beforeword · English transcription of the sent letter
Intended recipient: Murray Shanahan
Author: Kirill Shebetov · beforeword
Sent to: m.shanahan@imperial.ac.uk
Sending date in the supplied materials: 2026-10-02 · 00:28
The sending timezone is not supplied.
Status: shown in the Sent folder; no reply published.
The published materials do not establish receipt or reading by the recipient.
Record: https://beforeword.xyz/letters/researchers/shanahan/en/#record
Record data: https://beforeword.xyz/letters/researchers/shanahan/record.json
Russian translation: https://beforeword.xyz/letters/researchers/shanahan/#letter
Transcribed from author-supplied screenshots. Screen line wraps are joined and overlaps included once; wording and punctuation are retained.
beforeword’s context and links to discussed publications are separate on the letter page.

--- BEGIN ENGLISH TRANSCRIPTION ---

Subject: The human side of your account of LLM belief

Dear Murray,

Your clarification of “Talking About Large Language Models” places the discussion of belief in the use of words, with attention to the capabilities and behaviour of the system being described.

beforeword proposes a requirement for explanations that rely on such attributions: keep the recorded words, the interpretation added to them, and the grounds for that interpretation separately inspectable. This includes human self-descriptions.

For “I believe this”, a human source may warrant a different interpretation from an AI source. The proposed requirement is to make the grounds for that difference explicit wherever the attribution supports a further conclusion.

The reading procedure and worked examples:
https://beforeword.xyz/model/en/

Please consider a brief critical response on the human case—particularly whether this requirement clarifies belief-attribution or asks for more justification than ordinary usage requires.

Kirill Shebetov
beforeword

+

--- END ENGLISH TRANSCRIPTION ---