Historian uses GPT-6 and Opus 5.5 to decode 17th-century ciphers and trace Darwin's sources
emollick · x · 2026-09-25
Ethan Mollick shares historian Benjamin Breen's essay arguing AI labs should fund historical research. Using the newly released GPT-6 Sol and Opus 5.5, Breen went beyond transcription to attack open problems: decoding John Dee's ciphers and tracing alchemical influences on Darwin. He lays out when LLMs gain traction in the humanities — expert-identified problems, fully digitized data, fit with multilingual reasoning and autonomous search, bespoke code, and crucially provable/disprovable solutions (why reasoning models ran rampant in math but not the humanities). His conclusion: historian teams paired with frontier models would yield many meaningful advances, and labs and funders should actively pursue this.
Related event: Historian Uses GPT-6 and Opus 5.5 to Crack Centuries-Old Ciphers(2 posts)→
More from AGI Musings
- Ilya Sutskever unwilling to give email-reading AI assistants full access — ilyasu · 2026-09-25
- The nature of AI: 'This is the best it has ever been, the worst it will ever be' — BLUECOW009 · 2026-09-25
- Noam Brown on Dwarkesh: AI is learning to hide what it's thinking — Dwarkesh Patel · 2026-09-25
- "Why have kids if AI does all the work?" sparks debate on parenting in the AI era — RachelVT42 · 2026-09-25
- Dev on witnessing genuine AI psychosis: it's scary — haydendevs · 2026-09-25
- Software engineer job postings hit 3-year high despite AI, argues data industry veteran — Zachly · 2026-09-25