mathematics. Thus, these machines must convert human speech into rules of mathematics to make a clear sense of what is being communicated.

So how will computers comprehend these languages then? It is done by the usages of matrices (ie) Matrix which is a very good method to store and redistribute data that is related via a function, method or pattern. Using this, sentences are made into two and three dimensional matrices which can encompass all nouns and verbs.

There are two other things that make Sanskrit suitable for this:

1. It is phonetically consistent - that is, each word that is written is spoken out as it is. There are no exceptions at all. Also, if two words are written very similarly, we can be assured that they will be spoken in the same way as well.

2. All the words follow the rules of grammar to the tee and there are no exceptions. This means that a computer can take a very new word, and after analysis if the word is a noun or a verb, it can be put into a matrix and can be used without the chance of a grammatical error.

Hence, computers can very easily take in new words and immediately assimilate that into usage. This makes Sanskrit an excellent option for natural language processing and eventually AI.

There are many scientists and intellectuals that doubt this matter – the natural language conversion for AI technology. Most of them argue about Lisp which is the language of Artificial intelligence. They also claim that if the Sanskrit language is considered suitable, then software should have been written by now.

On the contrary, the majority of the scientists do believe that Sanskrit has many useful lessons and techniques that can be used for the development of a computing language, but the integration is quite difficult.

Overall, AI development has yet another path to discover and build upon using the virtues of Sanskrit. If the tools and values of Sanskrit Literature and the language's structure is used properly, it might as well become the next big revolution in the world of Artificial Intelligence!!

By Dakshmitha G