Ardor, by Roberto Calasso, Chap. 7, Ātman

“From the Ṛgveda to the Bhagavad Gītā a way of reasoning is developed that never acknowledges a single subject, but rather presupposes a dual subject. This is because the constitution of the mind is dual: consisting of a gaze that perceives (eats) the world and of a gaze that contemplates the gaze directed at the world. The first expression of this idea appears with the two birds in hymn 1.164 of the Ṛgveda: “Two birds, a couple of friends, are perched on the same tree. One of them eats the sweet berry of the pippal; the other, without eating, watches.” There is no more basic revelation than this. And the Ṛgveda presents it with the clarity of its enigmatic language. The dual constitution of the mind implies that two birds dwell perpetually within each of us: the Self, ātman, and the I, aham. Friends, alike, sitting on the tree at the same level, one might seem the double of the other. And so it is in the life of many, who never manage to distinguish between them. But, once their difference has been recognized, everything changes. Every moment becomes the superimposition of two perceptions that can add together, cancel each other out, multiply each other. When they multiply each other, according to the mysterious formula 1 × 1, thought springs forth. Even if, seen from outside, all remains the same. The answer seems still to be 1.

“Ātman, the Self, is a discovery. How to attain it was the ultimate doctrine for the disciples who had studied and assimilated all the Vedas. […]”

Calasso… Another giant I knew nothing about.

The subconscious, and the conscious. The gut decides, then, the head rationalizes. The veda, then, teaches the head to control the gut, or, at least, to understand it. How? Sanatkumāra says - you have to perfect your speech. Then, perfect your manas, mind, then, saṃkalpa, your intention, then dhyāna, meditation, and so on.

When have perfected this stack of domains, then you can cross the darkness - and achieve mokṣa, release.

Before release, if you can share all that was thus acquired, great! Release means losing it. You can’t step in the same river twice.

The western world is learning self knowledge very late. That is, perhaps, because, religiously, we’re inclined to self sacrifice, instead of self knowledge. And we’re individualistic, as Calasso says.

Being individuals is not bad - just different. It’s possible the Veda path to self-enlightenment was easier, while individualism in a caste system was hard. Stoicism was not individualistic either. It was a doctrine for dealing with the hand you were dealt.

“Self-referentiality, this movement of thought that was enough for Gödel to break up the whole edifice of formal systems, beginning with arithmetic…”

Gödel, Alonzo Church, Alan Turing. They worked on formal systems, the basis of computer languages. In the 1960s, the invention of the perceptron, then, in the 70s, deep learning. This last one is related to the subconscious.

To give an example, you learn to drive. It’s hard at first. You use your conscious mind, and it can do only one thing at a time. A car in front slows down, a passenger asks a question, the radio is on.

At some point, you “learn”. The driving function is passed to the subconscious. The conscious oversees. It intervenes when something unexpected happens. Deep learning learns the same way the subconscious learns. While no algorithm for consciousness is known.

Not only that, we don’t know what conscious “is”. The veda, as retold by Calasso, is as good as the latest scientific research on consciousness.

What computing can do is - learn from a large number of examples. It takes hundreds of thousands of examples to teach a robot how to catch a ball. A child, however, at some point, learns it effortless, in just a handful of tries. The child watches the world around, and infers physical rules, which objects are solid, which are not. When will an object stay in balance. The child is not taught these things.

We don’t know how consciousness factors into this learning. But we’re trying to replicate it with algorithms.

We know consciousness is single threaded. Which means it must be a function of a specific mechanism in the brain. And subconsciousness is multi threaded.

One day, perhaps not too far out, we’ll crack the problem of consciousness in computing. It’s easier to understand something you can build hands-on. Then, we may go back to teach an artificial consciousness all the levels of sophistication laid out by the veda, from how to perfect speech, to mind, intention, meditation, all the way to mokṣa, release.