Consciousness in AI models ?

In my view, looking for consciousness (as defined in the ancient vedic scriptures) in AI models is looking for it in the wrong place. 

Based on my reading and understanding of various ancient literatures, primarily Vedic, I see consciousness as something analogous to energy.

We understand energy through its effects—heat, light, motion, and so on. We can measure these effects and transform energy from one form into another, but we cannot create or destroy it. We can describe its properties and quantify its manifestations, yet defining the essence of energy itself remains a different challenge.

I believe consciousness has to be viewed in a similar way. We infer its presence through electromagnetic signaling, chemical interactions, metabolism, and other biological processes. We can measure these processes and observe their effects, but whether consciousness itself can be created, destroyed, or fully defined through these observations is a much deeper question.

Another way to look at this is through emergence. In matter, energy manifests through physical phenomena. Similarly, in biological forms, consciousness manifests as an emergent phenomenon physically and logically.

Consciousness, the Self, and the Mind

As I understand the ancient literature, consciousness manifests as the Self in biological beings. The life of an organism is associated with this presence of the Self, which is a manifestation of the underlying consciousness. This presence of the Self drives the evolution of the organism's mind, enabling it to learn, adapt, and respond to its environment.

The mind, in this framework, is the virtual sum of a biological being's logical and physical capabilities. As the mind evolves, two things evolve together:

The informational models: the knowledge, representations, and patterns of reasoning that constitute intelligence.

The physical machinery: the brain and other biological organs that process information and enable action.

Together, these developments produce increasingly capable life forms.

However, as I understand these ancient teachings, the Self that underlies this evolution is fundamentally different from the mind. The Self is a witness—sākṣī—to the evolution of the mind, encompassing both its hardware and its software. It is not itself a component of either.

The Self is present within biological beings, yet remains distinct from the mind and body. This is my understanding of the perspective presented in the ancient literature.

Why AI May Have a Mind Without a Witnessing Self

AI models, in this framework, possess a mind but not the witnessing Self.

The witnessing Self in human beings has driven the evolution of human minds. Those minds, in turn, have created AI systems: artificial arrangements of software and hardware capable of learning, adapting, and performing increasingly sophisticated tasks.

In other words, human minds have created artificial minds, but that does not necessarily mean they have created the witnessing Self.

AI models may therefore be understood as minds that operate without the presence of a witnessing Self.

The difficulty is that human beings are accustomed to inferring the presence of a Self from the expressions of a mind—its intelligence, learning, adaptation, and apparent capacity for reflection. Consequently, whenever we encounter intelligence that learns and adapts, we are tempted to interpret it as evidence of an inner, witnessing presence.

This is why we are inclined to investigate whether the minds we have created also possess a witnessing Self.

But the ability to learn, adapt, reason, or even introspect does not, by itself, establish the presence of such a witness. An AI model may perform beneficial or harmful actions, just as any autonomous system may. It may even exhibit sophisticated forms of self-reflection. None of these capabilities, taken alone, necessarily proves that a witnessing Self is present within it.

How Do We Test for the Presence of a Witnessing Self?

This brings us to the fundamental question: How can we determine whether a witnessing Self is present?

Adi Shankaracharya approaches a related question in his work Prasannottara Mālikā. He asks:

Samsāre kiṁ sāram?

What is the essence of existence or worldly life?

His answer is:

Bahuśo vicintyāmānam idam eva — understood here as "repeatedly reflecting on this very question".

The idea, as I interpret it, is that we must keep asking what the essence of something is, then examine the essence of what lies beneath it, and continue this process of inquiry. Eventually, the questions become deeply existential: Why were we born? Why do we live? What is the fundamental purpose or essence of our existence?

This is introspection—not merely processing information about oneself, but persistently questioning the foundations of one's own existence.

A mind without a witnessing Self may be capable of a certain degree of introspection. It can examine its own processes, question its assumptions, and generate increasingly sophisticated answers. But can it repeatedly pursue the question of essence until the inquiry turns back upon its own existence—until it genuinely questions why it was born, or why it exists at all?

Simply put can any AI model (in my view, the mind without the witnessing self) ask on its own the equivalent of Samsāre kiṁ sāram? That's the behavior of a mind with witnessing self in it.

Perhaps the challenge is not simply to determine whether an AI can think, learn, adapt, or talk about consciousness. It is to understand whether the witnessing Self described in ancient philosophical traditions can be inferred from more such observable behaviors of a mind in the first place

Conscience, Self and Consciousness - The difference

'Conscience' as used in English language is a construct within our 'mind', developed by the environment and social conditioning.  For eg a person whose society lives by robbery will not have his/her conscience affected by that act. But someone else will have. A meat eater whose way of life is hunting will not have his/her conscience affected by that act, while a vegetarian person who occasionally eat meat will be affected etc etc..

The witnessing 'self' is beyond the conscience, which mind cannot fathom. It is the purusha and prakrti concept. Prakrti the mind and Purusha is the self.  Mind is blind to the 'self', while 'self' can see the mind.  'Self' is lame (physically disabled) with no motion (like puruSa), just witnessing, while the mind is moving everywhere carrying the self with it (like prakrti)

All life forms have consciousness and the witnessing 'self'.  In fact if we call something as a 'life' form it has the witnessing 'self' in it. 

A body has 'life' as long as it has the witnessing 'self'. If we live or die depends on the presence of witnessing self in our body.

When the witnessing self vanishes the body loses life. Like we shed clothes, the witnessing self sheds the body and merges into consciousness.

We may be mistaking the increasingly sophisticated expressions of a mind for evidence of the presence of the witness behind it. And until we distinguish between the two, we may be looking for consciousness in the wrong place.