Nexus: A Brief History of Information Networks from the Stone Age to AI

In Chapter 4, Harari discusses “the fallibility of human beings” (70) and the need for self-correction in various systems, including mythology, politics, and bureaucracy. Humans have long sought error-free mechanisms to identify and correct mistakes, often imagining superhuman solutions, such as divine guidance in religion or AI in the modern world. Harari criticizes this desire for infallibility by highlighting how religion has historically provided legitimacy to social orders through claims of divine, error-free authority. 

However, even religious systems are ultimately reliant on fallible humans, such as prophets, priests, oracles, and spiritual leaders, whose messages can be corrupted or misinterpreted. Religious institutions, like priests and oracles, were established to vet and stabilize claims of divine communication, but they, too, were subject to error and corruption. This dilemma, present in religion, holds relevance to modern debates about AI and the desire for a “truth-seeking” (71) technology that could bypass human fallibility.

Religious texts like the Bible and the Quran serve as a technology to bypass human fallibility, providing fixed, identical texts that are intended to preserve divine words across time and space. Unlike oral traditions, bureaucratic documents, or archives, a book contains a consistent set of texts that can be reproduced identically, allowing widespread access to the same information.