A mythological Perspective on Artificial Intelligence

To most people, artificial intelligence is a modern marvel that became known to most on the 30th of November 2022 after OpenAI release ChatGPT as a free public demo. Its ability to generate images, engage in conversations, and generate human-like text made it an overnight sensation reaching millions of users all over the world. Unknown to most, the earliest AI-like ideas emerged, not from science as it is known today but from the roots of mythology.

The spark behind artificial intelligence dates back to c. 3000–1000 BCE, a period when ancient societies imagined the development of artificial beings that are endowed with autonomy, purpose, and intelligence. This idea was driven by human attempt to understand agency, creation and the need to control the nature of the mind. Contrary to the mechanical intelligence that we have today, artificial intelligence then was associated with divinely animated artifacts, an aspect that led to the development of an intuitive belief that intelligence must not be innate, as it could be constructed.  

In ancient Mesopotamia, gods were said to manufacture intelligent servants who would perform various duties on their behalf. For example, the Atrahasis Epic from the 18th century BCE explains how beings were created as part of engineered workers to engage in manual labor. It was widely believed that idols and statues could house divine consciousness that was capable of hearing, seeing, and acting as instructed. However, the created being could only partake in these activities after being activated ritually through the mīs pî (mouth-opening) ceremony. This myth framed intelligence as an entity that could be instilled inside specific kind of matter, which foreshadows later ideas of activation and programming in the present-day society.

In Ancient Egypt, great emphasis was placed on the belief that agency, intelligence, and action could inhabit artificial forms. Egyptians often understood animation as a ritualistic and technical process that was controlled by specific rules, procedures, and symbols. Similarly, temple statues were treated as living entities. They believed that conducting the “Opening of the Mouth” ritual appropriately following the provided procedural sequences would cause intelligence to indwell in a statue. Like present-day statues, the intelligence in these beings was not permanent as it was activated conditionally, and would function only when it was correctly initialized. This meant that agency and intelligence were not innate, but were transferable under strict observance of protocols.

One of the most technologically imaginative and explicit versions of AI dates back to the 3rd century BC. During this time, Talos a mythical figure was created by Hephaestus, the god of invention and blacksmiths and gifted to King Minos. Talos was a giant bronze automaton that was built with the intention of guarding Crete against pirates and invaders. It was often operated autonomously and patrolled the island three times a day in a bid to defend the inhabitants from invaders by hurling rocks at any approaching ships. Talos had a single vein in his body which was filled with ichor, the blood of gods and connected from the neck to the ankle. The divine fluid was sealed off with a bronze nail struck in the ankle to prevent the ichor from flowing out. Like modern-day robots, Talos enforced the laws of the law and the desired divine order by acting as a security system.

Fig 1: Apollonius of Rhodes' Talos (c. 500 B.C.)

Figure 1: Talos, an automaton created to defend the shores of Crete.

Later on in the 8th century BC, Hephaestus created Homer's Golden Maidens (c. 800 B.C.) which were far more developed golden servants who possessed cognitive abilities. Unlike modern-day AI assistants, the golden female assistants could speak, engage in skilled labor, and understand the provided commands. Contrary to Talos, these automata possessed a mind of their own, they had a voice and vigor as well as intelligence, presenting the earliest concept of artificial intelligent.

 However, the narrative around artificial intelligence changed when Pandora, the first human woman created by Hephaestus.

Figure 2: Pandora Box

Figure 2: A visual illustration of a Pandora Box

According to Greek mythology, Pandora, the woman was given a pandora box and was instructed by god Zeus not to open it. However, curiosity got the better of her and she opened the Pandora Box putting an end to the state of paradise that earth had enjoyed. The box released all kinds of troubles, diseases, and death, plaguing humankind for a lifetime. Only hope remained when she closed it again. Rethinking this tale, the Pandora introduces the idea that AI could carry unintentional consequences, which creates the need to consider modern AI ethics in its construction.

Modern View of Intelligence

The separation of intelligence from organic form aligns with modern perspectives of intelligence, best explained using the concept of substrate-independent. This line of thought explains that intelligence or consciousness often depend on functional organization, such as a story or software and not on the material that they run on. This perception explains that if the functional structure of a biological component could be replicated in a distinct medium, then the resulting artificial structure would possess similar mental properties.

In the ancient world, artificial intelligence involved well-aligned tools that were designed with the aim of preserving stability instead of disrupting it. This perspective aligns with the construction of modern AI. Although ancient understanding of artificial intelligence was mystical rather than mechanical, it created a perspective that influenced modern-day AI including task-specific agents, initialization of programs, and rule governed behavior that is guided by substrate-independent analogy. Additionally, artificial intelligence in the ancient world was understood as a controlled extension of divine order rather than a threat, a concept that resonates with present-day debates about AI systems.

Although these myths serve as more of fantasies, they encode philosophical beliefs that are reflected in the creation of modern-day AI. While they are mostly magical, they showcase the conceptual possibility of artificial intelligence as something that was imagined years before modern science was conceived.