{
  "question_id": "W300",
  "slug": "ai-speeds-lab-research-2x",
  "title": "A lab says AI has made its research 2× faster",
  "type": "binary",
  "resolution": {
    "levels": [],
    "values": [],
    "annotations": []
  },
  "forecasts": [
    {
      "question_id": "W300",
      "forecast_date": "2026-10-03",
      "evidence_cutoff": "2026-10-03",
      "method": "wide-v2.1",
      "opened": "2026-10-03",
      "linked_questions": [],
      "reading_adopted": "A historical-baseline claim (e.g., Anthropic stating a sustained AI-attributable >=2x acceleration backed by its public AECI trend-break analysis plus an explicit compute/headcount attribution) counts; claims resting on redacted indicators, third-party (e.g., METR) estimates merely reproduced in a developer document, proxies (LoC, agent-workdays, automation share), or forecasts do not.",
      "summary": null,
      "ambiguities": [],
      "key_drivers": [],
      "would_update_on": [],
      "forecasts": [
        {
          "period_end": "2026-12-31",
          "p_yes": 0.01
        },
        {
          "period_end": "2027-03-31",
          "p_yes": 0.04
        },
        {
          "period_end": "2027-06-30",
          "p_yes": 0.073
        },
        {
          "period_end": "2027-09-30",
          "p_yes": 0.115
        },
        {
          "period_end": "2027-12-31",
          "p_yes": 0.155
        },
        {
          "period_end": "2028-03-31",
          "p_yes": 0.197
        },
        {
          "period_end": "2028-06-30",
          "p_yes": 0.237
        },
        {
          "period_end": "2028-09-30",
          "p_yes": 0.275
        },
        {
          "period_end": "2028-12-31",
          "p_yes": 0.308
        },
        {
          "period_end": "2029-03-31",
          "p_yes": 0.332
        },
        {
          "period_end": "2029-06-30",
          "p_yes": 0.356
        },
        {
          "period_end": "2029-09-30",
          "p_yes": 0.376
        },
        {
          "period_end": "2029-12-31",
          "p_yes": 0.396
        },
        {
          "period_end": "2030-03-31",
          "p_yes": 0.411
        },
        {
          "period_end": "2030-06-30",
          "p_yes": 0.425
        },
        {
          "period_end": "2030-09-30",
          "p_yes": 0.437
        },
        {
          "period_end": "2030-12-31",
          "p_yes": 0.45
        },
        {
          "period_end": "2031-03-31",
          "p_yes": 0.46
        },
        {
          "period_end": "2031-06-30",
          "p_yes": 0.469
        },
        {
          "period_end": "2031-09-30",
          "p_yes": 0.479
        },
        {
          "period_end": "2031-12-31",
          "p_yes": 0.488
        }
      ],
      "passes": []
    },
    {
      "question_id": "W300",
      "forecast_date": "2026-10-10",
      "evidence_cutoff": "2026-10-10",
      "method": "wide-v2.1",
      "opened": "2026-10-03",
      "linked_questions": [],
      "reading_adopted": null,
      "summary": "Anthropic, the lab closest to this bar, said in August and September 2026 that artificial intelligence speeds its research significantly but not yet twofold; an outside evaluator put it near one and a half times. Crossing twofold within one to two years is plausible, but the question also requires a published number with a clear method, which labs often withhold. A lab report affirming a sustained twofold speedup, with its method, would most change this forecast.",
      "ambiguities": [],
      "key_drivers": [],
      "would_update_on": [],
      "forecasts": [
        {
          "period_end": "2026-12-31",
          "p_yes": 0.03
        },
        {
          "period_end": "2027-03-31",
          "p_yes": 0.074
        },
        {
          "period_end": "2027-06-30",
          "p_yes": 0.125
        },
        {
          "period_end": "2027-09-30",
          "p_yes": 0.182
        },
        {
          "period_end": "2027-12-31",
          "p_yes": 0.239
        },
        {
          "period_end": "2028-03-31",
          "p_yes": 0.292
        },
        {
          "period_end": "2028-06-30",
          "p_yes": 0.342
        },
        {
          "period_end": "2028-09-30",
          "p_yes": 0.386
        },
        {
          "period_end": "2028-12-31",
          "p_yes": 0.426
        },
        {
          "period_end": "2029-03-31",
          "p_yes": 0.458
        },
        {
          "period_end": "2029-06-30",
          "p_yes": 0.485
        },
        {
          "period_end": "2029-09-30",
          "p_yes": 0.508
        },
        {
          "period_end": "2029-12-31",
          "p_yes": 0.528
        },
        {
          "period_end": "2030-03-31",
          "p_yes": 0.544
        },
        {
          "period_end": "2030-06-30",
          "p_yes": 0.559
        },
        {
          "period_end": "2030-09-30",
          "p_yes": 0.572
        },
        {
          "period_end": "2030-12-31",
          "p_yes": 0.584
        },
        {
          "period_end": "2031-03-31",
          "p_yes": 0.595
        },
        {
          "period_end": "2031-06-30",
          "p_yes": 0.604
        },
        {
          "period_end": "2031-09-30",
          "p_yes": 0.612
        },
        {
          "period_end": "2031-12-31",
          "p_yes": 0.62
        }
      ],
      "passes": []
    },
    {
      "question_id": "W300",
      "forecast_date": "2026-10-17",
      "evidence_cutoff": "2026-10-17",
      "method": "wide-v2.1",
      "opened": "2026-10-03",
      "linked_questions": [],
      "reading_adopted": null,
      "summary": "Anthropic already measures this and says artificial intelligence has sped up its research meaningfully but has not yet doubled it. An outside evaluator puts the speed-up near one and a half times. Company leaders expect much faster gains within a year or two, but the required proof of method is demanding. The development that would most change this forecast is Anthropic's next risk report declaring a sustained doubling caused by its models.",
      "ambiguities": [],
      "key_drivers": [],
      "would_update_on": [],
      "forecasts": [
        {
          "period_end": "2026-12-31",
          "p_yes": 0.035
        },
        {
          "period_end": "2027-03-31",
          "p_yes": 0.088
        },
        {
          "period_end": "2027-06-30",
          "p_yes": 0.147
        },
        {
          "period_end": "2027-09-30",
          "p_yes": 0.207
        },
        {
          "period_end": "2027-12-31",
          "p_yes": 0.263
        },
        {
          "period_end": "2028-03-31",
          "p_yes": 0.31
        },
        {
          "period_end": "2028-06-30",
          "p_yes": 0.352
        },
        {
          "period_end": "2028-09-30",
          "p_yes": 0.387
        },
        {
          "period_end": "2028-12-31",
          "p_yes": 0.418
        },
        {
          "period_end": "2029-03-31",
          "p_yes": 0.444
        },
        {
          "period_end": "2029-06-30",
          "p_yes": 0.467
        },
        {
          "period_end": "2029-09-30",
          "p_yes": 0.488
        },
        {
          "period_end": "2029-12-31",
          "p_yes": 0.506
        },
        {
          "period_end": "2030-03-31",
          "p_yes": 0.521
        },
        {
          "period_end": "2030-06-30",
          "p_yes": 0.535
        },
        {
          "period_end": "2030-09-30",
          "p_yes": 0.549
        },
        {
          "period_end": "2030-12-31",
          "p_yes": 0.562
        },
        {
          "period_end": "2031-03-31",
          "p_yes": 0.573
        },
        {
          "period_end": "2031-06-30",
          "p_yes": 0.584
        },
        {
          "period_end": "2031-09-30",
          "p_yes": 0.594
        },
        {
          "period_end": "2031-12-31",
          "p_yes": 0.605
        }
      ],
      "passes": []
    }
  ],
  "declined": []
}