The Voynich Ninja

(28-11-2024, 04:04 AM)RadioFM Wrote: You are not allowed to view links. Register or Login to view.I think a good base test for future chatbots would be to give them a message coded using any of the advanced Steganographia technique or a pseudo-Voynichese cipher (I think we got a couple of those here in this forum). If an AI ever manages to extracts the plaintext from one of those without any hints at all, then I'd say it's worth looking into.

I'd love to give it a try. I vaguely remember seeing a few examples of Voynichese-looking ciphers on voynich.ninja, but cannot find any of them now. Could you suggest any specific keywords to search for?

(28-11-2024, 04:04 AM)RadioFM Wrote: You are not allowed to view links. Register or Login to view.I think it'd be best to give people a base test, something reproducible from the comfort of their PCs, so that they can check for themselves whether the latest chatbot can or cannot crack Voynichese.
I think a good base test for future chatbots would be to give them a message coded using any of the advanced Steganographia technique or a pseudo-Voynichese cipher (I think we got a couple of those here in this forum). If an AI ever manages to extracts the plaintext from one of those without any hints at all, then I'd say it's worth looking into.

That's an awesome idea. If someone can create such a test message, I will certainly use it like this in the video.

When the script is done I will check with everyone who contributed how they want to be credited in the video (forum alias or real name). This will be one with a lot of your input.

I will use my first post in this thread to start building a provisional structure.

I quickly wrote a rough structure in my first post: You are not allowed to view links. Register or Login to view.

Very much work in progress, all input welcome. There are very likely also essential parts from this thread that I missed.

• ChatGPT can help you write code to use in your research--> I have an example of this where I used code to help me analyze text files for a video, not the most exciting one; other stories are welcome.

Not sure if much more exciting, but I had it walk me through the free to use programs I needed and exactly what I had to do, step by step (explain like im 5 style..) to create something that would calculate first and second order entropy, then I had it alter the code to ignore line breaks. 

Now I have something set up in Visual Studio Code where I have a transcription of the VM, and below I can paste whatever text I like in and it will show me the VM readings and new text readings. 

I used this to compare different styles of texts in 1 language, rather than comparing the VM to various languages. Is it "song like" or "poem like" as people have said, for example.. do poems/song like texts even have different entropy readings than other texts? Then other tests, like do constructed languages better resemble voynich? 

None of it was terribly scientific, and I got distracted by other stuff but from looking at my notes I found the most promising avenue to explore/compare in future was constructed languages and also text that is in a normal language but contains elements of "constructed language"? I'm not sure the correct term for it, but text such as below, where it contains lots of names of things which are very predictable in nature, most ending "el" or "al" for example and also writing out numbers, for example if it carried on it would write "twenty−second" "twenty−third" "twenty−forth" and all the "twenty-" repetition would effect the entropy scores. 

"And behold the names of those angels: the first of them is Samjaza, the second Artaqifa, and the third Armen, the fourth Kokabel, the fifth Turael, the sixth Rumjal, the seventh Danjal, the eighth Neqael, the ninth Baraqel, the tenth Azazel, the eleventh Armaros, the twelfth Batarjal, the thirteenth Busasejal, the fourteenth Hananel, the fifteenth Turel, and the sixteenth Simapesiel, the seventeenth Jetrel, the eighteenth Tumael, the nineteenth Turel, the twentieth Rumael, the twenty−first Azazel."

Is the VM just listing stuff... ¯\_(ツ)_/¯, but it's nice to have this code on hand to explore ideas

I am in favor of using AI to find matches to the astrological charts with those in other works. Specifically, if errors or idiosyncrasies in the VM's illustrations are found in other documents, that may start a trail of crumbs.From there we may know if the manuscript was source or derivative material. Either way, encrypting it could make sense, sigh!