This is actually really amazing. They scan charred, burnt scrolls that can’t be physically opened without crumbling, and the AI is able to use subtle deformations the ink on the scrolls caused to reconstruct letters.
The scrolls were discovered in the eighteenth century, when workmen came across the remains of a luxury villa that might have belonged to the family of Julius Caesar’s father-in-law. Deciphering the papyri, Sommerschield says, could “revolutionize our knowledge of ancient history and literature”. Most classical texts known today are the result of repeated copying by scribes over centuries. By contrast, the Herculaneum library contains works not known from any other sources, direct from the authors.
It’s definitely a process that will yield some false positives, if for no other reason than Romans were notorious users of abbreviations to save on resources which can make things ambiguous. But from an object cataloguing mountainous backlog perspective? fuck yeah. Scan all the tablets everywhere, and then pick out the super interesting ones for human review.
I’m a little bit surprised that all they got from that section of text was the word “purple”. I mean it’s a pretty amazing restoration, it looks like there’s a whole lot more than one decipherable word there!
I can definitely see the word porphyry here, which is super cool, but I forgot that the Greeks and Romans had another annoying resource-saving technique of not putting spaces between words or sometimes just using interpuncts•like•this. We didn’t start using word spaces in texts until centuries later.
I can get behind continuous writing when the word boundaries are otherwise obvious, but it’s super irritating for foreigners hundreds of years later trying to make out the details.
The researchers got fucking lucky this person had nice educated handwriting, at least. Sometimes you just get shit like this grocery list instead.
This is actually really amazing. They scan charred, burnt scrolls that can’t be physically opened without crumbling, and the AI is able to use subtle deformations the ink on the scrolls caused to reconstruct letters.
It’s definitely a process that will yield some false positives, if for no other reason than Romans were notorious users of abbreviations to save on resources which can make things ambiguous. But from an object cataloguing mountainous backlog perspective? fuck yeah. Scan all the tablets everywhere, and then pick out the super interesting ones for human review.
I’m a little bit surprised that all they got from that section of text was the word “purple”. I mean it’s a pretty amazing restoration, it looks like there’s a whole lot more than one decipherable word there!
Lol I missed this image entirely.
I can definitely see the word porphyry here, which is super cool, but I forgot that the Greeks and Romans had another annoying resource-saving technique of not putting spaces between words or sometimes just using interpuncts•like•this. We didn’t start using word spaces in texts until centuries later.
I can get behind continuous writing when the word boundaries are otherwise obvious, but it’s super irritating for foreigners hundreds of years later trying to make out the details.
The researchers got fucking lucky this person had nice educated handwriting, at least. Sometimes you just get shit like this grocery list instead.