Printed in the Fall 2023 issue of Quest magazine.
Citation: Crow, John L "Can Artificial Intelligence Embody the Divine Light?" Quest 111:4, pg 9 & 47
By John L. Crow
The Higher Ego is, as it were, a globe of pure divine light, a Unit from a higher plane, on which is no differentiation.
—H.P. Blavatsky
For the last few decades, anxiety about artificial intelligence has been increasing. We see it expressed in movies such as The Terminator and the Matrix series. According to Hollywood narratives, soon an AI system will gain independent consciousness and seek to eliminate humanity. Psychologists have called this fear AI anxiety, and it is classed alongside other modern fears, such as climate anxiety.
This anxiety has become rampant partly because of the complexity of modern technology. The average person simply cannot understand what is going on in the “black box” in which AI runs. This lack of understanding is compounded by statements from AI programmers saying even they don’t know or understand the resulting data and algorithms created by the AI once it is set in motion.
Despite the complexity of AI, computers are fundamentally the same as they been since the 1950s. They run commands given to them and wait idle until given the next command. Humans, however, can think independently because of the divine light, the spirit of life, descending from the Source.
The easiest way to understand why AI will never become conscious is to relate it to the golem, a creature found in Jewish folklore. According to the legends, a golem is a creature made by the Kabbalist entirely of earthen material, like clay. It is animated when the Kabbalist places paper or something similar within it containing Hebrew writing taken from the Torah and/or the name of God. Through the logos, that is, the words of God, the golem comes alive. It can move, carry objects, and perform tasks as commanded by its creator. But it cannot think for itself. It only moves to where it is told to go, and it only carries what it is commanded to carry. It completes all the tasks commanded to the best of its ability. If it fails to function correctly, it can learn how to perform the tasks better. Left unattended, it might wreak havoc, but in all cases, it would return to its inanimate state when the object with the writing was removed.
AI systems and its technology are similar to golems. Both are made of earthly materials (silicon being the base component of circuitry). Words bring the golem to life, just as programming brings AI into being. Most importantly, both appear to be alive, but are not. All the AI and the golem can do is perform the tasks given to them. Both can learn to do them better, but they can never develop independent thought. At their core, they lack the divine light that animates life.
This is not to say that these artifacts cannot do tasks given to them better than humans. Indeed both AI and golems are created to perform tasks that humans cannot do easily or do not want to do.
One task current AI systems are being created to perform is to process spoken and written language. As a result, the AI demonstrates amazing natural language abilities. These abilities have caused journalists and others to declare that these AI systems are alive. But these individuals overlook the fact that to gain these abilities, the AI system had to process millions of texts written by humans. The AI appears to be alive because it is doing a wonderful job of imitating those who are or were alive.
The algorithms and linguistic models created by these AI systems need to be “trained” on prewritten texts, which are frequently curated by the creators of the AI system. During this process, the AI system makes mistakes and is corrected by users. After millions of calculations and numerous corrections, the models allow it to predict the next word in the sentence based on the context and the surrounding words.
In addition to linguistically based AI, other forms can create images or videos, either by using a sentence description or by combining preexisting videos and images. They can be instructed to create a picture in the style of certain artists. If the system has processed and examined the artist’s work, it can make a facsimile in the same style showing what was requested of it
Like linguistic AI, these systems have been trained through images and video created by humans. They are masterful at storing and combining portions of pictures and videos. They can create simulations of works by well-known artists and directors, but that is because they were trained on the images and video created by humans, who, unlike AI programs, have the ability to be creative, invent new styles, and imagine new forms of art.
Like the golem, AI will never become conscious, because they are both the products of man. We lack the ability to bring the divine light into our creations. At best, AI are but temporary reflections of our divine light.
No doubt AI systems will continue to advance in complexity while becoming more useful in performing the tasks for which we create them. But they are our creations, and they will never become conscious. So we do not need to fear a future in which AI systems will seek to eliminate humanity. Without our divine light reflected in them, AI systems are as useless as a lump of clay.
John L. Crow, PhD is a religious studies scholar at Florida State University, where he teaches courses on religion, digital literacy, and technology use in online learning. His research focuses on the history of Eastern religions emerging in the West in the late nineteenth and early twentieth centuries. For more information, visit www.johnlcrow.com.