Claim · 2011 · publication · Stated in source · Supported

“Machine Intelligence Survey” (2011) reports a survey result (Sandberg and Bostrom, informal poll at the FHI Winter Intelligence conference (FHI Technical Report 2011-1)): “HLMI”; horizon as stated: median estimates: 10% chance of HLMI 2028; 50% chance 2050; 90% chance 2150. The record says nothing about whether the event happened.

One recorded claim. The claim panel gives its qualification, the source passages with their place in each source, and how it was checked.

What it connects

Machine Intelligence Survey → states a prediction → HLMI

Where it appears

Predictions on record →

Claim and evidence

Machine Intelligence Surveystates a prediction ↓HLMI

Machine-checkedStated in sourceSupported

Supporting evidence: primary · 1 source

Date: 2011 · publication Records dated 2011 →

Scope and limitsRests on transcriptions by a reviewer from a different model family of a copy the preparing assistant could not reach; that reviewer read the copy again on 2026-11-05 and confirms the quoted wording. The report calls the poll "informal". The question assumed that no global catastrophe halts progress and allowed the answer "never". Each figure is the median of the respondents' estimates for one probability level, not a forecast by the authors or a combined distribution. The figures belong to the report, not to any respondent. The poll date (16 January 2011), the report year (2011) and the capture (2016) are separate dates.

The evidence 05

primary sourceSupports the claim

Machine Intelligence Survey (FHI Technical Report #2011-1)

Future of Humanity Institute, University of Oxford · 2011

Archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1)

Cite the original source

Cite the original document, not Selekoa. Check these details against it.

Paraphrase

The report states that an “informal poll” was conducted at the FHI Winter Intelligence conference on machine intelligence on “16/1 2011”.

Location in source Printed p. 1 / PDF p. 2

Machine-checked · 2026-11-03

Source details and check notes

Authors: Anders Sandberg; Nick Bostrom
Version: Internet Archive capture 20161224100323 of the PDF, read by a reviewer from a different model family; fhi.ox.ac.uk refused the preparing assistant. Technical report 2011-1; the poll was taken at the FHI Winter Intelligence conference on 16 January 2011.
Canonical address: Open the canonical source ↗
Accessed: 2026-11-05
Language: English (en)
Archive copy: Open archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1). An archive copy shows only what the archive retrieved on 2016-12-24. It does not prove what the page said earlier, and it can differ from the copy checked on 2026-11-05.
Attribution: Anders Sandberg, Nick Bostrom, “Machine Intelligence Survey (FHI Technical Report #2011-1)”, 2011.
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: Paraphrase, with the record’s words in quotation marks, of a passage read by a reviewer from a different model family on 2026-11-03. The capture's length is not known here, so each passage is a labelled paraphrase with the record's words in quotation marks; the three median sentences are given in one paraphrase. Copy read: Internet Archive capture 20161224100323, read by a reviewer from a different model family on 2026-11-03 and again on 2026-11-05; the preparing assistant could not reach it. Machine-checked; no human editorial review.

primary sourceSupports the claim

Machine Intelligence Survey (FHI Technical Report #2011-1)

Future of Humanity Institute, University of Oxford · 2011

Archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1)

Cite the original source

Cite the original document, not Selekoa. Check these details against it.

Paraphrase

It gives the median estimates of when there will be a given chance of HLMI: “10% chance of HLMI was 2028”, “50% chance of HLMI was 2050” and “90% chance of HLMI was 2150”.

Location in source Printed p. 1 / PDF p. 2

Machine-checked · 2026-11-03

Source details and check notes

Authors: Anders Sandberg; Nick Bostrom
Version: Internet Archive capture 20161224100323 of the PDF, read by a reviewer from a different model family; fhi.ox.ac.uk refused the preparing assistant. Technical report 2011-1; the poll was taken at the FHI Winter Intelligence conference on 16 January 2011.
Canonical address: Open the canonical source ↗
Accessed: 2026-11-05
Language: English (en)
Archive copy: Open archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1). An archive copy shows only what the archive retrieved on 2016-12-24. It does not prove what the page said earlier, and it can differ from the copy checked on 2026-11-05.
Attribution: Anders Sandberg, Nick Bostrom, “Machine Intelligence Survey (FHI Technical Report #2011-1)”, 2011.
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: Paraphrase, with the record’s words in quotation marks, of a passage read by a reviewer from a different model family on 2026-11-03. The capture's length is not known here, so each passage is a labelled paraphrase with the record's words in quotation marks; the three median sentences are given in one paraphrase. Copy read: Internet Archive capture 20161224100323, read by a reviewer from a different model family on 2026-11-03 and again on 2026-11-05; the preparing assistant could not reach it. Machine-checked; no human editorial review.

primary sourceSupports the claim

Machine Intelligence Survey (FHI Technical Report #2011-1)

Future of Humanity Institute, University of Oxford · 2011

Archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1)

Cite the original source

Cite the original document, not Selekoa. Check these details against it.

Paraphrase

Question 1 asks by what year respondents would assign a 10%, 50% or 90% chance of the development of human-level machine intelligence, “assuming no global catastrophe halts progress”, and the next sentence allows the answer “never”.

Location in source Printed p. 1 / PDF p. 2, question 1

Machine-checked · 2026-11-05

Source details and check notes

Authors: Anders Sandberg; Nick Bostrom
Version: Internet Archive capture 20161224100323 of the PDF, read by a reviewer from a different model family; fhi.ox.ac.uk refused the preparing assistant. Technical report 2011-1; the poll was taken at the FHI Winter Intelligence conference on 16 January 2011.
Canonical address: Open the canonical source ↗
Accessed: 2026-11-05
Language: English (en)
Archive copy: Open archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1). An archive copy shows only what the archive retrieved on 2016-12-24. It does not prove what the page said earlier, and it can differ from the copy checked on 2026-11-05.
Attribution: Anders Sandberg, Nick Bostrom, “Machine Intelligence Survey (FHI Technical Report #2011-1)”, 2011.
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: Paraphrase, with the record’s words in quotation marks, of a passage read by a reviewer from a different model family on 2026-11-05. The capture's length is not known here, so each passage is a labelled paraphrase with the record's words in quotation marks; the three median sentences are given in one paraphrase. Copy read: Internet Archive capture 20161224100323, read by a reviewer from a different model family on 2026-11-03 and again on 2026-11-05; the preparing assistant could not reach it. Machine-checked; no human editorial review.

primary sourceSupports the claim

Machine Intelligence Survey (FHI Technical Report #2011-1)

Future of Humanity Institute, University of Oxford · 2011

Archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1)

Cite the original source

Cite the original document, not Selekoa. Check these details against it.

Paraphrase

The appended questionnaire defines a human-level machine intelligence as one that can substitute for humans in virtually all cognitive tasks, including those requiring scientific creativity, common sense or social skills.

Location in source PDF p. 12, appended questionnaire, definition before question 1

Machine-checked · 2026-11-05

Source details and check notes

Authors: Anders Sandberg; Nick Bostrom
Version: Internet Archive capture 20161224100323 of the PDF, read by a reviewer from a different model family; fhi.ox.ac.uk refused the preparing assistant. Technical report 2011-1; the poll was taken at the FHI Winter Intelligence conference on 16 January 2011.
Canonical address: Open the canonical source ↗
Accessed: 2026-11-05
Language: English (en)
Archive copy: Open archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1). An archive copy shows only what the archive retrieved on 2016-12-24. It does not prove what the page said earlier, and it can differ from the copy checked on 2026-11-05.
Attribution: Anders Sandberg, Nick Bostrom, “Machine Intelligence Survey (FHI Technical Report #2011-1)”, 2011.
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: Paraphrase, with the record’s words in quotation marks, of a passage read by a reviewer from a different model family on 2026-11-05. The capture's length is not known here, so each passage is a labelled paraphrase with the record's words in quotation marks; the three median sentences are given in one paraphrase. Copy read: Internet Archive capture 20161224100323, read by a reviewer from a different model family on 2026-11-03 and again on 2026-11-05; the preparing assistant could not reach it. Machine-checked; no human editorial review.

primary sourceSupports the claim

Machine Intelligence Survey (FHI Technical Report #2011-1)

Future of Humanity Institute, University of Oxford · 2011

Archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1)

Cite the original source

Cite the original document, not Selekoa. Check these details against it.

Paraphrase

The report says that 35 surveys were collected, suggesting a response rate of about 41%.

Location in source Printed p. 1 / PDF p. 2, paragraph before the estimates

Machine-checked · 2026-11-05

Source details and check notes

Authors: Anders Sandberg; Nick Bostrom
Version: Internet Archive capture 20161224100323 of the PDF, read by a reviewer from a different model family; fhi.ox.ac.uk refused the preparing assistant. Technical report 2011-1; the poll was taken at the FHI Winter Intelligence conference on 16 January 2011.
Canonical address: Open the canonical source ↗
Accessed: 2026-11-05
Language: English (en)
Archive copy: Open archived copy, captured 2016-12-24: Machine Intelligence Survey (FHI Technical Report #2011-1). An archive copy shows only what the archive retrieved on 2016-12-24. It does not prove what the page said earlier, and it can differ from the copy checked on 2026-11-05.
Attribution: Anders Sandberg, Nick Bostrom, “Machine Intelligence Survey (FHI Technical Report #2011-1)”, 2011.
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: Paraphrase, with the record’s words in quotation marks, of a passage read by a reviewer from a different model family on 2026-11-05. The capture's length is not known here, so each passage is a labelled paraphrase with the record's words in quotation marks; the three median sentences are given in one paraphrase. Copy read: Internet Archive capture 20161224100323, read by a reviewer from a different model family on 2026-11-03 and again on 2026-11-05; the preparing assistant could not reach it. Machine-checked; 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. The document states a prediction: an event, by a date or within a horizon, in the document’s own words. The record is dated by the document, never by the target date. It documents what the text states and when; it says nothing about whether the event happened, and it implies no agreement or influence.

Date. Date of the survey report; when the survey was conducted is in the record’s survey details. Poll taken on 16 January 2011 ("16/1 2011"); the report is numbered 2011-1.

Record ID pr-c-fhi-2011-hlmi