The Gita, AI and I

One wintry evening in Delhi I received an unusual call from New York. At the other end was the soft voice of a charming man, representing a company that was bringing out the world’s classics using Artificial Intelligence (AI).

The book they had chosen from India was the Bhagavad Gita, and he wanted to know whether I would be willing to be the author, a sort of guide for the book. 

“But what is this animal supposed to be at the end — creating a classic using AI?” I asked skeptically, also wondering if the voice at the other end was human. 

He explained gently that readers would be encouraged to pause while reading the text and ask a question if they had a doubt.

At the end of each chapter, they would engage in a discussion. The answers would come from my cloned voice based on twelve to fifteen hours of interviews conducted with me in advance.

Fascination and horror           

My reaction was one of fascination and horror. I was instantly captivated by the possibilities of such an interactive book — how this might breathe life into the classics.

In fact, I wished I had had such a personal tutor while reading the Gita when I was young. Or when I was reading those incomprehensible Germans — Hegel and Kant — when studying philosophy at the university.

But just as quickly I was overcome with doubts and suspicions. AI opened in my mind troubling questions of  identity, selfhood and authenticity. I felt uncomfortable being cloned, having a twin suddenly.

My questions about AI

Ironically, these are the same questions — Who am I? — that the Gita also addresses. Who indeed will be my cloned twin providing answers to questions about human identity in this AI book? 

The friendly voice at the other end added that their company had already published half a dozen classics. Well-known names like Salman Rushdie and Margaret Atwood had signed up for future classics.

I was flattered to be in such august company. But I told him that they might have made a mistake in choosing me. The guide for the Gita would more appropriately be a Sanskrit scholar who had spent a lifetime interpreting the text or a spiritual Guru.

I was neither. I was an agnostic, in fact, and although I admired the Gita, I did not fully buy its message.  “There are plenty of Gurus and scholars around,” I added. “Perhaps, I can suggest some names,” I said. 

What a stroke of luck!

No, I was the right person, he insisted. Wasn’t I the one who had brought the great epic of India, the Mahabharata, to life in my book, The Difficulty of Being Good. He wanted me to do the same with the Gita.

A scholar or a Guru, he felt, would take the audience back to the “wonder that was ancient India.”  He wanted someone to bring the Gita to the 21st century. Sensing my hesitation, he suggested I think about it and we could speak in a week or two. 

When I recounted my conversation to my wife over dinner, she was surprised that I was hesitating. While pouring a liberal helping of aubergine curry on her basmati rice, she exclaimed, “How lucky can you get! Imagine spending the last years of your life unravelling the mysteries of the great book of India.” 

“But it’s a deeply religious book, and most of it is directly from God’s mouth!” I don’t even know if God or anything transcendental exists.”  

She knew that I had never been able to make the leap of faith. She could hear the angst in my voice and tried to reassure me. 

She added that I was attracted to the Mahabharata, the fifth Veda. “The Gita is, after all, embedded in the epic.” 

As we rose from the dining table, she repeated, “What a stroke of luck to have this fall in your lap!”

What new could I offer?

She was right, of course. It didn’t take long to persuade me that this was an opportunity of a lifetime. To read the Gita at this stage in my life was a privilege given to few.

But I was still bothered. I was daunted by the hundreds of commentaries on the Gita over the centuries. In an argumentative country everyone had his own opinion of the Gita. What new could I offer to the reader?

My three interns

In the spirit of the age, I acquired three AI interns to assist me with research on the Gita: Chat GPT, Perplexity and Claude.

ChatGPT is the most empathetic, Perplexity is best at research and gives links to follow-up and Claude has a literary style. Gemini is a good standby.

They are not only surprisingly competent but cheerfully, tirelessly deliver the results at the speed of light. When I have occasionally pointed out that something is not quite right, they readily try again.

There is no hint of complaint or the need of approval. Each time they reply with calm assurance and keep getting better at their work. What I got last week was better than six months ago.

I feel grateful that they take away so much of the drudgery of looking for a specific text, cross-referencing it and even summarizing it before I decide to read the actual book. 

Gandhi and Himmler

With the invisible labor of research eased, I had more time to think about philosophical questions posed by the Gita. I find myself playing thought games.

“What if the Buddha had been Arjuna’s charioteer?” is one of them. Another is a paradox: How does one explain that MK Gandhi, the apostle of peace and non-violence, and Heinrich Himmler, the Nazi apostle of war and violence, were both  inspired by the Gita?

Himmler trained SS officers quoting from the Gita.  Gandhi did the same with volunteers fighting for India’s freedom.

But I also have lingering worries about my love affair with my three interns. Much as I feel gratitude, I feel somehow less connected to their findings. And I am clearly troubled by my lack of ownership for some of their conclusions.

I try and reassure myself, saying “Oh it’s only me; I’m just growing old.” But no, I find an alienating distance from the AI output. It would have felt different if I had patiently done the work myself. 

AI as one’s operating system

I think while I write. My thoughts are fuzzy before I put them down but they get clearer as I read them in a proper sentence. Writing for me has become a tool for thinking.

I am also a compulsive re-writer. My wife feels I waste a lot of time re-writing, but I find that both the writing and the thinking becomes clearer in the end.

There is great reward to find an elusive thought expressed in a clear, original brief and bold sentence. AI in comparison seems to flatten thought. It is a little bit too easy, too fluent, too bland. 

Thinking in the age of AI

Thinking is an individual lived experience. I worry about the young who rely too much on AI and do not practice their mental muscles while growing up.

As for me, AI has opened up profound questions about the self, identity and authenticity. “Who am I?” is a question that I feel is a lived experience, not to be answered by a machine.

Two thousand years ago the Gita too wrestled, oddly enough, with the same metaphysical problem of the self.  It concluded that our normal human belief in our individuality is, in fact, the problem. 

Identity as an illusory perception

The Gita claims that my day-to-day sense of identity is an illusory perception created by the human ego. It posits that there is a true, permanent self or soul (atman) which underlies our transient, perceiving selves, and it is identical with the cosmic spirit (brahman).

Awareness of the oneness of everything is, thus, the Gita’s central teaching. Human beings mistakenly identify their minds and bodies with reality.

If one agrees with the Gita, the exalted notion of individuality or the lived experience of an individual, so prized especially by modernity, turns out to be false. The Gita calls it “maya,” an illusion.

Both the Gita and the AI have made me doubly aware that the exalted notion of the human individual of my modern, liberal life is fabricated. Grappling with these two fictional selves has left me totally befuddled.

Three aspects of being human

AI raises fundamental questions about what it is to be human. The only things I can be sure of at this moment are the thoughts and feelings going through my head.

If I look into my consciousness, I cannot find the thinker of these thoughts, nor the feeler of these feelings. The Buddha faced this problem as well but reached the opposite conclusion to the Gita.

Both Eastern and Western philosophers agree that we can never know the “thing in itself,” as Kant put it.  We can only know our own perceptions — and yours are different from mine.

The Gita, of course, goes further. It calls our world of subjective perceptions an illusion.  On a lighter note, this sort of thinking might  have led Mark Twain to quip that there is no difference between fiction and non-fiction — only fiction has to make sense! 

Sentient AI?

The fear that AI might become sentient one day is real. Before that happens, however, AI has formidable hurdles to cross.

One of them is that human experience is inherently subjective. It cannot be explained in objective terms.

A second hurdle for AI to achieve consciousness is the need to replicate my brain’s biological neuron-specific physiology. Our subjective experience is based on a human brain’s nearly 86 billion neurons and over 500 trillion synapses.

Consciousness is an emergent property of the specific biological matter of my brain and a silicon-based system, no matter how complex, will never be able to replicate them.

Spontaneity as a human USP

Both these hurdles lead to a third one. A vital property of consciousness is spontaneity, a freedom to choose my goals and objectives. This is central to the human experience.  

AI has so far achieved an excellent ability to follow commands. The autonomy that underlies spontaneity is what the Gita calls atman, an intangible “life force,” which is also the self-awareness behind all thought, intention and action. 

AI, as it currently exists, operates based on algorithms and data without conscious awareness or purposes of its own. To become truly sentient, AI still has far to go to achieve  this spontaneity and freedom.