Back to the scorecard

Interest in neural methods grows

Reference judgmentCorrectResearch snapshot · October 15, 2026
Source confidence & research status
i
Editorial paraphrase · evidence v0.3. AI-assisted editorial assessments; original book editions not independently collated. Attribution and outcome confidence are recorded separately.

Source state: AUTHOR RETROSPECTIVE CHECKED.

Attribution confidence
MEDIUM
Outcome confidence · deadline / review
MEDIUM / MEDIUM
Timing confidence / kind
MEDIUM · DOCUMENTED BY
Population
Technology markets and relevant users; no universality unless the forecast expressly requires it.

Our reading of the claim

Reference (stated reasonable reading)

Correct

Criterion: Interest and practical work on neural/evolutionary methods increase while conventional computation remains important.

Deadline: 2009 · Included in this comparison

At original deadlineFULFILLED
By review dateFULFILLED

The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development. The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.

Deadline and follow-up notes

At deadline: The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development.

By review: The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.

Evidence: N60 (opens in a new tab)N35 (opens in a new tab)

View this reading in the scorecard →
01

The forecast and its attribution.

EDITORIAL PARAPHRASE

Interest in neural methods grows

How My Predictions Are Faring — Ray Kurzweil

1999 · Attribution / year unconfirmed
P01 printed p.31, Computers #24

AUTHOR RETROSPECTIVE REPRINT

Source publication: 2010-10 · Access: text checked

Official author archive. Cover says October 2010, but the available file contains later updates. Reprinted forecast passages are used for attribution; the author’s favorable self-grades are not independent outcome evidence. Original book editions were not inspected. Page locators below are PRINTED pages, not PDF viewer indices.

Catalogue entry P01
Open listed source (opens in a new tab)
Normalized claim
Interest in neural methods grows
Success criteria
Interest and practical work on neural/evolutionary methods increase while conventional computation remains important.
Claim type
Trend
02

Timeline

PREDICTION MADE1999Attributed forecast year; original edition not collated
ORIGINAL TARGET2009Deadline supplied by this assessment
FULL REALIZATIONBy 2009Evidence bound, not first achievement year

DOCUMENTED BY: Documented by 2009

End of named target year.

Outcome at original deadlineFULFILLED
Outcome at reviewFULFILLED

Supplied target wording: 2009. The original deadline is preserved if a later forecast revises it. Forecast horizon: not established. Drift is a calendar-year difference, not a precise elapsed duration.

03

The evidence.

CONTEXT · Publication date not recorded

Hinton, Osindero and Teh: Fast learning for deep belief nets (opens in a new tab)

Primary institutional, research, or issuer source. Dates in the claim refer to the observation or product event, not the review date.. Primary institutional, research, or issuer source. Dates in the claim refer to the observation or product event, not the review date.

Source checked as of 2026-10-15
PRIMARY PAPER

Source publication: 2006-07 · Access: web text or indexed extract checked

Primary institutional, research, or issuer source. Dates in the claim refer to the observation or product event, not the review date.

Catalogue entry N60
CONTEXT · Publication date not recorded

Krizhevsky et al.: ImageNet Classification with Deep Convolutional Neural Networks (opens in a new tab)

Primary institutional, research, or issuer source. Dates in the claim refer to the observation or product event, not the review date.. Primary institutional, research, or issuer source. Dates in the claim refer to the observation or product event, not the review date.

Source checked as of 2026-10-15
PRIMARY PAPER

Source publication: 2012 · Access: web text or indexed extract checked

Primary institutional, research, or issuer source. Dates in the claim refer to the observation or product event, not the review date.

Catalogue entry N35

Capability, adoption, and timing

A related breakthrough is not automatically fulfillment. Review must check the actual capability, how widely it was required to be used, the specified technology, and the original deadline. A prototype cannot establish a claim about routine or widespread use.

04

Research notes & history

One paper alone is not a complete bibliometric series. This qualitative directional assessment does not independently quantify the genetic-algorithm component or the share of all instructions executed.

Open questions

  • Original-edition collation remains an attribution task; it does not erase the stated outcome assessment.
  • An earlier documented instance could refine timing; the documented-by date is not claimed to be the invention year.
Imported from
kurzweil_verification_v0_3_package/predictions.canonical.v0.3.json
Legacy identifier
None
Forecast family
K054
Parent statement
Not assigned
File checksum
93ae54d9ea2f1bd1817c481b85e2b707812bcd934c74d0f1715fa098d9b5f51b
Inspect preserved source fields
release_record
{"population_scope":"Technology markets and relevant users; no universality unless the forecast expressly requires it.","attribution_confidence":"MEDIUM","interpretation_confidence":"HIGH","outcome_confidence":{"at_deadline":"MEDIUM","at_review":"MEDIUM"},"timing_confidence":"MEDIUM","date_uncertainty_affects_verdict":false,"interpretations":{"reference":{"criterion":"Interest and practical work on neural/evolutionary methods increase while conventional computation remains important.","at_deadline":"FULFILLED","at_review":"FULFILLED","rationale":"The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development. The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.","relaxed_or_changed_conditions":[],"kind":"REFERENCE_READING","documented_by_year":2009,"at_deadline_rationale":"The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development.","at_review_rationale":"The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.","evidence_source_ids":["N60","N35"],"verdict":"CORRECT","same_as_reference":false,"eligible_for_deadline_metrics":true,"deadline_year":2009,"outcome_assessment_complete":true},"strict":{"criterion":"Interest and practical work on neural/evolutionary methods increase while conventional computation remains important.","at_deadline":"FULFILLED","at_review":"FULFILLED","rationale":"The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development. The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.","relaxed_or_changed_conditions":[],"kind":"REFERENCE_READING","documented_by_year":2009,"at_deadline_rationale":"The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development.","at_review_rationale":"The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.","evidence_source_ids":["N60","N35"],"verdict":"CORRECT","same_as_reference":true,"eligible_for_deadline_metrics":true,"deadline_year":2009,"outcome_assessment_complete":true},"broader":{"criterion":"Interest and practical work on neural/evolutionary methods increase while conventional computation remains important.","at_deadline":"FULFILLED","at_review":"FULFILLED","rationale":"The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development. The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.","relaxed_or_changed_conditions":[],"kind":"REFERENCE_READING","documented_by_year":2009,"at_deadline_rationale":"The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development.","at_review_rationale":"The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.","evidence_source_ids":["N60","N35"],"verdict":"CORRECT","same_as_reference":true,"eligible_for_deadline_metrics":true,"deadline_year":2009,"outcome_assessment_complete":true}},"timing":{"first_full_achievement_year":null,"documented_by_year":2009,"achievement_interval_years":null,"kind":"DOCUMENTED_BY","label":"Documented by 2009","drift_years":null,"drift_range_years":null},"deduplication":{"is_canonical":true,"is_repeated_identical_claim":false,"counting_unit":"DISTINCT_FORECAST_VERSION","note":"Different deadlines or materially different thresholds remain separate versions; identical legacy repetitions are linked, not recounted."},"revision_history":[],"is_new_in_v03":true,"counting_policy":{"unit":"distinct_forecast_version","identical_repetitions_count_once":true,"alternative_interpretations_count_once":true,"legacy_conditional_rows_count":false},"id":"K054","slug":"k054-interest-in-neural-methods-grows","title":"Interest in neural methods grows","prediction_paraphrase":"Interest in neural methods grows","wording_kind":"EDITORIAL_PARAPHRASE","original_quote":null,"primary_category":"AI & software","tags":[],"subcategory":"","claim_type":"TREND","family_id":"K054","parent_statement_id":null,"legacy_ids":[],"origin":"NEW_SOURCE_EXTRACTION","publication":{"forecast_year":1999,"year_basis":"AUTHOR_ADAPTATION_OR_RETROSPECTIVE","original_book_edition_checked":false,"forecast_year_status":"ATTRIBUTED_1999"},"target":{"text":"2009","deadline_year":2009,"precision":"YEAR","year_start":2009,"year_end":2009,"interpretation":"End of named target year."},"source_verification":{"state":"AUTHOR_RETROSPECTIVE_CHECKED","confidence":"MEDIUM","book_text_not_independently_collated":true},"success_criteria":"Interest and practical work on neural/evolutionary methods increase while conventional computation remains important.","assessment":{"at_deadline_rationale":"The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development.","at_review_rationale":"The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.","verdict_qualifier":null,"outcome_evidence_basis":"DIRECT_AND_CONTEXTUAL_EVIDENCE_WITH_EXPLICIT_EDITORIAL_INTERPRETATION","attribution_is_separate":true,"first_event_date_required_for_verdict":false,"at_review_semantics":"FULL_REALIZATION_HAS_OCCURRED_BY_REVIEW_NOT_NECESSARILY_CURRENT_PRODUCT_SALES","review_state":"EDITORIAL_ASSESSMENT_COMPLETE","verdict":"CORRECT","as_of":"2026-10-15","at_deadline":"FULFILLED","at_review":"FULFILLED","confidence":"MEDIUM","disputed_interpretation":false,"benchmark_ready":true,"rationale":"The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development. The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.","first_full_achievement_year":null,"full_achievement_documented_by_year":2009,"drift_years":null,"drift_range_years":null,"still_unfulfilled_as_of_review":false,"eligible_for_deadline_metrics":true,"withheld_reason":null},"prediction_sources":[{"source_id":"P01","locator":"P01 printed p.31, Computers #24","role":"FORECAST_ATTRIBUTION"}],"evidence":[{"source_id":"N60","locator":"Primary institutional, research, or issuer source. Dates in the claim refer to the observation or product event, not the review date.","role":"OUTCOME_EVIDENCE"},{"source_id":"N35","locator":"Primary institutional, research, or issuer source. Dates in the claim refer to the observation or product event, not the review date.","role":"OUTCOME_EVIDENCE"}],"limitations":"One paper alone is not a complete bibliometric series. This qualitative directional assessment does not independently quantify the genetic-algorithm component or the share of all instructions executed.","open_questions":["Original-edition collation remains an attribution task; it does not erase the stated outcome assessment.","An earlier documented instance could refine timing; the documented-by date is not claimed to be the invention year."],"legacy_relations":[]}

Review history

— Evidence v0.3: EDITORIAL_ASSESSMENT_COMPLETE; supplied verdict CORRECT. The 2006 deep-belief-net paper is a concrete pre-deadline advance consistent with renewed neural-method development. The 2012 ImageNet result and subsequent AI systems decisively reinforce the broad direction; no majority-of-all-computation threshold is imposed on this weaker claim.

Related forecasts.

No forecast family has been resolved for this record. Related statements and revised deadlines will be linked during research.