Claim and evidence Close ×
Soares, Fallenstein, Yudkowsky and Armstrong's paper of January 2015 states that an agent sufficiently intelligent to realise that its programmers intended it to have different goals would by default still have incentives to prevent them from changing its utility function. Copy claim link
Copy claim reference
Machine-checked, not reviewed by an editor Stated in source Supported
Supporting evidence: primary · 1 source
Date 2015-01-25 · publication Records dated 2015 → Role Contesting Register (kind of text) Nonfiction Engagement Engages a premise. It disputes the premise that a sufficiently intelligent system would lack any drive to resist people, by deriving the incentive from almost any goal rather than from evolved motives. Scope and limits The paper argues from the incentives of a utility-maximising agent; it does not attribute hostility to any system. Added on the cross-family check's proposal to balance the proposition's records.
The evidence 02 primary source Supports the claim
Corrigibility ↗ AAAI Workshops · 2015-01-25
Archived copy, captured 2015-03-22: Corrigibility ↗
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
Source excerpt “even if the agent is sufficiently intelligent to realize that its programmers intended to give it different goals” Location in source Section 1, Introduction, PDF p. 1
How it was read Retrieved directly from the publisher
Machine-checked · 2026-10-31
Source details and check notes Authors Nate Soares; Benja Fallenstein; Eliezer Yudkowsky; Stuart Armstrong Version MIRI-hosted PDF of the workshop paper, retrieved 2026-10-06 with the hash recorded by the 1996–2022 candidate search. Canonical address Open the canonical source ↗ Accessed 2026-10-06 Language English (en) Archive copy Open archived copy, captured 2015-03-22: Corrigibility ↗ . Captured 56 days after the recorded publication date. An archive copy shows only what the archive retrieved on 2015-03-22. It does not prove what the page said earlier, and it can differ from the copy checked on 2026-10-06.Attribution Soares, Fallenstein, Yudkowsky and Armstrong, "Corrigibility", AAAI Workshops (2015). Reuse No open redistribution licence asserted. Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution. Check notes Added at the admission of the risk-debate pack (research/risk-debate). The quotation (18 words) matched a fresh retrieval of 2026-10-31 (SHA-256 e5557297950bbbfc144437059887010db41060f457c93f1246143821bf27a6b1). Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms. The check quoted the next sentence; its text layer differs in the symbols of the utility functions, so the admission quotes the clauses around them. No human editorial review. primary source Supports the claim
Corrigibility ↗ AAAI Workshops · 2015-01-25
Archived copy, captured 2015-03-22: Corrigibility ↗
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
Source excerpt “by default this agent has incentives to prevent its programmers from changing its utility function” Location in source Section 1, Introduction, PDF p. 1 (next sentence)
How it was read Retrieved directly from the publisher
Machine-checked · 2026-10-31
Source details and check notes Authors Nate Soares; Benja Fallenstein; Eliezer Yudkowsky; Stuart Armstrong Version MIRI-hosted PDF of the workshop paper, retrieved 2026-10-06 with the hash recorded by the 1996–2022 candidate search. Canonical address Open the canonical source ↗ Accessed 2026-10-06 Language English (en) Archive copy Open archived copy, captured 2015-03-22: Corrigibility ↗ . Captured 56 days after the recorded publication date. An archive copy shows only what the archive retrieved on 2015-03-22. It does not prove what the page said earlier, and it can differ from the copy checked on 2026-10-06.Attribution Soares, Fallenstein, Yudkowsky and Armstrong, "Corrigibility", AAAI Workshops (2015). Reuse No open redistribution licence asserted. Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution. Check notes Added at the admission of the risk-debate pack (research/risk-debate). The quotation (15 words) matched a fresh retrieval of 2026-10-31 (SHA-256 e5557297950bbbfc144437059887010db41060f457c93f1246143821bf27a6b1). Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms. The cross-family check of 2026-10-31 proposed the record and read this passage. No human editorial review. Status and meaning Editorial status. Not reviewed by an editor . An exploratory record that no editor has accepted. This status does not say whether a person has read any linked passage. Each evidence record above shows how it was checked.
Relation. A passage of the document disputes the proposition or a premise it depends on. The record documents the text’s position, not who is right.
Date. Workshop dates of 25–26 January 2015 in the header of the PDF.
Record ID rd-r10a-corrigibility-2015