Interesting article and though labelled as “artificial intelligence” this text reading software seems similar to that developed for printed and handwritten texts. Palaeographers will tell us that in manuscripts, and especially the ones by ecclesiastical copyists, a variety of fairly standard letterings were used over the centuries. So that it might actually easier to read the lettering in a 10th cent. manuscript than in a 20th century handwritten letter. The LANGUAGE is another matter: the Latin used by the Church, notaries and civil authorities in their records is not Cicero’s Latin. For example, court chancellors might hear pleas being given in a trial in the local vernacular in Sicily or Iceland and then write their notes translating directly into often shaky Latin prose. My suspicion is that a good deal of the material scanned and converted into ASCII code or something like that will be about as readable as many of those “txt” transcripts of old books that you can download on some major websites. Ah and then there are the GREEK manuscripts from Byzantine sources, where the same lettering and linguistic problems occur.
yes, that’s going to make it interesting, but I wonder why this needs to be done under the aegis of AI ? Surely not because it’s faster and more efficient. I suspect then plan is to search for something specific by the pretence of innovation.
Interesting article and though labelled as “artificial intelligence” this text reading software seems similar to that developed for printed and handwritten texts. Palaeographers will tell us that in manuscripts, and especially the ones by ecclesiastical copyists, a variety of fairly standard letterings were used over the centuries. So that it might actually easier to read the lettering in a 10th cent. manuscript than in a 20th century handwritten letter. The LANGUAGE is another matter: the Latin used by the Church, notaries and civil authorities in their records is not Cicero’s Latin. For example, court chancellors might hear pleas being given in a trial in the local vernacular in Sicily or Iceland and then write their notes translating directly into often shaky Latin prose. My suspicion is that a good deal of the material scanned and converted into ASCII code or something like that will be about as readable as many of those “txt” transcripts of old books that you can download on some major websites. Ah and then there are the GREEK manuscripts from Byzantine sources, where the same lettering and linguistic problems occur.
yes, that’s going to make it interesting, but I wonder why this needs to be done under the aegis of AI ? Surely not because it’s faster and more efficient. I suspect then plan is to search for something specific by the pretence of innovation.
This could well be.