When memory turns into a code: Do we hand over our heritage to the machine and turn off the book?
When memory turns into a code: Do we hand over our heritage to the machine and turn off the book?
Written by Dr. Muhammad Mamdouh Abdullah
Head of the NGO Development Sector at the Misr El Kheir Foundation
There is no doubt that digitization has brought about a tremendous revolution in the world of knowledge. It has become possible to store thousands of books in one device, search through millions of pages within seconds, and make old and rare books available to researchers anywhere. Artificial intelligence has also become capable of indexing, classifying, comparing, translating, and extracting information at an unprecedented speed. These are real advantages that make the trend towards digitizing heritage, historical, religious and cultural books important and necessary.
But the danger begins when digitization turns from a means of protecting the book and providing access to it into a complete alternative to it, and when we reach the conviction that the electronic copy replaces the printed copy, then we begin to destroy the original books or stop printing them, until the digital memory becomes the only source of knowledge.
The old book is not just words that can be converted into an electronic file. The original copy may carry historical and scientific value in terms of the type of paper, ink, binding, seals, dedication phrases, notes written by its owners in the margins, and differences between editions. In many cases, the original itself is evidence that the researcher can refer to to verify the authenticity of the text.
The risk increases when the only digital copy becomes the approved reference. A small error in scanning or in optical character recognition technology may change a word or sentence. This wrong copy is then transmitted to databases, search engines, and artificial intelligence systems, and the wrong information is repeated thousands of times until it appears to be true. In historical, religious and heritage books, changing a single word may have a significant impact on the meaning.
Let us imagine a rare historical book that was converted into a digital file and then its paper copies were destroyed. If the file is later modified, intentionally or as a result of a technical error, the researcher will not have an independent source with which he can compare. If printed copies remain in multiple libraries, any change can be detected by comparison.
Another danger is controlling the knowledge portal. A printed book in a library or home remains available without the need for a platform, subscription, operating system, or Internet connection. As for the digital book, it may become linked to a platform, database, or company, and if the access conditions change, the service stops, or the storage media becomes damaged, the content may become unavailable even though it was saved digitally.
The increasing reliance on artificial intelligence may lead to humans moving from reading the source to reading the machine’s interpretation of the source. If the researcher asks the artificial intelligence what a specific author said and merely answers without referring to the book, he may obtain a truncated summary, an inaccurate interpretation, or information taken from an incomplete copy. Here the machine becomes an intermediary between humans and the source, and the way it displays information may affect future generations’ understanding of heritage.
However, the solution should not be to reject technology. The problem is not in digitization, but in transforming it into an alternative to the original. What is required is to build a dual memory: a paper copy preserved according to scientific standards, and multiple digital copies preserved in more than one institution and site, while preserving images of the original pages along with searchable text.
Every digital file must carry clear data about its source, the copy that was photographed, the date of digitization, and the party that carried it out, with a system in place to verify that files have not been changed. Artificial intelligence techniques should not be used as a final source, but rather a tool that helps the researcher access the source, while making the page number and original page image available whenever possible.
The most important thing is for libraries, universities, and documentation centers to adopt a clear policy that prevents the destruction of heritage and historical books just because they have been digitized, and for important paper editions to continue to be produced, because multiple copies represent one of the most important elements of protecting memory from error, tampering, or loss.
The future is not in choosing between book and screen, but in combining them. We use technology to facilitate access to knowledge, and we use artificial intelligence for research and analysis, but we preserve the book as a physical witness to the knowledge that has reached us.
Memory that relies on a single digital copy can be changed or lost, but memory that preserves the original and benefits from technology at the same time gives future generations something more important than information: it gives them the ability to verify information.
Therefore, preserving the printed book is not a return to the past, but rather an insurance for the future. Because a nation that abandons the origins of its memory may find itself one day owning millions of files, but it does not have the means by which to know which ones were the originals.
