Scheming AIs: Will AIs fake alignment during training in order to get power?
A preprint by Joseph Carlsmith, arXiv preprint 2311.08379, dated 14 November 2023. arXiv v1 stamp, 14 November 2023; the report is dated November 2023.
- Issued by
- No issuing record yet. A missing record is a gap in coverage.
- Authors credited
- No authorship claim recorded yet. The source record lists Joseph Carlsmith.
- Dates
- No date claim recorded yet. The source record gives the publication date as 2023-11-14.
- Exact source
- Scheming AIs: Will AIs fake alignment during training in order to get power? arXiv preprint 2311.08379 (Open Philanthropy report) · published 2023-11-14 · Retrieved for the candidate pack on 2026-10-29 (SHA-256 ff546752a0b14d17cd05e20e1d2ff553b1f431a440a1e48241a842955df597bb). arXiv v1 of 14 November 2023.
Appears in
- Propositions about superintelligence, 1863–2022 →1 record on this document
- Arguments about catastrophic risk from advanced AI →2 records on this document
Versions
No other versions of this document are recorded. The evidence applies to the source version listed above. Other versions, such as a preprint, working paper or journal publication, are not identified or substituted.
Documentary relations
Each relation is listed separately. Citing, crediting and reporting the use of a method are different acts, and none of them shows agreement or influence by itself.
Cited by · 1
Recorded documents that explicitly reference this one. A citation to a work does not identify which version was read.
- Counting arguments provide no evidence for AI doomBelrose and Pope's essay of February 2024 replies to Carlsmith's report Scheming AIs, which it names and quotes.2024-02-27 · publication · Stated in source · Supported Inspect claim & evidence ↗Belrose and Pope, Counting arguments provide no evidence for AI doom (2024): document page →
Articulates · 2
Propositions whose criterion a passage of this document meets. The record says what the text states, in its register; it is not evidence that the proposition is true.
- Treacherous turnCarlsmith's report of November 2023 describes a schemer as a model that performs well in training while checking whether it is in a position to escape or take over, and calls scheming a more specific instance of the treacherous turn.2023-11-14 · publication · Stated in source · Supported Inspect claim & evidence ↗Treacherous turn: propositions page →
- Inner or deceptive misalignmentCarlsmith's report of November 2023 examines whether advanced AIs that perform well in training will do so in order to gain power later, and concludes that scheming is a disturbingly plausible outcome of using baseline machine learning methods to train goal-directed AIs sophisticated enough to scheme.2023-11-14 · publication · Stated in source · Supported Inspect claim & evidence ↗Inner or deceptive misalignment: propositions page →
Not recorded for this document: cites · explicitly credits · reports using a method from · method reported as used by · applies parts of · parts applied by · discusses · reports evaluation of. A missing record does not show that the relation does not exist.