AI and Multilingualism Q5: The One Misconception This Book Most Urgently Corrects

SILENCE ON THE RANKS AND LABS! Despite all that has been said on the positive dimension of silence, meaning outside silence, silence around you, people still consider silence in history, silence in anthropology, silence- in archaeology, as meaning the absence of what we may or might be looking for. Let me and us be clear: Silence is the absence of any noise to our ears! But that does not mean there is nothing to be found if we should try to look for something. That reminds me of an episode in the Dentistry Ward of Hospital Val de Grâce in Paris, a military hospital. It was in 1968. All dentistry students who had finished their basic years of study opted to serve their military service in a military hospital, which was practice for their future professions. I was complaining about a toothache. The dentist graduate who was looking after me made a first X-ray image of the tooth and saw nothing. So, he did what he had been told to do. He adjusted the X-ray apparatus and made a second X-ray picture. On this one, there was a cyst at the tip of the root of the tooth. With these contradictory pictures, he asked his fellow dentist graduate next to him: "Which picture is the proper picture?" His fellow graduate told him: "A picture can always see nothing, but if a picture sees something, it does not hallucinate. So, you should proceed with the cyst." Today we know an AI machine can hallucinate. But in July or August 1968, there was no AI machine. X-ray machines did not hallucinate, though they could be blind. But in our field of cultural and spiritual archaeology, anthropology, and even theology and spirituality, when using AI machines, agentive or not, generative or not, we must keep in mind that "nothing" does not prove there is nothing, and it can hallucinate something where there is nothing at all, and it is better to double up the test, the procedure. AI applied to the 7,500 languages in the world is on the point of committing a mass murder of languages, a nearly complete genocide of languages. AI means LLM. LLM means 3 trillion data points. Only 6 to 12 languages can have such an LLM. All other 5,488 languages will have to accept being translated into one of the happy few LLMs by a machine trained for its own and only happy language selected by the masters of the world, hence translating the unhappy language into a happy few language is treacherous, distortive, unfair, and the original unhappy language <will end up dead, absent, except if 1- The world decides to compensate for this loss … 2- … Compensate by teaching the languages and cultures concerned to those who are their natural audience, the natives of these languages. 3- Develop libraries for these languages, working on all possible conservational media. 4- AND OF COURSE A BUDGET THAT HAS TO BE PAID BY THE TECHNOLOGY THAT CREATES THE HAZARD; Have a good read.