Voynich Manuscript Breakthrough

[p1t1o]

https://transacl.org/ojs/index.php/tacl/article/download/821/174

So apparently there has been some progress in decoding the Voynich Manscript - using "An AI" (which I think might just be 2018-speak for "we used computers" but it did involve some "machine learning") and combined with an analysis of a document called "The Universal Declaration of Human Rights" which was produced in 380 languages,  there are strong signs that the language used is a ciphered version of hebrew, with the letters of each word then put into an alphabetised anagram (ie: letters put into alphabetical order, where "saturn" becomes "anrstu").

Apparently the first sentence decodes to:

“She made recommendations to the priest, man of the house and me and people.”

The analysis is statistically strong, but it still takes a little word-wrangling to make it make sense, but if it is consistently spitting out recognisable hebrew words, they must be close, right?

 

[YNM]

 

It's a hebrew D&D ?

[p1t1o]

10 minutes ago, YNM said:

<snipt>

An XKCD always gets a like

My gut says its an illustrated novel written in a language made up for the purpose, Tolkien invented Elvish, its been done before.

[Scotius]

"Human nature doesn't change." I wholeheartedly agree. And what is in our nature? Laziness. Humans are lazy, and avoid any unnecessary activities they can. And what could be so super important and secret to be written by hand (and "manuscript" indicates a substantial amount of work) in such convoluted way? It had to take months if not years of work. Whoever created this book, must have been extremely devoted to this task. He also wasn't an unlearned nobody commoner. I wonder if it would be possible to use graphological analysis of the manuscript and other handwritten texts from the period and look for possible matches?

[p1t1o]

6 minutes ago, Scotius said:

"Human nature doesn't change." I wholeheartedly agree. And what is in our nature? Laziness. Humans are lazy, and avoid any unnecessary activities they can. And what could be so super important and secret to be written by hand (and "manuscript" indicates a substantial amount of work) in such convoluted way? It had to take months if not years of work. Whoever created this book, must have been extremely devoted to this task. He also wasn't an unlearned nobody commoner. I wonder if it would be possible to use graphological analysis of the manuscript and other handwritten texts from the period and look for possible matches?

I think its pretty safe to assume they have tried every technique known to man so far.

I'd go so far as to say, if your cryptographic technique hasnt been tried on the Voynich manuscript, its probably not a very good technique.

[magnemoe]

13 minutes ago, p1t1o said:

I think its pretty safe to assume they have tried every technique known to man so far.

I'd go so far as to say, if your cryptographic technique hasnt been tried on the Voynich manuscript, its probably not a very good technique.

They have tried all ways to decipher it. 
Its not just random characters, the word and character usage indicate language.

My guess its an sort of joke an aristocrat tried to pull on another. It might involve an invented language or just an scrambling of an ordinary text. 
Yes this would be expensive but they was rich so no huge issue to pay some for an year just to pull off an good joke. 

Second is that its some sort of scam, hype the book then sell it. 

The actual even is lost is history or might have flopped and the book was stored for a long time. 
 

[Mikki]

The letters are known to be used in northwest india/pakistan during medival times around 1000 years ago by a small population which did not survive as a whole, the language and designart (The manuskript) is mostlikely transferred to europe by gypsies, who commonly don`t use to write themselfes or deliver writings at all.
Its mostlikely a very unique piece of multicultural influences, some indian researchers have pointed years ago into this direction.

But who knows truely what it is.

Most plants have been identified to grow in this particular region until today and have their counterparts in many books used in the mediterranean area.

Edited January 30, 2018 by Mikki

[NSEP]

AI = 1

Humans = 0

[YNM]

9 minutes ago, NSEP said:

AI = 1

Humans = 0

In playing D&D ?

[NSEP]

2 minutes ago, YNM said:

In playing D&D ?

And chess, go, and uhhhh, decyphering a little bit of the Voynich Manuscript, maybe? If they aren't just using the computer and giving it fancy names?

[p1t1o]

@NSEP

It erks me that they said "an AI". Is this really AI? The software still had to be programmed to do what it did, so what if it absorbed the data from a bunch of languages first, the programmer is still the guy responsible for the result.

I know "AI" has a broad range of definitions, but I hate using it for everything that needs a computer, even if it did "learn".

When the general population reads "AI" (especially "an" AI, like its a standalone entity and not an executable on someones windows desktop) they automatically think "Artificial life!" and I hate propagating that kind of reaction.

 

 

[YNM]

4 hours ago, Scotius said:

Whoever created this book, must have been extremely devoted to this task. He also wasn't an unlearned nobody commoner. I wonder if it would be possible to use graphological analysis of the manuscript and other handwritten texts from the period and look for possible matches?

I think that the problem with Voynich is that it's shown by a man who owns it out of some... colection in what building ? But the (purported ?)history of the owner seems to be as weird as the manuscript itself.

500 years isn't too long to keep a book intact to be honest.

Edited January 30, 2018 by YNM

[sevenperforce]

56 minutes ago, p1t1o said:

@NSEP

It erks me that they said "an AI". Is this really AI? The software still had to be programmed to do what it did, so what if it absorbed the data from a bunch of languages first, the programmer is still the guy responsible for the result.

I know "AI" has a broad range of definitions, but I hate using it for everything that needs a computer, even if it did "learn".

When the general population reads "AI" (especially "an" AI, like its a standalone entity and not an executable on someones windows desktop) they automatically think "Artificial life!" and I hate propagating that kind of reaction.

Well, what's the difference between an advanced machine learning algorithm and "artificial life"? Is there some tipping point where machines suddenly gain a spark of consciousness? 

Is something "artificially intelligent" if it is able to learn in a way a human cannot, even if it never becomes self-aware?

I, for one, prefer the use of "AI" to refer to things like this, because it bring these questions to the forefront and inculcates the general population with the idea that perhaps "AI" is not what they think.

[YNM]

@sevenperforce "Smart programs" ? Siri, Google assistant, Cortana falls into the same category. True AI that has the same human intelligence and looks magical (and looks silly - what's a human) isn't here yet at all.

Though yes, I'd still tuck "intelligent" into it, probably not AI though, as this program just has "learned" something. That something is hard to tell in concept and hard to see, but it's producing results.

[sevenperforce]

3 minutes ago, YNM said:

@sevenperforce "Smart programs" ? Siri, Google assistant, Cortana falls into the same category. True AI that has the same human intelligence and looks magical (and looks silly - what's a human) isn't here yet at all.

Though yes, I'd still tuck "intelligent" into it, probably not AI though, as this program just has "learned" something. That something is hard to tell in concept and hard to see, but it's producing results.

Siri is more-or-less identical to what the general population expects "artificial intelligence" to be like, but is not qualitatively different than Clippy.

[p1t1o]

58 minutes ago, sevenperforce said:

Well, what's the difference between an advanced machine learning algorithm and "artificial life"? Is there some tipping point where machines suddenly gain a spark of consciousness?

Big-hitting questions for sure, questions which I am confident are not relevant to decryption software.

58 minutes ago, sevenperforce said:

Is something "artificially intelligent" if it is able to learn in a way a human cannot, even if it never becomes self-aware?

I  dont think my PC is artificially intelligent, no.

But yes, there are definitions of "AI" that might include things like my PC or your phone, or the decryption software used with the Voynich manuscript, but when a scientist, or anyone of subjective significance says "We used an AI." it automatically invokes one or more of the stronger definitions of AI because one is unlikely to assume that they would say "We used an AI" and be talking about something which you would only class as an AI after a conversation about definitions

 

58 minutes ago, sevenperforce said:

I, for one, prefer the use of "AI" to refer to things like this, because it bring these questions to the forefront and inculcates the general population with the idea that perhaps "AI" is not what they think.

A fair and valid point.

For me it is similar but converse. I dislike people assuming (or implying) AGI when someone means something far less paradigm-shifting, because they are getting an inaccurate idea of what is really going on.

Edited January 30, 2018 by p1t1o