PulseAugur / Brief
EN
LIVE 15:08:29

Brief

last 24h
[3/3] 224 sources

Multi-source AI news clustered, deduplicated, and scored 0–100 across authority, cluster strength, headline signal, and time decay.

  1. MEME · Mastodon — sigmoid.social Español(ES) ·

    I asked AI to try to translate or decipher a page from the controversial Voynich Manuscript, although it said it was not possible, it offered to generate a page

    An AI model was prompted to translate or decipher a page from the Voynich Manuscript. While the AI stated it could not perform the translation, it offered to generate a page that closely resembled the provided illustration. The AI cautioned that any generated content would be entirely fictional. AI

    I asked AI to try to translate or decipher a page from the controversial Voynich Manuscript, although it said it was not possible, it offered to generate a page

    IMPACT AI used to generate fictional content based on historical artifacts.

  2. RESEARCH · arXiv cs.CL English(EN) ·

    A Quantitative Confirmation of the Currier Language Distinction

    Researchers have quantitatively confirmed a linguistic distinction within the Voynich manuscript, previously proposed by Currier. Using a Beta-Binomial mixture model on character-pair substitution ratios, they were able to recover Currier's A/B language split with high accuracy. This analysis suggests a genuine structural property of the text and identifies three functional regimes for character pairs, which could inform future theories about the manuscript's writing system. AI

    A Quantitative Confirmation of the Currier Language Distinction

    IMPACT Provides a novel analytical framework for linguistic pattern detection that could be adapted for other historical or encoded texts.

  3. RESEARCH · Hugging Face Daily Papers English(EN) ·

    A Quantitative Confirmation of the Currier Language Distinction

    Researchers have quantitatively confirmed a proposed language distinction within the Voynich manuscript, known as Currier's A/B split. Using a Beta-Binomial mixture model on character counts, they were able to recover this distinction without prior labeling. A supervised classifier trained on this data achieved 89.2% accuracy in predicting folio identities, suggesting the character pairs exhibit distinct functional patterns. AI

    A Quantitative Confirmation of the Currier Language Distinction

    IMPACT Provides a novel quantitative method for analyzing historical texts, potentially applicable to other linguistic or cryptographic challenges.