AI Engineer Claims to Have Cracked Linear A
Points and comments are a snapshot, not live.
An AI engineer claims to have deciphered the Bronze-age Minoan writing system Linear A.
Tom Di Mino, a self-taught AI engineer and amateur linguist, claims to have deciphered Linear A, a Minoan script used from 1800 to 1450 BC. He believes the script maps to an extinct Semitic language that was a precursor to biblical Hebrew. The key insight came on May 22 while analyzing prayer inscriptions; he identified a verb root meaning "to dwell" using an unknown sign. Di Mino has proposed readings for 40 signs, including 13 previously unknown ones, and produced a lexicon of 408 translated terms. Linguistics experts at Rutgers and Cambridge are reviewing his work.
What commenters are saying
Commenters are broadly cautious, awaiting verification from experts. The post's author, who knows Di Mino socially, notes his work is credible enough for review and that he has produced over 300 translations. Skeptics question the lack of a preprint or direct write-up. Some compare the process to prior decipherments like Linear B, Akkadian, and Maya, emphasizing the need for held-out test texts. Two camps emerge: those impressed by the systematic approach using Claude Code, and those insisting on peer-reviewed confirmation before accepting the claim.