Writing a fake Voynich by rules: copies, habits and the link between words

[quimqu]

Jesus, now that I've been playing around with Claude a bit, I suddenly see its writing style everywhere - this is awful.

Anyway, what I want to add is that most seriation research misrepresents the status quastionem. The current binding is very easy to beat because everything is messed up, which we have known for decades. The question to ask is whether a singulion structure beats an alternate "correct" nesting.

Hello Koen,

yes, Claude style (and previously ChatGPT style) are common in the forum. My writting in presenting the results is Claude aided, as for me it is very difficult to write, make myself understandable, and make the results clear in a mix of technical and non technical ways. I think you can see in this thread: replies are my onw directly and posts about results are Claude aided.

As you might know, I am a data scientist and I am the manager of a R&D department. As a technician, at the beginning, I was very sceptic about IA coding. My history with IA is summarized here: wow (at the start of everything), uuuhhh (scared, after some months), let's see (after some months), this isn't working (a few months ago), wow wow wow (from August 2026). Now I really think the AI will help (is helping) us pushing forward with the VM, at least with the code side. 

I sarted coding line by line, word by word, passing the transliterations to my python code. Then I started to use the help of ChatGPT and simmilar IA's for coding. It was a push forward, but it was very tricky and slow, as the code made mistakes, the IA did things I didn't want it to make.. But it was promising, tyring but promising. I stopped for some months but then, this August, everything changed. Not only with the Voynich but also at work: we discovered Claude Code. And it is amazing... I talked with other people in other companies and they think the same: coding is no longer a barrier, it's a commodity, the code that our IA scriba (Claude code) generates now is fully reliable. The thing is giving the instructions right. And what is great: you have your full mind focused on analysisng results, proposing test, cryticism, everything that enriches the investigation.

One note: there is a huge difference between making the IA guess what the Voynich is about and making absurde theories about it, and helping to analyze the transliterations.