(Yesterday, 01:12 PM)pjburkshire wrote: You are not allowed to view links. Register or Login to view.This July 2025 paper may have been commented on before but I didn't find it in a search. As I have stated before, I think the Voynich Manuscript is more medical and spiritual than alchemical but of course the Czech Shorthand part caught my eye.
Sounds good but it's copy-pasted AI slop: the author didn't even notice the made-up transliteration lines in the article. See MarcoP's posts about hallucinations: You are not allowed to view links. Register or Login to view.
Relying on AI so much doesn't usually lead to trustworthy conclusions. The AI was very probably just pretending to do the work of comparison and validation, because it's much faster and when the user is happy with the result, why not just skip the boring tasks?
Also, as MarcoP mentioned on reddit, all claims of "compressed" and "shorthand" are wrong:
Quote:In particular, you don't mention character entropy: that figure tells us that Voynichese may be an expanded natural language (eg a verbose cipher), it certainly isn't a compression. This was discussed by René Zandbergen: You are not allowed to view links. Register or Login to view..

