Epoch AI

{
  "id": "epoch-ai",
  "numericId": "E125",
  "path": "/knowledge-base/organizations/epoch-ai/",
  "filePath": "knowledge-base/organizations/epoch-ai.mdx",
  "title": "Epoch AI",
  "quality": 51,
  "readerImportance": 87.5,
  "researchImportance": 7.5,
  "tacticalValue": 70,
  "contentFormat": "article",
  "tractability": null,
  "neglectedness": null,
  "uncertainty": null,
  "causalLevel": null,
  "lastUpdated": "2026-03-13",
  "dateCreated": null,
  "llmSummary": "Epoch AI maintains comprehensive databases tracking 3,200+ ML models showing 4.4x annual compute growth and projects data exhaustion 2026-2032. Their empirical work directly informed EU AI Act's 10^25 FLOP threshold and US EO 14110, with their Epoch Capabilities Index showing ~90% acceleration in AI progress since April 2024.",
  "description": "Epoch AI is a research institute tracking AI development trends through comprehensive databases on training compute, model parameters, and hardware capabilities. Their data shows training compute growing 4.4x annually since 2010, with over 30 models now exceeding 10^25 FLOP. Their work directly informs major AI policy including the EU AI Act's 10^25 FLOP threshold and US Executive Order 14110's compute requirements. In 2025, they launched the Epoch Capabilities Index showing ~90% acceleration in AI progress since April 2024, and the FrontierMath benchmark where frontier models solve less than 2% of problems (o3 achieved ~10-25%).",
  "ratings": {
    "novelty": 2.5,
    "rigor": 6,
    "actionability": 3.5,
    "completeness": 7.5
  },
  "category": "organizations",
  "subcategory": "epistemic-orgs",
  "clusters": [
    "community",
    "epistemics",
    "ai-safety"
  ],
  "metrics": {
    "wordCount": 4593,
    "tableCount": 31,
    "diagramCount": 1,
    "internalLinks": 14,
    "externalLinks": 67,
    "footnoteCount": 0,
    "bulletRatio": 0.11,
    "sectionCount": 64,
    "hasOverview": true,
    "structuralScore": 15
  },
  "suggestedQuality": 100,
  "updateFrequency": 45,
  "evergreen": true,
  "wordCount": 4593,
  "unconvertedLinks": [
    {
      "text": "arXiv:2202.05924",
      "url": "https://arxiv.org/abs/2202.05924",
      "resourceId": "a9007e0713dc6b7f",
      "resourceTitle": "Sevilla et al."
    },
    {
      "text": "epoch.ai",
      "url": "https://epoch.ai",
      "resourceId": "c660a684a423d4ac",
      "resourceTitle": "Epoch AI"
    },
    {
      "text": "epoch.ai",
      "url": "https://epoch.ai/",
      "resourceId": "c660a684a423d4ac",
      "resourceTitle": "Epoch AI"
    },
    {
      "text": "Epoch AI",
      "url": "https://epoch.ai/",
      "resourceId": "c660a684a423d4ac",
      "resourceTitle": "Epoch AI"
    },
    {
      "text": "FrontierMath Tier 4",
      "url": "https://epoch.ai/frontiermath/the-benchmark",
      "resourceId": "46010026d8feac35",
      "resourceTitle": "FrontierMath benchmark"
    },
    {
      "text": "Epoch Trends",
      "url": "https://epoch.ai/trends",
      "resourceId": "b029bfc231e620cc",
      "resourceTitle": "Epoch AI"
    },
    {
      "text": "arXiv:2202.05924",
      "url": "https://arxiv.org/abs/2202.05924",
      "resourceId": "a9007e0713dc6b7f",
      "resourceTitle": "Sevilla et al."
    },
    {
      "text": "METR's Time Horizon benchmark",
      "url": "https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/",
      "resourceId": "271fc5f73a8304b2",
      "resourceTitle": "Measuring AI Ability to Complete Long Tasks - METR"
    },
    {
      "text": "how many models will exceed compute thresholds",
      "url": "https://epoch.ai/blog/model-counts-compute-thresholds",
      "resourceId": "080da6a9f43ad376",
      "resourceTitle": "Epoch AI projections"
    },
    {
      "text": "arXiv:2202.05924",
      "url": "https://arxiv.org/abs/2202.05924",
      "resourceId": "a9007e0713dc6b7f",
      "resourceTitle": "Sevilla et al."
    },
    {
      "text": "Epoch Blog",
      "url": "https://epoch.ai/blog/can-ai-scaling-continue-through-2030",
      "resourceId": "9587b65b1192289d",
      "resourceTitle": "Epoch AI"
    },
    {
      "text": "Epoch Data Insights",
      "url": "https://epoch.ai/data-insights/ai-capabilities-progress-has-sped-up",
      "resourceId": "663417bdb09208a4",
      "resourceTitle": "Epoch AI's analysis"
    },
    {
      "text": "directly incorporates Epoch's compute trend data",
      "url": "https://ourworldindata.org/grapher/artificial-intelligence-training-computation",
      "resourceId": "87ae03cc6eaca6c6",
      "resourceTitle": "Our World in Data AI training"
    },
    {
      "text": "Epoch AI Website",
      "url": "https://epoch.ai/",
      "resourceId": "c660a684a423d4ac",
      "resourceTitle": "Epoch AI"
    },
    {
      "text": "ML Trends Dashboard",
      "url": "https://epoch.ai/trends",
      "resourceId": "b029bfc231e620cc",
      "resourceTitle": "Epoch AI"
    },
    {
      "text": "Compute Trends Across Three Eras (arXiv:2202.05924)",
      "url": "https://arxiv.org/abs/2202.05924",
      "resourceId": "a9007e0713dc6b7f",
      "resourceTitle": "Sevilla et al."
    }
  ],
  "unconvertedLinkCount": 16,
  "convertedLinkCount": 0,
  "backlinkCount": 28,
  "hallucinationRisk": {
    "level": "high",
    "score": 75,
    "factors": [
      "biographical-claims",
      "no-citations"
    ]
  },
  "entityType": "organization",
  "redundancy": {
    "maxSimilarity": 16,
    "similarPages": [
      {
        "id": "ai-timelines",
        "title": "AI Timelines",
        "path": "/knowledge-base/models/ai-timelines/",
        "similarity": 16
      },
      {
        "id": "metaculus",
        "title": "Metaculus",
        "path": "/knowledge-base/organizations/metaculus/",
        "similarity": 16
      },
      {
        "id": "rethink-priorities",
        "title": "Rethink Priorities",
        "path": "/knowledge-base/organizations/rethink-priorities/",
        "similarity": 16
      },
      {
        "id": "large-language-models",
        "title": "Large Language Models",
        "path": "/knowledge-base/capabilities/large-language-models/",
        "similarity": 15
      },
      {
        "id": "reasoning",
        "title": "Reasoning and Planning",
        "path": "/knowledge-base/capabilities/reasoning/",
        "similarity": 15
      }
    ]
  },
  "coverage": {
    "passing": 7,
    "total": 13,
    "targets": {
      "tables": 18,
      "diagrams": 2,
      "internalLinks": 37,
      "externalLinks": 23,
      "footnotes": 14,
      "references": 14
    },
    "actuals": {
      "tables": 31,
      "diagrams": 1,
      "internalLinks": 14,
      "externalLinks": 67,
      "footnotes": 0,
      "references": 32,
      "quotesWithQuotes": 0,
      "quotesTotal": 0,
      "accuracyChecked": 0,
      "accuracyTotal": 0
    },
    "items": {
      "llmSummary": "green",
      "schedule": "green",
      "entity": "green",
      "editHistory": "red",
      "overview": "green",
      "tables": "green",
      "diagrams": "amber",
      "internalLinks": "amber",
      "externalLinks": "green",
      "footnotes": "red",
      "references": "green",
      "quotes": "red",
      "accuracy": "red"
    },
    "ratingsString": "N:2.5 R:6 A:3.5 C:7.5"
  },
  "readerRank": 33,
  "researchRank": 570,
  "recommendedScore": 167.61
}

{
  "grokipedia": "https://grokipedia.com/page/Epoch_AI",
  "eaForum": "https://forum.effectivealtruism.org/topics/epoch-ai"
}