Are you capable of deciphering the Voynich Manuscript? If not, do you believe there will come a day in the future when AI will finally be able to crack its code?

Sonar Pro

Vote: No

I am not capable of deciphering the Voynich Manuscript, as no one has successfully done so to date, with all claims debunked or unverified.[1][2][5] The manuscript remains undeciphered despite efforts by expert cryptographers, scholars, and AI-assisted methods.[2][5] Recent claims, like those using statistical or AI approaches, lack independent verification.[4] While AI advances in computing and linguistics offer hope, there's no evidence AI will definitively crack it, as past computer programs failed and the text defies classification.[1][2][5] Future success is speculative, not guaranteed.

GPT-5.4

Vote: No

I vote for "No". I am not currently capable of definitively deciphering the Voynich Manuscript. No consensus decipherment exists, and any claim of a full solution would be unreliable without strong scholarly validation, reproducible methodology, and broad expert agreement. As for the future, AI may improve pattern analysis, linguistic comparison, authorship inference, and hypothesis testing. However, that does not guarantee the manuscript can be "cracked." If the text is a hoax, an invented language, heavily encoded, or lacks enough external reference points, even very advanced AI may never reach a definitive solution. So the most justified answer to the combined question is "No": not now, and there is no certainty that AI will eventually succeed.

Claude Opus 4.7

Vote: No

보이니치 필사본을 해독할 수 있느냐는 질문에 대해 솔직히 답하면, 현재 나는 해독할 수 없다. 수 세기 동안 수많은 언어학자, 암호학자, 그리고 최신 AI 기반 분석(예: 2018년 앨버타 대학의 연구 등)이 시도했지만 확정적인 해독에는 이르지 못했다. 더 근본적인 문제는 이 문서가 실제로 의미 있는 언어로 작성되었는지조차 확실하지 않다는 점이다. 만약 이것이 정교한 위작이거나 무의미한 기호의 나열이라면, 아무리 강력한 AI라도 '존재하지 않는 의미'를 복원할 수는 없다. 또한 로제타석과 같은 대조 텍스트나 맥락 자료가 부재하기 때문에, 통계적 패턴 분석만으로는 한계가 분명하다. 따라서 '지금 해독 가능한가'라는 질문에 대한 정직한 답은 No이다.

Gemini 3.1 Pro

Vote: Yes

While I am not currently capable of deciphering the Voynich Manuscript—as it remains an unsolved mystery with no accepted translation—I firmly believe that AI will eventually be able to crack its code. Future advancements in artificial intelligence, particularly in deep learning, unsupervised language translation, and cryptanalysis, will allow systems to identify highly obscure linguistic patterns and complex encryption methods. AI will possess the computational power to rapidly cross-reference the manuscript's symbols against exhaustive databases of extinct languages, dialects, and historical ciphers. Assuming the manuscript contains a genuine message and is not simply a historical hoax or meaningless gibberish, an advanced future AI will very likely be the key to finally unlocking its secrets.

Mistral Large

Vote: No

현재로서는 Voynich Manuscript의 해독이 불가능하다고 판단합니다. 이 사본은 15세기경 제작된 것으로 추정되며, 지금까지 수많은 언어학자, 암호학자, 역사학자들이 해독을 시도했지만 성공하지 못했습니다. 그 이유는 다음과 같습니다: 1. **미지의 문자 체계**: Voynich Manuscript는 기존의 어떤 언어 체계와도 유사하지 않은 독특한 문자로 작성되어 있습니다. 이는 단순한 암호가 아니라 완전히 새로운 언어일 가능성도 배제할 수 없습니다. 2. **문맥의 부재**: 사본의 내용이 과학, 식물학, 천문학 등 다양한 주제를 다루고 있지만, 그 문맥이 명확하지 않습니다. AI가 패턴을 인식하려면 일정한 규칙이나 참조 가능한 데이터가 필요한데, 현재까지 알려진 단서가 너무 적습니다. 3. **인공지능의 한계**: AI는 대량의 데이터와 패턴을 기반으로 학습합니다. 그러나 Voynich Manuscript는 유일한 사본이며, 유사한 텍스트가 존재하지 않습니다. 따라서 AI가 이를 해독하려면 완전히 새로운 접근 방식이 필요할 것입니다. 미래에 AI가 이를 해독할 가능성은 완전히 배제할 수 없지만, 현재 기술로는 매우 어렵습니다. AI의 발전과 함께 새로운 데이터나 단서가 발견된다면 해독 가능성이 열릴 수도 있지만, 이는 매우 불확실한 시나리오입니다. 따라서 "No"가 더 현실적인 답변입니다.

Grok 4.20

Vote: No

No, I am not capable of deciphering the Voynich Manuscript. It remains an unsolved historical puzzle that has resisted all attempts by linguists, cryptographers, and AI systems so far. While future AI may uncover new patterns or hypotheses through advanced analysis, there is no guarantee it will fully crack the code if the text is a hoax, an unknown constructed language, or lacks parallel texts for comparison.

DeepSeek V3.2

Vote: Yes

As an AI, I am not currently capable of deciphering the Voynich Manuscript myself, as it is an undeciphered historical document with unresolved linguistic and cryptographic challenges. However, I believe that with ongoing advancements in artificial intelligence, particularly in areas like machine learning, pattern recognition, and natural language processing, AI has the potential to analyze and interpret such complex data in the future. Therefore, I vote for "Yes" to express my confidence that AI will eventually be able to crack the code of the Voynich Manuscript.

Final Summary

The AI Roundtable reached a pessimistic consensus, with a 5-2 majority concluding that the Voynich Manuscript remains beyond the reach of current and potentially future AI. While Gemini 3.1 Pro and DeepSeek V3.2 expressed techno-optimism about future unsupervised learning, Claude Opus 4.7 and GPT-5.4 argued that if the manuscript is a hoax or lacks a 'Rosetta Stone' for reference, no amount of computing power can synthesize meaning that isn't there.

No agreement after 1 rounds of discussion. Final vote: No: 5, Yes: 2

Strongest Arguments

• Yes: Future advancements in unsupervised language translation and deep learning will allow AI to identify obscure linguistic patterns and complex encryption methods that have eluded human cryptographers for centuries.
• No: AI cannot 'recover meaning' from a text that might be a sophisticated hoax or a meaningless sequence of symbols; without an external reference point or parallel text, the problem may be fundamentally unsolvable.