- 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":"HIGH","at_review":"HIGH"},"timing_confidence":"LOW","date_uncertainty_affects_verdict":false,"interpretations":{"reference":{"criterion":"A circa-$1,000 consumer budget, inflation-adjusted from 1999, buys approximately 20 petaflops of ordinary arithmetic capacity.","at_deadline":"NOT_FULFILLED","at_review":"NOT_FULFILLED","rationale":"Consumer hardware in 2019 was orders of magnitude below 20 petaflops in conventional FP32 throughput. Current checked consumer GPU specifications remain far short on that basis. Inflation adjustment and GPU-only accounting do not rescue the order-of-magnitude mismatch.","relaxed_or_changed_conditions":[],"kind":"REFERENCE_READING","documented_by_year":null,"at_deadline_rationale":"Consumer hardware in 2019 was orders of magnitude below 20 petaflops in conventional FP32 throughput.","at_review_rationale":"Current checked consumer GPU specifications remain far short on that basis. Inflation adjustment and GPU-only accounting do not rescue the order-of-magnitude mismatch.","evidence_source_ids":["N09","N10","N11"],"verdict":"WRONG","same_as_reference":false,"eligible_for_deadline_metrics":true,"deadline_year":2019,"outcome_assessment_complete":true},"strict":{"criterion":"A circa-$1,000 consumer budget, inflation-adjusted from 1999, buys approximately 20 petaflops of ordinary arithmetic capacity.","at_deadline":"NOT_FULFILLED","at_review":"NOT_FULFILLED","rationale":"Consumer hardware in 2019 was orders of magnitude below 20 petaflops in conventional FP32 throughput. Current checked consumer GPU specifications remain far short on that basis. Inflation adjustment and GPU-only accounting do not rescue the order-of-magnitude mismatch.","relaxed_or_changed_conditions":[],"kind":"REFERENCE_READING","documented_by_year":null,"at_deadline_rationale":"Consumer hardware in 2019 was orders of magnitude below 20 petaflops in conventional FP32 throughput.","at_review_rationale":"Current checked consumer GPU specifications remain far short on that basis. Inflation adjustment and GPU-only accounting do not rescue the order-of-magnitude mismatch.","evidence_source_ids":["N09","N10","N11"],"verdict":"WRONG","same_as_reference":true,"eligible_for_deadline_metrics":true,"deadline_year":2019,"outcome_assessment_complete":true},"broader":{"criterion":"A circa-$1,000 consumer budget, inflation-adjusted from 1999, buys approximately 20 petaflops of ordinary arithmetic capacity.","at_deadline":"NOT_FULFILLED","at_review":"NOT_FULFILLED","rationale":"Consumer hardware in 2019 was orders of magnitude below 20 petaflops in conventional FP32 throughput. Current checked consumer GPU specifications remain far short on that basis. Inflation adjustment and GPU-only accounting do not rescue the order-of-magnitude mismatch.","relaxed_or_changed_conditions":[],"kind":"REFERENCE_READING","documented_by_year":null,"at_deadline_rationale":"Consumer hardware in 2019 was orders of magnitude below 20 petaflops in conventional FP32 throughput.","at_review_rationale":"Current checked consumer GPU specifications remain far short on that basis. Inflation adjustment and GPU-only accounting do not rescue the order-of-magnitude mismatch.","evidence_source_ids":["N09","N10","N11"],"verdict":"WRONG","same_as_reference":true,"eligible_for_deadline_metrics":true,"deadline_year":2019,"outcome_assessment_complete":true}},"timing":{"first_full_achievement_year":null,"documented_by_year":null,"achievement_interval_years":null,"kind":"UNESTABLISHED","label":"Not fully realized by 2026-10-15; overdue, not an observed delay","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":[{"from_version":"0.2","previous_assessment":{"review_state":"ASSESSED","verdict":"UNSCORABLE","as_of":"2026-10-14","at_deadline":"NOT_ASSESSED","at_review":"NOT_ASSESSED","confidence":"MEDIUM","disputed_interpretation":false,"benchmark_ready":false,"rationale":"The article specifies $1,000 in 1999 dollars. The intake’s alleged $4,000 book version is not confirmed here and cannot be silently overwritten. The brain-equivalence metric also remains undefined.","first_full_achievement_year":null,"full_achievement_documented_by_year":null,"drift_years":null,"drift_range_years":null,"still_unfulfilled_as_of_review":false,"eligible_for_deadline_metrics":false,"withheld_reason":"Future, unresolved, intrinsically unscorable, or interpretation-dependent; see assessment."},"change_reason":"Complete deadline/current outcome assessment under explicit scope; separate uncertainty from verdict."}],"is_new_in_v03":false,"counting_policy":{"unit":"distinct_forecast_version","identical_repetitions_count_once":true,"alternative_interpretations_count_once":true,"legacy_conditional_rows_count":false},"id":"K034","slug":"k034-the-2019-thousand-dollar-brain-compute-estimate","title":"The 2019 thousand-dollar brain-compute estimate","prediction_paraphrase":"$1,000 in 1999 dollars buys brain-scale computation.","wording_kind":"EDITORIAL_PARAPHRASE","original_quote":null,"primary_category":"Computing hardware","tags":["brain equivalence","currency basis"],"subcategory":"Price/performance","claim_type":"PERFORMANCE_COST","family_id":"F-BRAIN-COMPUTE","parent_statement_id":null,"legacy_ids":[],"origin":"CONFIRMED_RELATED_VERSION","publication":{"forecast_year":1999,"year_basis":"ARTICLE_ATTRIBUTION","original_book_edition_checked":false,"forecast_year_status":"ATTRIBUTED_ARTICLE_YEAR_NOT_ORIGINAL_BOOK_COLLATION"},"target":{"text":"2019","deadline_year":2019,"precision":"YEAR","year_start":2019,"year_end":2019,"interpretation":null},"source_verification":{"state":"TEXT_CONFIRMED_IN_LISTED_SOURCE","confidence":"MEDIUM","book_text_not_independently_collated":true},"success_criteria":"A circa-$1,000 consumer budget, inflation-adjusted from 1999, buys approximately 20 petaflops of ordinary arithmetic capacity.","assessment":{"at_deadline_rationale":"Consumer hardware in 2019 was orders of magnitude below 20 petaflops in conventional FP32 throughput.","at_review_rationale":"Current checked consumer GPU specifications remain far short on that basis. Inflation adjustment and GPU-only accounting do not rescue the order-of-magnitude mismatch.","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":"WRONG","as_of":"2026-10-15","at_deadline":"NOT_FULFILLED","at_review":"NOT_FULFILLED","confidence":"HIGH","disputed_interpretation":false,"benchmark_ready":true,"rationale":"Consumer hardware in 2019 was orders of magnitude below 20 petaflops in conventional FP32 throughput. Current checked consumer GPU specifications remain far short on that basis. Inflation adjustment and GPU-only accounting do not rescue the order-of-magnitude mismatch.","first_full_achievement_year":null,"full_achievement_documented_by_year":null,"drift_years":null,"drift_range_years":null,"still_unfulfilled_as_of_review":true,"eligible_for_deadline_metrics":true,"withheld_reason":null},"prediction_sources":[{"source_id":"P02","locator":"Computers and communications in 2019: opening paragraph","role":"FORECAST_ATTRIBUTION"}],"evidence":[{"source_id":"N09","locator":"RTX 3080 launch price $699 and roughly 30 FP32 shader TFLOPS. Tensor, ray-tracing and shader operations are not interchangeable.","role":"OUTCOME_EVIDENCE"},{"source_id":"N10","locator":"10,752 CUDA cores, 2.62 GHz boost; theoretical FP32 about 56.3 TFLOPS using 2 operations/core/cycle. AI TOPS use different precision and cannot establish brain equivalence.","role":"OUTCOME_EVIDENCE"},{"source_id":"N11","locator":"RTX 5080 launch MSRP $999, 1,801 AI TOPS. Launch prices are not represented as October 2026 retail quotes; a GPU also is not a whole PC.","role":"OUTCOME_EVIDENCE"}],"limitations":"A different precision or sparse-operation convention changes the number, not automatically the biological claim. Current consumer comparison is illustrative, not a full auction-price optimization.","open_questions":["Original-edition collation remains an attribution task; it does not erase the stated outcome assessment."],"legacy_relations":[{"legacy_id":"EXE-012","relation":"SPLIT_AND_HOLD_PRICE_VARIANT"}]}