I regret that I never found this post until now.It appears that the whole thread has quietly gone back under the rock from whence it emerged, but as a technologist who has a good level of experience with neural networks, I've still got to ask: "What the heck was that all about?!?"
I hope this post does not unfairly taint people's opinion of what role AI (and neural networks in particular) might play in Voynich Manuscript research. So let me add the following:
It is highly unlikely that AI will ever "crack the Voynich Manuscript" -- certainly not by any of the approaches that I've ever seen suggested in posts and publications. That being said, there are ways that AI and machine learning models can make helpful advances in several areas of Voynich research.


![[+]](https://www.voynich.ninja/images/collapse_collapsed.png)