Vedanta and machine consciousness

In a previous post, we had seen how the concept of consciousness differs between Indian philosophy rooted in the Vedas, and modern Western scientific models. Specifically, while Western science models consciousness as an attribute of a being, Indian philosophy models consciousness as the fundamental substance of the universe. Beings are considered to reflect this universal consciousness to various degrees, and not have consciousness themselves. 

To explain this, we had given the analogy of the space just outside of the Sun that appears dark, but which is full of light. We cannot see this light, unless there is something to reflect it. 

Image source: Pinterest

In this post, we will delve into the notion of consciousness further. If all we do is reflect universal consciousness, then what makes us appear conscious, and say a stone, appear unconscious? What kind of beings reflect this universal consciousness, and what beings don't? 

Indian philosophy has an answer to this too. Consider objects around us in broad daylight. All of them reflect sunlight-- much in the same way that all beings (living or non-living) reflect consciousness. However, on a few surfaces like glass, water, polished metal etc. we not only can see sunlight reflected off them, but also an image of the Sun itself-- the source of the light. 

This metaphor helps to explain sentience and non-sentience in systems of being. Sentient beings not only reflect consciousness, they also can-- to different extents-- form an image of the universe itself! 

By this model, when we consciously create something, we are actually channelizing or using the universal consciousness into some creation of ours-- somewhat like the image below. Our creation may collapse eventually, but the universal consciousness remains eternally. 

Image source: Pinterest

Sentience is not seen as a binary attribute-- as in, it is either present or absent. Sentience is a continuum with beings like humans having several orders of magnitude more sentience than say, a rock. Hence, while both rock and human can reflect consciousness, humans can also sometimes offer a glimpse into the very source of this universal consciousness. 

Beings are said to comprise of two kinds of "bodies"-- called the sthula sharira and the sookshma sharira translating roughly to "material body" and "subtle body". In my book called The Theory of Being. I had called this the two "realms" of existence of any being in this universe-- the material realm and the information realm. 

The material body is our physical body that exists in the physical universe. The "subtle body" is variously (and in my opinion, incorrectly) equated with terms like "soul" or "spirit" represents all the information related processing that is part of who we are. 

The subtle body in turn, is said to comprise of three components: pranamaya kosa, manomaya kosa, and vijnanamaya kosa. The pranamaya kosa represents all forms of information constructs that keep our material bodies alive. They contain our emotional dispositions, defence mechanisms, fight or flight responses, internal communication protocols for signalling danger, hunger, pain, etc. The manomaya kosa represents our "mind" that implements our "sense of self" and ego. This aspect of the subtle body encodes a representation of our self image, who we think we are, all the persona of ourselves that we project in social settings, what we desire, what we despise, etc. The vijnanamaya kosa represents our objective knowledge about the universe. It contains our models of reality around us, general knowledge constructs, episodic memory, etc. 

Sentient beings that can reflect an image of consciousness, rather than just scatter consciousness, are said to have highly evolved subtle bodies. A subtle body is (in mechanistic terms) essentially a large body of software that can keep a material body alive, build models of the external world and manage knowledge about it, and have a "sense of self". 

Of these, we already can build machines that can do the first two operations very well. We can build robots that are adaptive and self-configuring in order to keep themselves going. We also can build AI that can build semantic models based on data, which it can then use to perform a variety of intelligent tasks. 

The only missing element in building a full-fledged subtle body for our machines, seems to be a "sense of self". Hmmm.. What does it mean for machines to have a sense of self? What exactly do we mean by a "sense of self"?

A few weeks ago, I had given a talk on this very question, which may be found here: 

Hope you enjoy the presentation!