Abstract
This paper explores the historical and philosophical lineage of Artificial Intelligence, tracing its evolution from ancient myths and religious taboos to its formal establishment as a scientific discipline at the 1956 Dartmouth Conference. By examining the shift from energy-based physics to information-based cybernetics, the work frames AI as the realization of a long-standing human urge to replicate intelligence through rigorous scientific inquiry.
Introduction: From Myth to Science
The history of Artificial Intelligence is often treated as a modern phenomenon, beginning with the advent of the digital computer. However, as Pamela McCorduck argues, the field represents the moment a "human dream" metamorphosed into a formal science. This transition was not merely a technological milestone but a profound cultural shift, moving the concept of man-made intelligence from the realm of "make-believe" and legend into a structured scientific framework.
The Dual Heritage: Hellenic Curiosity and Hebraic Prohibition
McCorduck identifies a central tension in the human psyche that governed the pre-history of AI, categorized as the Hellenic versus the Hebraic tradition.
The Hellenic urge is characterized by curiosity and an enthusiasm for creating life-like machines. This is exemplified by the myths of Hephaestus, the Greek god of the forge, who constructed golden handmaidens to assist him. In this view, the creation of intelligence is a natural extension of human ingenuity and craftsmanship.
Conversely, the Hebraic tradition presents a prohibition against such creations. Rooted in the Second Commandment’s warning against "graven images," this perspective frames the attempt to create intelligence as a blasphemous act of hubris. Historically, opponents of AI have often relied on these ancient religious and moral arguments to frame the field as either wicked or fraudulent. The emergence of AI as a science required navigating—and eventually moving past—these deep-seated cultural anxieties.
Pre-Electronic Foundations
While the digital computer became the primary instrument for AI, the concept of the "thinking machine" predates electronic hardware by centuries. McCorduck highlights several key precursors:
- Ramon Lull (13th Century): Inspired by the Arabic zairja, Lull created the Ars Magna, a mechanical system intended to combine concepts and discover new truths through logical combinations.
- Charles Babbage: His work on the Analytical Engine in the 19th century laid the conceptual groundwork for programmable computation.
- Konrad Zuse: In 1945, Zuse designed Plankalkül, the first high-level programming language, demonstrating that the logic of intelligence could be codified independently of the specific hardware used to execute it.
The Paradigm Shift: From Energy to Information
The catalyst that allowed AI to emerge specifically in the mid-1950s was a fundamental paradigm shift in the scientific community. For centuries, physics had been dominated by the study of energy. However, the rise of cybernetics introduced a new primary unit of analysis: information.
This shift allowed researchers to view the computer not just as a calculator, but as a "symbol manipulator." The computer became the "instrument of choice" because it provided a medium where the ancient dream of man-made intelligence could be tested and simulated with precision.
The Birth of the Discipline: Dartmouth 1956
The formal launch of Artificial Intelligence as a distinct field occurred at the 1956 Dartmouth Conference. Organized by John McCarthy, Marvin Minsky, Nathaniel Rochester, and Claude Shannon, the event was supported by a modest $7,500 grant from the Rockefeller Foundation.
It was during this period that John McCarthy coined the term "Artificial Intelligence." He chose this nomenclature specifically to differentiate the work from "cybernetics" or "automata theory," which were the prevailing terms of the time. The conference established the core conjecture that remains the bedrock of the field: that every aspect of learning or any other feature of intelligence can, in principle, be so precisely described that a machine can be made to simulate it.
Conclusion: The Unfinished Agenda
McCorduck observes that the origins of most scientific fields often appear "loony" to their contemporaries. The "lunacy" of early AI is simply a reflection of a mysterious but ancient urge to reproduce ourselves. Despite the progress made since 1956, the fundamental goals outlined in the original Dartmouth proposal remain largely unfulfilled. If that same proposal were submitted today, it would still receive a serious reading, suggesting that while the field has moved into "different territory," the primary challenges of simulating human intelligence remain as provocative as they were at the field's inception.
Summary updated