AI becomes the linguist's secret weapon in decoding ancient languages

For centuries, cracking lost languages has required a foundational clue—a bilingual text like the Rosetta Stone or ties to a known linguistic relative. But some of history's most stubborn writing systems lack even that basic foothold.

The problem: languages without anchors#

Linear A, the Bronze Age script of the Minoan civilization on Crete, remains isolated from any confirmed connection to other languages. Etruscan, spoken in pre-Roman Italy, is only marginally better off: fragments from tomb inscriptions have preserved bits of vocabulary, yet grammar and meaning remain largely cryptic. These linguistic orphans have frustrated scholars for generations, resisting the traditional tools that made earlier decipherments possible.

AI as intellectual amplifier#

Recently, artificial intelligence has entered the arena—not as a replacement for human expertise, but as a turbocharger for educated intuition. In June 2026, a self-taught AI programmer and amateur linguist made waves with a proposed solution to Linear A. His method began with a single hypothesis: that an unknown word in a prayer text stemmed from a Semitic root meaning "to dwell" or "to inhabit."

Using AI-powered scripts, he then systematically tested that phonetic pattern across the entire corpus of known Linear A inscriptions. The result: proposed values for 40 characters and a draft lexicon of 408 words—all suggesting Linear A belongs to the Semitic language family, alongside Hebrew and Aramaic.

Where human insight meets machine power#

What's crucial to understand is the precise division of labor. The initial leap—the creative spark that launched the investigation—came from the human researcher. The AI's contribution was systematic and scalable: running through thousands of pattern-matching tests at speeds no person could achieve. The machine tested hypotheses; the human dreamed them up.