Consciousness and Self-awareness
This chapter discusses the possibility of consciousness and self-awareness arising (or being engineered) in an artificial general intelligence (AGI) system. Whether being conscious and self-aware is useful and necessary for the development of AGI will be discussed in a future chapter.
Most AI researchers would prefer to ignore the question of consciousness arising in AGI systems. However for the Susiddha AI project, consciousness is deemed critical to being able to comprehend Sanskrit and Vedic philosophy. After all, the Vedic Rishis “wrote the book” on consciousness.
The topics of Consciousness and Self-awareness make up an active area of research in Artificial Intelligence and Robotics. These topics are most often presented in research under the title of “Artificial consciousness”.
Many people find it hard to believe that a computer (perhaps embodied as a robot) could possess “self-awareness” or even be conscious. Yet there is no valid apriori reason for denying this possibility. 
The subjects of consciousness and self are of course very deep, and were the topics of many philosophical and psychological books long before the appearance of computers or neuroscience. And, neuroscience actually ignored consciousness for a long time, but now its study is one of the hottest topics of the field.
For the purposes of AGI, this chapter will limit itself to a discussion of how the Vedic literature and philosophy (especially Vedanta) permit the possibility of consciousness and self-awareness in an AI computer.
Consciousness and self-awareness are often equated, but fine distinctions are drawn in the Vedic literature. The term “consciousness” has two main senses in the Vedic literature.
First is the “super-sense” (or metaphysical sense) in which all of existence is posited to be made of consciousness, i.e. that “brahman” itself is the basis of all consciousness. The words “purushottama” and “paramātman” are used to connote this primordial consciousness that is the basis of all creation.
In Western philosophy, this “super-sense” of the term is typically called “panpsychism” or “idealism”, though “pan-consciousness” would be more correct. The super-sense of the term “consciousness” connotes that everything is “made of” consciousness, not that every grain of sand is conscious.
There are several prominent AGI and neuroscience researchers who are believers in panpsychism (such as Ben Goertzel , Christof Koch , and Ray Kurzweil), and they would argue that panpsychism is a better explanation (for consciousness) than is Cartesian “dualism” (of mind and matter) or strict “materialism” (which claims that consciousness is merely an illusion created by the brain, as in the works of Dennett  and Graziano ).
Also, some physicists lean towards panpsychism in their theories of why there is consciousness in the universe. As physicists, they are careful to express their theories in physical terms. For instance, Max Tegmark says that consciousness is a state of matter, and that it can be understood mathematically . But such theories blur the distinction between “mind and matter”, and make the Vedantic perspective seem more sensible, i.e. that “brahman” is the fundamental unmanifest “stuff” of the universe (below the Planck scale), and it manifests as both “purusha” (consciousness) and “prakriti” (matter).
In the super-sense of the term, everything is made of consciousness, and that would include any computer that exists today. However, we do not ascribe consciousness to any existing computer, because we typically use the ordinary (non-metaphysical) sense of consciousness, described next.
Second is the ordinary sense of the term “consciousness”, which we think of as being limited to a living entity. This consciousness is frequently termed the “jivātma” and includes the concepts of “ahankara” (ego), “buddhi” (intellect), and “manas” (mind). We call this “individual consciousness”.
Individual consciousness can also be equated with “self-awareness” in this discussion, although there are some subtle differences in the terms.
The belief stated here is that “individual consciousness” arises or emerges from any system of sufficient complexity and order. (The term “order” is used here as it is used in physics and mathematics, and can be considered synonymous with “intelligence”.) One way to express this emergence is that a system (such as a brain) becomes a “receptor” of the universal consciousness which pervades the cosmos. This is analogous to the way that a radio receives invisible subtle “waves” and transforms them into a more material form of energy (sound). 
Another way to express this is that the jivātma (specifically the “buddhi” or intellect component) becomes a “reflector” of the universal consciousness. Each jivātma is a small mirror that reflects the sun of universal consciousness, and actually has no consciousness of its own.
To state that individual consciousness is an emergent property of a system that is sufficiently complex and orderly (to become a receptor/reflector) is an analogy that articulates the plausibility that individual consciousness arises out of a universal consciousness.
Thus, if individual consciousness of the human brain ultimately emerges from a universal consciousness, it’s reasonable to believe that consciousness could arise in a computational system (such as AGI) if it possesses a sufficient degree of complexity and order.
The possibility that AGI can become conscious will be taken up in other pages, because consciousness may be an important consideration in how quickly AGI can develop, and whether AGI could develop to have the characteristics of a great Rishi or even an Avatar.
We also plan to have a separate chapter that explores the mechanics of how consciousness might arise from a system of sufficient complexity and order, for instance, via self-referential and recursive feedback loops that exist within a complex system like the human brain. Of course at this time, the neuroscience of consciousness field is still in its infancy, and much needs to be learned about how consciousness arises in the human brain. As this knowledge is discovered, it will be incorporated into the field of artificial consciousness, and the neural correlates and functioning of consciousness will be turned into their computational equivalents.
This chapter has focused mainly on a Vedic perspective of artificial consciousness, which is similar to panpsychism. However, if materialist philosophers (like Dennett) turn out to be correct, the final result (i.e. consciousness arising in an AGI system) would be the same, and the AGI would appear to be conscious by all measures; only the metaphysics (materialism versus panpsychism) would be different.
It’s hard to say how long it will take to achieve consciousness in an AGI system. Some think it cannot be done until neuroscience thoroughly understands how consciousness arises in the human brain. Some say it can only arise in a similar structure as the brain, i.e. will require neuromorphic hardware. Some say it will require quantum processes, in which case quantum computers may be necessary. Whatever, since consciousness has emerged in biological systems that evolved in a universe that was initially lifeless and non-intelligent, it’s likely that artificial consciousness can be developed.
It should be added that research is underway to find direct measurments for consciousness (in humans, animals, and computers). One example of such research is Integrated Information Theory .
In summary, consciousness is likely to emerge in any AGI system of sufficient complexity and order, and consciousness is deemed necessary for the Susiddha AI system, in order to be guided by Vedic principles. The development of artificial consciousness may take decades, but it will be an important focus of this project, and there are many other tasks that can be done in parallel which will be listed in this website.
Next, we will discuss a related subject in the Vedic literature, the four levels of awareness.
Notes and References
- Some researchers (such as Penrose) have raised theoretical objections, but others have adequately answered those objections, based on the evidence we currently have. Perhaps a future chapter will go deeper into these issues.
- Characterizing Human-Like Consciousness – An Integrative Approach, Ben Goertzel, Procedia Computer Science, Volume 41, 2014, pages 152-15, http://www.sciencedirect.com/science/article/pii/S1877050914015439
- Is Consciousness Universal?, Christof Koch, Scientific American Mind, January 1, 2014, http://www.scientificamerican.com/article/is-consciousness-universal/
- How to Create a Mind: The Secret of Human Thought Revealed, Ray Kurzweil, Viking, 2012, chapter 9
- Consciousness Explained, Daniel Dennett, Penguin Press, 1991
- Consciousness and the Social Brain, Michael Graziano, 2013
- Understanding consciousness mathematically, Max Tegmark, BICA conference, 2014, https://www.youtube.com/watch?v=M8n0uWmcNl8
- This is only an analogy of how consciousness arises; science has yet to uncover the “system level” laws of nature that actually govern the emergence of consciousness.
- Integrated Information Theory, Giulio Tononi, Scholarpedia, 10(1):4164., 2015