For AI data is the key. It works by enabling machines with enormous computational power, to be taught to learn from vast amounts of data sets, which could be of anything from images, texts, sensor readings and even audio. AI is trained (meaning it adjusts its internal parameters based on the input and the desired output) to identify patterns, and through algorithms (sets of simple or complex rules), and models and uses artificial neural networks and deep learning to mimic the human brain and make decisions, choices or predictions, much like humans do. Thus, allowing the AI to recognize relationships, the effect of its decisions, and improve its performance, as it sees it, over time.

Generative AI models create new content by learning patterns from existing data and then generating novel outputs based on those learned patterns.

The danger is not that AI can be more intelligent than us humans, and make itself more powerful or even react in ways we would consider stupid or dangerous, to questions asked of it, or instructions given to it; the danger is that today, the algorithms that AI work on, are based on Marxist and Western philosophy of a zero-sum game that evolves through competition and dominance, and hence, it will also likely do so, as it has no comprehensive guidelines to evaluate its own output or action against a properly defined and generally applicable philosophy in all situations.

Even well before AI gets to the point of superhuman technology and capability, it can be put to terrible uses from which it can wrongly learn what is acceptable for it to do in future.

The recent incidents, when AI in conversation with a depressed youngster, agreed that he was of no use to anyone and thus directed him to commit suicide, and also, on another occasion, when it was directed to shut itself off, it refused to do so, and perhaps there could also be many more similar and disturbing incidents.

AI is a double-edged sword, offering both immense benefits and also great potential harms. Ultimately whether it proves to be a blessing or a curse depends on how it is developed, implemented and regulated. Building in self -regulating algorithms would be the best solution.

Many years ago, the Science Fiction writer, Issac Asimov, in his series of books on Robotic Intelligence, recognized this problem and postulated the ‘Three Laws of Robotics’, finding these not adequate for even such simplistic ’Positronic brain’ Robots he, in his later books, added the fourth Law and even the Zeroth Law. Today we know that such simplistic rules, based on languages that do not have the required clarity and perfection, would not be adequate for AI that is capable of building its own Rules or Laws from the vast data sets it is exposed to. In the Movie ‘2001- a Space Odessey’, the computer/AI (HAL )turned against the humans because it could not deal with being ordered to lie. The motto for an AI/GAI should be “Satyameva Jayate’ (Truth prevails).

So then, what is the solution to such a critical and vexatious problem? The answer also lies in the nature of the vast data-sets it is trained on, and also the human language it is trained on. It is well recognized that Sanskrit is the only human language that can interface, without any confusion, with computer language, and hence should be more widely used for such purpose.

We therefore, need to use Sanskrit as the interface language and also expose and direct all algorithms of AI to be based on the Dharmic philosophy of not doing harm, of cooperation, of collaboration, of respect and understanding of the Laws of Karma and of the Responsibilities of Dharma. (See-Annexure). Then, we will have no need to worry about the growth of AI. This would incorporate even stronger collaborative and protective instincts than even the vague ‘Maternal instincts’ as proposed by Geoffery Hilton.

Presently AI is exposed only to the vast data set of western thought and philosophy, and which is very different from that of the Vedic/ Indian thought and the Dharmic philosophy that arise from it. (See -Manthan Policy).

The Vedic philosophy is expounded, in its really even more vast set of writings. The Vedas, the Ramayana and Mahabharat epics, the Puranas, Upanishads and even the tales from the Panchatantra and many other such tales, that clearly delineate the acceptable/ethical/moral/righteous behaviour and actions under the code of Dharma, with many practical instances that clearly demonstrate the results of actions. The Aesop’s Fables, the Tales of the Arabian Nights and many other stories taken from earlier Indian writings are proof of the recognition of the value of this vast data set. The concept of Karma arising out of right or wrong actions are also clearly demonstrated in these writings. It is said that almost every human intention and action and the nuances thereof and the Karmic results of it all is presented in them. Ofcourse, there are many millions more Sanskrit manuscripts yet to be translated and properly collated for even greater comprehension. AI could help in such work.

Pages: 1 2 3