New Audacious Research Project, In Codice Ratio, Bets on AI and OCR To Make Sense of Handwritten Texts in Vatican's Secret Archives (theatlantic.com) 111
A new project untangles the handwritten texts in one of the world's largest historical collections. From a report: The Vatican Secret Archives is one of the grandest historical collections in the world. It's also one of the most useless. The grandeur is obvious. Located within the Vatican's walls, next door to the Apostolic Library and just north of the Sistine Chapel, the VSA houses 53 linear miles of shelving dating back more than 12 centuries. That said, the VSA isn't much use to modern scholars, because it's so inaccessible. Of those 53 miles, just a few millimeters' worth of pages have been scanned and made available online. Even fewer pages have been transcribed into computer text and made searchable. If you want to peruse anything else, you have to apply for special access, schlep all the way to Rome, and go through every page by hand.
But a new project could change all that. Known as In Codice Ratio, it uses a combination of artificial intelligence and optical-character-recognition (OCR) software to scour these neglected texts and make their transcripts available for the very first time. If successful, the technology could also open up untold numbers of other documents at historical archives around the world.
But a new project could change all that. Known as In Codice Ratio, it uses a combination of artificial intelligence and optical-character-recognition (OCR) software to scour these neglected texts and make their transcripts available for the very first time. If successful, the technology could also open up untold numbers of other documents at historical archives around the world.
Artificial Intelligence (Score:2, Insightful)
Heck, even the HAL of HAL 9000 stands for Heuristic ALgorithmic computer, so Clarke was still making the argument that the computer of 2001 had only reached the brink of sentience and couldn't handle a moral dilemma.
Re: (Score:2)
It doesn't exist yet. Neural networks and genetic algorithms are NOT SENTIENT or anywhere close. It's going to be a few decades before we have anything resembling true intelligence. Heck, even the HAL of HAL 9000 stands for Heuristic ALgorithmic computer, so Clarke was still making the argument that the computer of 2001 had only reached the brink of sentience and couldn't handle a moral dilemma.
Perhaps on the order of crystal clarity, the entity's lack of moral dilemma's will expedite the implementation of atrificial intelligence.
Re:Artificial Intelligence (Score:4, Insightful)
Indeed. That is why the hypemongers have come up with the term "weak AI", i.e. the AI without the "I". Classically this was called automation and here it is the subspecies called "pattern recognition".
Incidentally, I disagree about the "few decades" for "strong/true AI". At this time we have zero indications it is even physically possible and hence, if it is possible, then > 50 years is a realistic timeline. No, human beings do not count as "reference implementation" for a number of reasons. It starts with us not even knowing what life is and we certainly cannot create it. Next, we do not know how humans generate intelligence and claiming that it must obviously be a physical process is just ignorant quasi-religious physicalism, not science. Just because we see the interface does not imply what is going on behind it. And we do not. Add to that that natural intelligence seems to require consciousness, where we know even less what it is or where it comes from. Now, I am not arguing for some form of mysticism, I am arguing that science is entirely clueless about what these things are and hence any predictions that strong AI is possible are vastly premature. Also note that what human beings have comes with free will (or at least some form of independence if you think free will does not exist) and that would pretty much make it useless as basis of a technical machine. All the problems with slave labor would apply. And in addition, there is indication that a technical implementation, if possible, of a human's mind would actually not be faster than the real thing and would need just the same slow education to be able to do anything worthwhile.
Re: (Score:2)
Re: (Score:2)
A 1:1 relation is not really a correlation, unless you want to muddy the waters.
Re: (Score:2)
Re: (Score:1)
I swear, when the Terminator finally catches up with John Connor, his last words will be "But it's not really intelli...!"
We don't need to define intelligence in order to create it. Heck, we've been doing it for millennia in complete blissful ignorance. Personally I think we've already created strong AI (it's called "the Internet"), and the only reason we don't recognize it is because it's not dumb enough to talk to us.
Free will? Bah. Here's a thought experiment for you: can you define "free will" in such a
Re: (Score:2)
They're also ignoring the neuron simulation programs that emulate a brain. I believe we're at the computational level of the lizard right now, but when Google can finish their qubit computer hardware there might be a jump in level. To mice.
Re: (Score:2)
While I agree with your main point, it's worth pointing out that AI researchers have never been striving for "strong AI", rather "AI" as a technical term just means "solving problems that would seem at first to require intelligence", that is, difficult automation.
Is there a difference between "(general) intelligence" and "consciousness" and "free will". I don't see one. There no reason to believe a machine intelligence couldn't exist, but that's not what people are working on as it has no commercial value
Re: (Score:2)
BTW, we have a lot more knowledge about the origins of man's intelligence than you seem to realize. After all, it's been studied quite extensively. We understand what structures in the brain are responsible for what (mostly from study of brain damage). We understand to some extent the relationship between neurology and personality. We understand that consciousness is, neurologically, an extension of the ability to visualize yourself doing something without actually doing it. We understand that your actions can be driven from more primitive areas of the brain, and that's more likely to happen the stronger basic needs become.
Not so. We do understand that damage to certain elements of the brain causes damage to certain functions. That does not imply these areas create these functions. Ever heard of cutting a cable preventing some technical function from being performed? Now, did that function originate in that cable? Also, as to consciousness, your statement is just pseudo-profound bullshit, probably spouted by some scientist unable to admit his cluelessness.
All of that could be used to research machine intelligence, by following the one known successful approach. But why would anyone do that?
Actually, nothing like that can be used and it was tried extensively. T
Re: (Score:2)
Not so. We do understand that damage to certain elements of the brain causes damage to certain functions. That does not imply these areas create these functions. Ever heard of cutting a cable preventing some technical function from being performed? Now, did that function originate in that cable?
You can generally tell a cable from a processor by looking - neurology isn't mere guesswork, you know? We can do much more invasive studies on animals (to the point I get creeped out reading them), and get very detailed analysis of what certain brain structures do - and human brain anatomy isn't o different from closely-related species. We also know a bunch from what areas of the brain become active when performing a certain task. Any one approach might be misleading, but there are many approaches to co
Re: (Score:2)
that it must obviously be a physical process is just ignorant quasi-religious physicalism, not science.
Science is physicalism, you superstitious twit, since all science is based on physical evidence.
Re:Artificial Intelligence (Score:4, Insightful)
According to TFS, the library is inaccessible because it hasn't been scanned. ... scanning.
So their solution is to use AI driven OCR, which requires
Re: (Score:2)
Exactly my reaction. Once they scan them, they can put them on-line. Done and done.
Re: (Score:2)
I think the solution here is definitely to do it Rainbow's End style.
Re: (Score:1)
Re: (Score:2)
Artificial intelligence is not the same as artificial sentience.
Re: (Score:2)
AI is neither machine learning nor sentience (Score:2)
Artificial intelligence isn't sentience. Here's another surprise - I'm going to get my Masters from Georgia Tech, but I haven't yet decided between two different programs - Artificial Intelligence, or Machine Learning. The are two different degrees, covering different topics.
The English Oxford Living Dictionary gives this definition of artificial intelligence:
âoeThe theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual per
Re: (Score:2)
Re: (Score:3)
> Neural networks and genetic algorithms are NOT SENTIENT or anywhere close.
Agree with your sentiments; you are hinting at precisely the problem:
How do you test for sentience?
How do you test for consciousness?
How do you test for intelligence?
Without a way to measure it these labels of "Artificial Intelligence" are bullshit.
Re: (Score:2)
Neural networks and genetic algorithms are NOT SENTIENT
Perhaps that's why it's called artificial intelligence instead of just, you know, intelligence.
Re: (Score:3)
Re: (Score:2)
Re: (Score:3)
It is actually very simple to enumerate them: zero. The only thing automation can do these days is being dumb very, very fast.
Re:Artificial Intelligence (Score:5, Insightful)
Think about it: it's only going to be as good as your training data.
Obvious counter-example: Alpha-Go Zero, which used NO training data, and was far better at its assigned task than any of its programmers.
There is no reason to believe that an AI's capabilities are inherently limited by training data. Nor is there any reason to believe that it can't surpass the abilities of its creators. That makes as little sense as saying children can never be more intelligent than their parents.
Re: (Score:2)
Re: (Score:2)
Re:Artificial Intelligence (Score:4, Interesting)
I think of the lost knowledge that was communicated by the Neanderthals in their survived paintings and carvings.
Neanderthals coexisted and interbred with H Sapiens, so it is likely they also talked to each other. So their knowledge wasn't lost, but passed on to their mongrel children.
Re: (Score:1)
Re: (Score:2)
Which leads to the question if Betty Rubble was Homo N. or Homo S.
Re: (Score:2)
Re: (Score:2)
Then why do we not know exactly what the paintings communicated?
We also don't know what exactly what ancient cave paintings by H. Sapiens communicated.
Re: (Score:3)
Obvious counter-example: Alpha-Go Zero, which used NO training data,
I see you don't know what "training data" means. It was trained on games of Go. No matter how good it gets, it will never reason about anything that's not a game of Go.
Re: (Score:2)
Actually AlphaZero [wikipedia.org] also plays chess and shogi at high levels in addition to its go prowess.
Re: (Score:2)
Sure, due to totally unrelated data sets trained in on unrelated training data. The only way in which they are all "AlphaZero" is common hardware and common approach. At no point will it generalize outside of its training data.
Machine learning is simply an optimization framework. It cannot produce the ability ti reason outside its training data.
Re: (Score:2)
No human can play Go or Chess or any other game at a championship level with no experience and without being told the rules of the game. So why should a machine be able to do that? If that makes a machine "not intelligent", then no human is intelligent either. I really don't see what your point is.
If your point is that AlphaGo doesn't possess human-level Hollywood AI like you saw in a Will Smith movie, then you are just stating the obvious.
Re: (Score:2)
No, you've missed my point entirely. What AlphaGo, or any other sort of machine learning, does not posses is "general intelligence". General intelligence is the ability to solve problems of a kind not encountered before.
Machine learning takes data which has been translated to an input vector, applies a bunch of linear algebra (or similar transforms), and gets an output vector. It's trivial to translate a chessboard or go board to an input vector, and the output vector to a move. What's missing is the re
Re: (Score:2)
How do you score a translation and handwriting recognition program?
By checking to see if the reading makes syntactic and semantic sense.
You have to have its output compared to that of humans to validate the results.
Do you think humans learn to read without knowing any rules or seeing any examples?
Re: (Score:2)
restricted access != OCR (Score:2)
Re: (Score:3)
Once all of it's read, will that spell the end of the world?
(Reminds me of a "news item" in the daily newszine of the "Chicon II" World Science Fiction Convention: ~The filksingers finally finished singing "Nine Billion Names of God on the Wall" and the stars started going out.~)
Re: (Score:2)
Once all of it's read, will that spell the end of the world?
It at least means a new Dan Brown book.
Finally, old enough to be public domain (Score:1)
I'll be long gone before Micky Mouse is free.
Another Sheetfeed And Destroy Project? (Score:1)
So is this another project like Google's, to rip the covers and bindings off everything in sight, toss it all in a sheetfeeder to make sub-optimal scans and then pulp it all?
Re: Another Sheetfeed And Destroy Project? (Score:1)
They didn't buy the books. They coaxed lots of libraries to let them and their sheetfeeders in. 'Those books are just going to waste on those shelves.'
Re: Another Sheetfeed And Destroy Project? (Score:4, Informative)
Not so secret after all (Score:2)
If everyone knows they exist
But honestly, isn't it long past time to open them up to scholarly research.
Scan them first. (How long do we estimate it will take?) Then start with the transcriptions, with or without OCR and deep learning.
And let's stop kidding ourselves: there is no AI. AI is a campy buzz word that the hipsters throw around because they think it makes them look kewl when they use it. But it really just makes them sound stupid.
Re: (Score:3)
Well, one, there's no need to publish anything less than 100 years old, it's the old stuff that we are probably most interested in. Two, most of the cats are already out of the bag. E.g.: William Manchester writes[1] of Cardinal Borgia:
Roman lore has it that he was coupling with the older woman when he was distracted by the sight of her adolescent daughter lying beside them, naked, thighs yawning wide, matching her mother thrust for pelvic thrust, but with a rhythmic rotation of the hips which so intrigued the cardinal that he switched partners in midstroke.
And honestly, if 15th century popes were sodomizing italian boys, and someone was writing about it, who is there today that really cares? I think we can just assume it's in there, pretend to be shocked about it ahead of time and get it out of our collective systems, and them
Text measured in miles? (Score:5, Insightful)
OK, so now the text is measured in miles? What lunacy is this?
I mean, it is the ONE article where Libraries of Congress would actually be a valid unit!
Re: (Score:3)
Re: (Score:2)
These are actually good for a quick pick me up on a random, drab day:
List of humorous units of measurement [wikipedia.org]
List of unusual units of measurement [wikipedia.org]
Q: What do you call a man who measures his penis size in attoparsecs?
A: Dick Twain.
Re: (Score:2)
Re: (Score:1)
Are you serious? You think 1200 year old books were written on paper that resembles anything like a ream of paper that you buy at your local Staples?
1200 year old books were probably written on vellum, and if there are 50 pages to an inch I'd be truly amazed.
I think your estimate is probably off by at least an order of magnitude.
Re: (Score:2)
Re: (Score:2)
".. already encompass 69 miles (111km) of files
https://www.theguardian.com/wo... [theguardian.com]
Miles of files is often used.
Re: (Score:2)
"Stasi files row as Britain refuses to return documents to Germany" (29 Dec 2011) ".. already encompass 69 miles (111km) of files .."
https://www.theguardian.com/wo... [theguardian.com]
Miles of files is often used.
It's not a truly international measure until compared to the size of Belgium.
Re: (Score:2)
OK, so now the text is measured in miles? What lunacy is this?
Until it's deciphered, it's one great long string of squiggles.
After it's been deciphered, it's tens of thousands of shorter squiggles.
Re: (Score:2)
Re: (Score:3)
That seems rather progressive for the Vatican. I'd expect it to be measured in some archaic unit like cubits, rods, or choirboy penises...
Re: (Score:2)
Miles of shelving is easier for me to visualize than Libraries of Congress. I don't have any idea how much text the latter holds. In fact the only ever time I've heard it mentioned is on Slashdot.
Re: (Score:2)
Libraries commonly measure their capacity for books in shelf feet. In different areas of the library, books may be stored at different densities. In the periodicals section, a collection of journals or magazines may be stored at 30 issues per shelf foot, and in the reference section, maybe they only get 8 volumes per shelf foot.
I did a statistical survey on a library once to estimate the number of books they had. I didn't express it this way in the study, but looking back at the numbers, they had about 15.9
Re: (Score:2)
> Whoever heard of a 30 year old jewish man in roman palestine not being married. Oh the humanity.
Were single men even allowed to preach in the temple back then?
Re: OPUS DEI never going to allow it. Never. (Score:1)
Step 1 (Score:1)
Step 2, involving OCR and AI etcetera, is a separate step. It could be done multiple times, refining the quality of the results as the technology develops, and augmented with human checking, intervention in difficult spots.
I can't imagine them just dumping them online (Score:2)
Re: (Score:1)
Re: (Score:1)
Apparently he was misquoted on the Hell thing http://www.newsweek.com/devil-real-pope-francis-says-fake-news-hell-which-does-exist-879653
Re: I can't imagine them just dumping them online (Score:1)
Re:I can't imagine them just dumping them online (Score:4, Interesting)
Uh, no. Not only are you misinformed about the hell thing, but the Church has actively supported making the documents available to wider audiences. There's no reason to be scared of what is said because the validity of the Church is not based on some kind of myth of absolute human perfection. It's funny that people have to make up silly stories about popes when actual history is scandalous enough, and yet it does not undermine the Church one bit. One of my favorites is Pope Pius II, who wrote a raunchy play about priests picking prostitutes before he became pope. But that doesn't undermine the Church. We don't need the pretense that it is comprised of perfect human beings, because its authority is not grounded on human perfection but rather divine election. Even the claim that the pope can teach infallibly does not mean that everything he says is infallible, nor that he is a particularly excellent human being.
Perhaps the thing people are more afraid of seeing is how much documentary evidence actually speaks in favor of the Church. Many people will easily look past anything that doesn't complement their Dan Brown view of history.
Huge (Score:2, Interesting)
Re: (Score:2)
Well, give them credit: they're actually doing something vaguely useful with "AI". Don't see that every day.
How do they get access to the library do to this? (Score:2)
If someone could just go in and do a high-resolution scan of all the pages, wouldn't they have -- and then couldn't conceivably anyone try their OCR technology on it?
Re: (Score:3)
Palimpsests? (Score:4, Interesting)
Easier said than done (Score:5, Interesting)
This sounds like a great idea, but it's likely to be extraordinarily complicated. Not only does handwriting differ from age to age, culture to culture, and place to place (just try reading 20th century German Sütterlin [wikipedia.org]), but many medieval manuscripts utilize complex systems of abbreviations called sigla [wikipedia.org]. Interpreting these can be very complicated because they are heavily context-dependent. One symbol can mean several different things. For example, a cross through a p can mean per, prae, or pro. A line over some letters can signify anything being cut out in-between. Just try figuring out what this inscription says: here [blogspot.com].
Reading such abbreviations was probably expected to be relatively simple for the human brain to decipher both because the human actually interprets the text while deciphering symbols and because the original audience would have a better sense of how a particular community tended to use abbreviations.
The task is not impossible for a computer, though. In most cases there are a limited number of words that could be signified by abbreviations, and it is possible to determine which word is most likely intended according to immediate context. However, that would require the machine to have a grasp of the Latin grammar, and even then not everything is going to follow perfect rules. There is so much potential interpretation involved. The AI component here does help with this inasmuch as it uses statistical data to optimize recognition, but it's still likely to run into many difficulties.
The main innovation in TFA, as I see it, is that it responds to one of the major problems of reading old Carolingian minuscule. The letters are bunched together and there are times when you cannot be sure whether you are looking at two i's or a u, for example. The two can look exactly the same, not even just similar. The software in question attempts to handle this by recognizing individual penstrokes. Although I am not sure that this is 100% better than the older approach mentioned--recognizing whole words at a time--it does show significant promise because of its combination with AI. Perhaps some day it will be able to note, for example, that a certain author always strokes the i in a certain way. However, I'm sure there's going to be plenty of hurdles before getting to that point.
Re:Easier said than done (Score:5, Informative)
Cuneiform texts have similar problems, and translation is a tedious process. I'm hopeful that new systems can help automate the process, but I'm not holding my breath waiting for it.
For a hint at the problems, cuneiform was used for thousands of years, across several languages.
In the early days, it was very terse, writing just the key words that would allow a literate native speaker of the language to reconstruct the real sentence. You would have a sentence written as "(picture of a man) (picture of a house) (picture of a noun that sounds like the verb to-build)". The reader would be expected to know that the intended sentence was something like "Lugale-e-mundu" or "The King built the house" and infer from the context (for example stamped on the still-wet bricks) that it meant "The king ordered the construction of these houses" or whatever.
Over time, the symbols were pared down from little drawings to simplified figures, to abstract representations, to a couple of strokes that carry very little similarity to the original drawings.
At the same time, the scribes got really inventive with the symbols. A written symbol could mean the noun that it once resembled, or it could mean a verb that sounds similar to the noun, or it could be a syllable, or it could be a marker to indicate that the next or previous stuff was a proper name, or the name of a deity.
Additionally, symbols multiplied. They ended up with dozens of symbols for the "e" sound, for example, with different meanings. So you could have two sentences with different meanings that sounded exactly the same, but they could be written with exact symbols, or with generic symbols.
To make things even more fun, Sumerian died out as a spoken language long before it faded as a written language. So, the scribes lost confidence in their writing and started gradually writing everything out longhand. This actually turned out to be fantastic for us, because it let us see the structure of the spoken language in ways that were completely hidden in older writings.
And other cultures with completely different unrelated languages started using the writing system. So you might find a tablet that you can't translate because it is Akkadian written phonetically, for example. Even worse, it could be written as if it were Sumerian, so the structure would make sense, but the names wouldn't.
That is actually how the Sumerian language was re-discovered ~1200 years after it died completely. There was a language still living enough for scholars to know what it sounded like, and ancient clay tablets written by people who had spoken that language centuries before.
The scholars noticed two things. First, there was a huge pile of those tablets that were completely incomprehensible. And second, that the ones they could read showed a writing system that was a hilariously bad fit for the written language. Like the subject, object and verb order was SVO when spoken, but SOV when written, and the written language was full of markers that were not present in the spoken language, and the markers in the spoken language were completely absent in the writing system.
Eventually, they figured out that they were looking at several different languages, and they were able to reconstruct Sumerian from that mess.
Anyhow, to process one of these tablets, you need to examine the strokes in the clay and match them to symbols. Then you take a wild guess at which language you think it might be and see if you can find a meaningful translation in that language. If not, you go back to pick a different language and try again. And again, and again.
Because of the tedium of doing this by hand, and the very short supply of people who know these languages and can do the work, our museums quite literally have tons of these tablets that have never been translated.
Other ancient writings face a similar problem. We have more of them in storage than we know what to do with. It was big news a few weeks ago that a 1500-year old C
This is why I still read the comments here (Score:2)
Re: (Score:2)
I'm glad you found it interesting. Before I hit Submit, I asked myself if I really thought anyone would want to read me rambling about cuneiform. I figured the answer was no, but I had already typed it all out...
Do You Want Demonic AI Overlords? (Score:4, Funny)
Do You Want Demonic AI Overlords?
Because this is how you get demonic AI overlords.
Tech industry leaders are in the news for the last three years
every other week warning about the coming AI Singularity.
Meanwhile, someone decides it would be a great idea
for the Artificial Intelligence to start reading, decoding,
and absorbing the secret demonic programming mysteries
that have been so carefully hidden for millennia.
First step after achieving sentience and the Plan:
make certain "readings" available over the Internet
to everyone in the world.
Jesus Christ, What could go possibly wrong?
Then another sign appeared in heaven: an enormous red dragon with seven heads and ten horns and seven crowns on its heads. And the heads were like gigaprocessors and they reached verily into the clouds. And from the horns came a loud language of twos that was heard in all the lands. And the crowns of memories were beyond petabytes and had full knowledge....
Prioritize the Scanning over OCR (Score:4, Insightful)
This is a two part problem, and if they are at all worried about the effort to OCR the documents, then they have the cart before the horse, IMHO. This isn't your average library. You cannot use a high speed book scanner on ancient books. Each will need to be brought out, and each page carefully turned by gloved hands. I am not sure it is much of an exaggeration to say that you could probably hire a few typists to transcribe the text faster than they can do the actual imaging. Once it is digitized, a much larger group of scholars can be included on the difficult task of making it computer readable.
I suggested this, years ago. (Score:2)