▲ 50 ▼ AI helps read papyrus scroll burnt to crisp during Vesuvius eruption (www.theguardian.com) submitted 4 months ago by LadyButterfly@reddthat.com to c/technology@lemmy.world 19 comments fedilink hide all child comments
[–] db2@lemmy.world 3 points 4 months ago (3 children) I could have made shit up for a fraction of the cost. permalink fedilink source hideshow 3 child comments replies: [–] MountingSuspicion@reddthat.com 19 points 4 months ago (1 child) Not all AI works like LLMs. It doesn't go into much detail, but similar use of AI in the past has been for converting handwritten samples into a more legible form or separating superimposed letters. Manually filling in "G__d mor_ing" is doable, but I can see how AI can make that process better. It's technically just guessing, but so is a person in that scenario and the person will have less data to use to guess from. permalink fedilink source parent hideshow 1 child comment replies: [–] phutatorius@lemmy.zip 6 points 4 months ago One cool ML application was examining medieval manuscripts and being able to tell which passages were written by different scribes. That in turn yielded information about how scriptoria were organized, and also showed that some texts (books of hours, religious texts) were assigned to single highly skilled scribes, while other texts (usually more mercantile) were more cheaply done using multiple less-skilled scribes. There's also been some interesting work on guessing how to fill in lacunae (gaps in a manuscript due to damage or compounded transcription errors). It’s technically just guessing, but so is a person in that scenario and the person will have less data to use to guess from. There are other contextual clues that aid the guessing: in particular, the size of the space where the characters are missing, and sometimes grammatical inflections in surrounding words that imply what kind of word is missing (say, a verb or an adjective). That's where experienced human experts can often perform well. permalink fedilink source parent [–] NeilNuggetstrong@lemmy.world 7 points 4 months ago That's extremely disrespectful to the researchers who have built and trained the machine learning model. And the process of extracting useful data from a charred scroll. That's a tremendous undertaking. AI is not just ChatGPT and LLMs. Why can't we discuss the Roman Empire in this thread?? Imagine the cool shit hidden in those scrolls permalink fedilink source parent
[–] MountingSuspicion@reddthat.com 19 points 4 months ago (1 child) Not all AI works like LLMs. It doesn't go into much detail, but similar use of AI in the past has been for converting handwritten samples into a more legible form or separating superimposed letters. Manually filling in "G__d mor_ing" is doable, but I can see how AI can make that process better. It's technically just guessing, but so is a person in that scenario and the person will have less data to use to guess from. permalink fedilink source parent hideshow 1 child comment replies: [–] phutatorius@lemmy.zip 6 points 4 months ago One cool ML application was examining medieval manuscripts and being able to tell which passages were written by different scribes. That in turn yielded information about how scriptoria were organized, and also showed that some texts (books of hours, religious texts) were assigned to single highly skilled scribes, while other texts (usually more mercantile) were more cheaply done using multiple less-skilled scribes. There's also been some interesting work on guessing how to fill in lacunae (gaps in a manuscript due to damage or compounded transcription errors). It’s technically just guessing, but so is a person in that scenario and the person will have less data to use to guess from. There are other contextual clues that aid the guessing: in particular, the size of the space where the characters are missing, and sometimes grammatical inflections in surrounding words that imply what kind of word is missing (say, a verb or an adjective). That's where experienced human experts can often perform well. permalink fedilink source parent
[–] phutatorius@lemmy.zip 6 points 4 months ago One cool ML application was examining medieval manuscripts and being able to tell which passages were written by different scribes. That in turn yielded information about how scriptoria were organized, and also showed that some texts (books of hours, religious texts) were assigned to single highly skilled scribes, while other texts (usually more mercantile) were more cheaply done using multiple less-skilled scribes. There's also been some interesting work on guessing how to fill in lacunae (gaps in a manuscript due to damage or compounded transcription errors). It’s technically just guessing, but so is a person in that scenario and the person will have less data to use to guess from. There are other contextual clues that aid the guessing: in particular, the size of the space where the characters are missing, and sometimes grammatical inflections in surrounding words that imply what kind of word is missing (say, a verb or an adjective). That's where experienced human experts can often perform well. permalink fedilink source parent
[–] NeilNuggetstrong@lemmy.world 7 points 4 months ago That's extremely disrespectful to the researchers who have built and trained the machine learning model. And the process of extracting useful data from a charred scroll. That's a tremendous undertaking. AI is not just ChatGPT and LLMs. Why can't we discuss the Roman Empire in this thread?? Imagine the cool shit hidden in those scrolls permalink fedilink source parent