Print This Post
27 May 2021, Gateway House

Natural intelligence is vital for AI ethics

On 25-26 May 2021, United Nations India held a Virtual Conference on Ethics of Artificial Intelligence (AI): Exploring pluri-perspectives in conjunction with OP Jindal University. Chaitanya Giri, Fellow, Space and Ocean Studies Programme, spoke on 'Operationalising AI ethics from a Pluri-Philosophical Lens', where he made the case for Sanskrit as an applicable language for instilling ethics in AI. He also emphasized the need for greater scientific inquiry in natural intelligence to avoid skewed technological progress with AI.

Former Fellow, Space and Ocean Studies Programme

post image

The remarks were delivered by the author at the Virtual Dialogue on the Ethics of Artificial Intelligence on 26 May 2021. 

India’s antiquity is intimately linked with the antiquity of Sanskrit. Over the years India has been home to opulent language dictionaries catalogued as Shabdkosh, Shabdbandh, the Amarkosh, the Dhatukosh, and many that grew of Sanskrit’s daughter languages. Languages grow prosperous over time. The prosperity comes via literature.  For India, Sanskrit’s antiquity and its motherly attribute as the progenitor of all Indo-European languages, make it the most abundant of languages. And there is a treasure-trove of Sanskrit available, that can be of great utility as we comprehend Artificial Intelligence and its ethics.

I am not a linguist, neither into epistemology, nor Artificial Intelligence. I am an Astrochemist, studying the chemistry taking place  in the universe, including the formation of life on planets that occur over billions of years. My study offers an atypical perspective on the evolution of life from inanimate prebiotic chemistry to sentient beings the way we are.

If we humans think in Modern English, today’s Lingua Franca, all  decent citizenry truly wants ethics to prevail. In common parlance, ethics evokes the correct-incorrect, right-wrong, and the moral-immoral binary.

But suppose we humans begin to use the massive lexicological expanse of Sanskrit. In that case, ethics become highly dependent on the settings it is used in. Let us choose a Sanskrit synonym for ethics – neeti – and do a metasearch of all words that uses it as a suffix – we come across so many of them. Rajneeti (statecraft)Ranneeti (warcraft)Arthneeti (economics) Dandneeti (lawfare)Kootneeti (strategy)Suneeti (benevolent ethics)Kuneeti (malevolent ethics) and many more.

Ethics in the reasoning of Sanskrit language, can have various dimensions and depend on the matrix in which it is viewed. The neeti of a carnivore, an herbivore, or an omnivore in an ecosystem is bound to be different. The difference is what maintains the order. Sanskrit, therefore, becomes key for the advancement of ethical AI regardless of its applications.

Sanskrit literature is a treasure-trove of Prakriti Darshana (the study of nature), evident from various treatises. This trove will be important as humans study ethics in the context of AI. This Prakriti Darshana is necessary because there is a lot to learn from nature and the various natural intelligence that has evolved before humans.

In today’s context, the term ‘intelligence’ is used quite loosely. We tend to call engineered systems operating on cause-and-feedback loop smartened by complex sets of instructions that go by the name of algorithms, as intelligent. Now that is a bit of a stretch. In reality, it is humans who have learned to develop better-engineered systems. We have spent thousands of human hours on such engineering and offered these systems billions of gigabytes of data. As a result, the systems now can compute a colossal number of tasks of varying complexity.

Significant technological progress in the algorithm-robotics-sensors trinity are making machines dexterous. They can now do unimaginable tasks much quicker. However, it has not reached a stage where AI has become emergent or acquired the capability to self-reproduce. The Self-Reproducing Automata, as imagined by the famous systems theorist Jon Von Neumann, is still far away. But that doesn’t mean a philosophical inquiry into understanding intelligence should not be delved into.

The distinction between the inanimate, animate, intelligent, sapient, and sentient is made in the Bhagavad Gita when Shri Krishna says, “The working senses are superior to inanimate matter; the mind is higher than such senses; intelligence is still higher than the mind, and the soul is higher than intelligence.”

Before going any further with AI and ethics, let us ponder, have we aptly termed AI?

We humans may be the most intelligent species on this planet. Still, we need not be under the misconception of being the only intelligent life-form on this planet. Intelligence on this planet rose at a time when 3.8 billion years ago, complex organic chemistry decided to undergo ’emergence’ and become ‘self-reproducing into unicellular and later into multicellular life-forms. The overarching intelligence-led biological life on Earth to pervade the planet, only to adapt and evolve to changing climes and environs. Human intelligence is a product of this overarching intelligence of all life on Earth.

And today, when we discuss AI and ethics, we are repeating the same mistake of anthropocentrism. Humans are not the centre of the universe, and indeed, AI, being an anthropogenic product, should not become one.

As a natural scientist, one also wonders, do ‘cancer cells’ have intelligence? Can swarming bees and birds be called intelligent? Can the billions of bacterial colonies present on different zones of a slimy microbial film growing on waterbodies cooperate for nutrition? The cognitive abilities of octopuses, apes, dolphins, and whales are well known.

When we study AI, we must also look at scientific communities that are part of endeavours known as ‘Search for Extraterrestrial Intelligence’ (SETI). That domain is a greater blackhole than AI. We do not worry so much about SETI because there is an uneducated certainty that humanity will encounter AI emergence well before ET Intelligence. But, what if we are wrong to presume that way?

There is so much intelligence out there. Are we humans doing anything to comprehend, interact, cooperate, and transact with these kinds of intelligence?

Nature has been kind that there are no apex predators above humans anymore. Now that we enjoy that apex position, we must still remember that no species has remained on the zenith of the intellect pyramid for too long. Hence, it will be wise to study the science of various kinds of natural intelligence before we mock ourselves with skewed technology advancement of AI.

The deliberation on natural intelligence is not vague. The Convention of Migratory Species, a respected body within the United Nations Environment Programme architecture, views dolphin intelligence seriously. Likewise, the Great Apes Survival Project, another UNEP project, takes the intelligence of gorillas, bonobos, chimpanzees, and orangutans with great diligence. Anti-whaling programs of the UN also give due significance to the brilliance of whales. The points made in this intervention therefore are parallel with the Goal Number 13, 14 and 15 of the UN Sustainable Development Goals.

The Prakriti Darshana in Indian philosophy has always emphasized natural intelligence. When multiple species help Lord Ram build the bridge to Lanka, it is evident. Even in modern literature, the Jungle Book by Rudyard Kipling, inspired by the Panchatantra and Jataka Mala, is full of Prakriti Darshana. India has a multi-millennia treasury of knowledge when it comes to the study of intelligence from nature. It is time to take it out from the vault, dust it, and open it.

Click here to view the event. Watch the speech here.

Chaitanya Giri is Fellow, Space and Ocean Studies Programme, Gateway House.

For interview requests with the author, or for permission to republish, please contact outreach@gatewayhouse.in.

© Copyright 2021 Gateway House: Indian Council on Global Relations. All rights reserved. Any unauthorized copying or reproduction is strictly prohibited.

TAGGED UNDER: , , ,