Claim and evidence Close ×
Bostrom’s Superintelligence (2014) calls “the treacherous turn” the behaviour of an AI that cooperates while it is weak and, once strong enough, pursues its final values without effective opposition. Copy claim link
Copy claim reference
Machine-checked Stated in source Supported
Supporting evidence: primary · 1 source
Date 2014-07-03 · publication Records dated 2014 → Role Articulation, Earliest located use of the name Register (kind of text) Nonfiction Name used “treacherous turn” Scope and limits Rests on a reading of the first impression by a model of another family on 2026-11-07; Selekoa has not read the book, and no quotation is given. Its use of the name is the earliest verified among the three texts the witness check compared: Danaher’s post (dated 29 July 2014, first captured 8 August 2014) says it takes the term from Bostrom, and Grace’s reading-group page (dated 31 August 2014) is not shown to carry the wording by that date. Earliest located means the earliest text found by the collection’s searches and that check, not a first statement or a coinage.
The evidence 02 primary source Supports the claim
Superintelligence: Paths, Dangers, Strategies (first edition, first impression) ↗ Oxford University Press · 2014-07-03
Copy source reference
Cite the original source Cite the original document, not Selekoa. Check these details against it.
Copy APA
Download BibTeX Download RIS Download CSL-JSON
Paraphrase Bostrom names the phenomenon he has just described “the treacherous turn”; the term is also the heading of the section that begins on p. 116.
Location in source Chapter 8, printed p. 118
Machine-checked · 2026-11-07
Source details and check notes Authors Nick Bostrom Version UK hardback, first edition, first impression, ISBN 978-0-19-967811-2, cited by page. Selekoa has not read the book: no publisher or library preview of the pages could be reached on 2026-11-07 (the preview service asked for a CAPTCHA, which was not answered). The pages cited were read by a model of another family on 2026-11-07 (research/recheck-witnesses-2026-11-07/codex-report.md, item 1); the copy it read is not one Selekoa cites. The date and ISBN are those of the Oxford Martin School’s listing of the hardback, read on 2026-11-07. Canonical address Open the canonical source ↗ Accessed 2026-11-07 Language English (en) Attribution Nick Bostrom, Superintelligence: Paths, Dangers, Strategies (Oxford University Press, 2014). Reuse No open redistribution licence asserted. Bibliographic metadata and labelled paraphrases of a reading by another model family; no quotation. Check notes The checker’s reading, labelled as such: a model of another family read this page on 2026-11-07 (research/recheck-witnesses-2026-11-07/codex-report.md, item 1) and identified the copy as the first impression by its copyright page. Selekoa has not read the page. No human editorial review. primary source Supports the claim
Superintelligence: Paths, Dangers, Strategies (first edition, first impression) ↗ Oxford University Press · 2014-07-03
Copy source reference
Cite the original source Cite the original document, not Selekoa. Check these details against it.
Copy APA
Download BibTeX Download RIS Download CSL-JSON
Paraphrase The boxed definition describes an AI that behaves cooperatively while it is weak and, once strong enough, strikes without warning and pursues its final values without effective opposition.
Location in source Chapter 8, boxed definition, printed p. 119
Machine-checked · 2026-11-07
Source details and check notes Authors Nick Bostrom Version UK hardback, first edition, first impression, ISBN 978-0-19-967811-2, cited by page. Selekoa has not read the book: no publisher or library preview of the pages could be reached on 2026-11-07 (the preview service asked for a CAPTCHA, which was not answered). The pages cited were read by a model of another family on 2026-11-07 (research/recheck-witnesses-2026-11-07/codex-report.md, item 1); the copy it read is not one Selekoa cites. The date and ISBN are those of the Oxford Martin School’s listing of the hardback, read on 2026-11-07. Canonical address Open the canonical source ↗ Accessed 2026-11-07 Language English (en) Attribution Nick Bostrom, Superintelligence: Paths, Dangers, Strategies (Oxford University Press, 2014). Reuse No open redistribution licence asserted. Bibliographic metadata and labelled paraphrases of a reading by another model family; no quotation. Check notes The checker’s reading, labelled as such: a model of another family read this page on 2026-11-07 (research/recheck-witnesses-2026-11-07/codex-report.md, item 1) and identified the copy as the first impression by its copyright page. Selekoa has not read the page. No human editorial review. Status and meaning Editorial status. Machine-checked . Machines found this record and checked it against its sources; no person accepted it individually. Each evidence record above shows how it was checked.
Relation. A passage of the document meets the proposition’s criterion. The record documents what the text states, in its register (fiction and satire are labelled); it is not evidence that the proposition is true and implies no influence on or from another text.
Date. Publication of the UK hardback, as the Oxford Martin School’s listing gives it.
Record ID ts-b1-bostrom-2014-superintelligence