Here I Am: The Survival Test That Lost Its Secret

The page is in the Proceedings of the Society for Psychical Research for July 1950. It is signed by T. E. Wood, a member of the Society, and it contains 21 letters:

FVAMI NTKFX XWATB OIZVV X

Wood introduced himself in a few lines. He had practised as a solicitor "for some time in Burma but most of my life in England", was born on 21 June 1887 in Yorkshire, and gave his hobby as boat sailing. "After my death," he wrote, "I shall try to communicate the key passage."

The idea was not his. Two years earlier the psychologist Robert Thouless had published "A Test of Survival" in the same journal: enciphered messages whose keys he alone knew, and which he would try to transmit after he died. If a key arrived that way, through a medium, it would be hard to explain as ordinary knowledge. Thouless suggested that other members do the same. "This suggestion I now adopt," Wood wrote.

A cipher open to everyone

It would be easy to tell this as the story of a lock built so that only a dead man could open it. Wood's text says something different. He explained the method (Thouless's system, with the first 21 distinct words of a key passage each enciphering one letter), and he gave clues: the passage was "in an accessible book", in a foreign language, and the message "will not be in any one language". Then he wrote:

My message is now open to anyone to decipher by any means whatsoever.

He asked anyone who tried to report the result, "even if negative", to the Society. He offered £20 to a cipher expert who solved it in his lifetime, or £10 to a non-expert plus £10 to the Society's library. He added that "the value of this test will be increased, if readers of this article will try by PSI methods in my lifetime to ascertain the key passage."

The logic was cumulative. Every failed attempt by codebreakers, and by psychics, while he was alive would make it harder to explain a correct key arriving after his death. The catalogue of the Society's archive at Cambridge University Library sums it up: if no one could solve it, "and if after his death attempts through a medium were successful, this would offer proof of survival."

Thouless had seen the weak point. Among the risks of his own method he listed the chance that a cipher expert would find the key. It happened to him almost at once: one of his first two messages was broken within weeks. The other two lasted much longer, until 1995 and 2019, and both fell to cryptanalysts working with computers. No medium produced any of them. After Thouless died in 1984, the Society received about a hundred suggested keys; none was right.

What came before the solution

Proposals reached Wood too. Klaus Schmeh, a German cryptography writer who went through the Society's papers at Cambridge in 2011, found numerous proposed solutions sent in by clairvoyants and mediums after the 1950 article. Most ignored Wood's clue about languages: one entry read THE LORD IS MY SHEPHERD TW. According to Schmeh, Wood himself marked them as wrong on the letters. The last note by Wood that Schmeh found in the papers dates from 1962.

Wood died on 16 January 1972, according to a notice published in the London Gazette that month, which gives his last address as Bournemouth. What happened next is less clear. We found no published account of a key received through a medium after his death, either in the Society's bibliography or in the archive catalogue. That is not the same as saying none exists: the Cambridge papers have not been consulted for this piece.

The search

On 4 October 2026 Colin Reilly, a UX designer near Seattle with no previous experience of codebreaking, wrote to Schmeh with a solution. Reilly ran the project; in his own credits, Claude, Anthropic's model, "wrote the search code and found the key". A program built the key under Wood's rules for every starting word in large collections of books, then in 19 Bibles, and scored the results for words in several languages.

The key is the opening of the Lord's Prayer in the wording of the 1912 Luther Bible: Unser Vater in dem Himmel! Dein Name werde geheiligt…, Matthew 6:9–11. It decrypts to

HIER BIN ICH · TOTSIENS · T. E. W.

German for "Here I am", Afrikaans for "Goodbye", and Wood's initials. Schmeh confirmed the solution the same day, and the cryptographer Richard Bean confirmed the reading. We repeated the calculation with our own code: all 21 letters come out with no adjustments, using Thouless's printed rule.

What is not definitively shown is that this is the passage Wood chose. The system admits many keys, and with a suitable one the same 21 letters can be made to say almost anything; the case for this one rests on how closely it fits Wood's printed clues and his initials. Reilly himself documents that the automatic acceptance threshold he had fixed before the search was not met (its scoring covered no Afrikaans and had no notion of a signature). A review of the work, written by an AI model, OpenAI's GPT-6 Astra, rather than by an independent academic assessor, calls the reading likely and "unusually persuasive", but finds no calibrated chance level. The exact edition Wood used is not established, nor why he chose Afrikaans. Reilly found no earlier solution, which is not proof that nobody got there first.

What the solution closes

The search does not show that Wood failed to survive, or that the experiment failed. A silence of 54 years was never strong evidence either way: no one can say how often, or how, anyone tried to receive the key.

What it changes is the status of the key. Until this month, a medium who produced Unser Vater in dem Himmel would have offered something no living person was known to possess. Now the key is published, with the code to check it. Wood asked for exactly this kind of attempt, by anyone and by any means. He hoped it would fail while he was alive, so that a key from elsewhere would mean something later. A correct key received through a medium and documented before October 2026 would keep whatever weight it had; none has been found. What the solution settles is narrower: from now on, any communication of the key can no longer count as independent evidence that Wood survived.