AI Compute Scaling Metrics

{
  "id": "ai-compute-scaling-metrics",
  "numericId": "E907",
  "path": "/knowledge-base/models/ai-compute-scaling-metrics/",
  "filePath": "knowledge-base/models/ai-compute-scaling-metrics.mdx",
  "title": "AI Compute Scaling Metrics",
  "quality": 78,
  "readerImportance": 82,
  "researchImportance": null,
  "tacticalValue": 92,
  "contentFormat": "article",
  "tractability": null,
  "neglectedness": null,
  "uncertainty": null,
  "causalLevel": null,
  "lastUpdated": "2026-03-13",
  "dateCreated": "2026-03-07",
  "llmSummary": "AI training compute is growing at ~4-5× per year with algorithmic efficiency improving ~3× per year (halving effective compute cost every ~8 months), while the compute landscape is shifting toward inference-dominant workloads (~50% in 2025, projected 67% in 2026). Big-4 hyperscaler capex is projected to approach \\$700B combined in 2026, with the US holding ~75% of global GPU cluster performance. The page tracks these empirical metrics alongside AI coding tool adoption as an acceleration indicator and Anthropic's scaling challenges as a case study, while documenting criticisms including capability-compute decoupling, benchmark limitations, and diminishing pre-training returns. Key safety implications include timeline compression from compound compute-plus-efficiency gains and a narrowing governance window as compute concentration enables tractable monitoring interventions.",
  "description": "Empirical tracking of measurable indicators of AI compute scaling: training compute growth, GPU deployment, capital expenditure, inference shifts, and safety implications.",
  "ratings": {
    "novelty": 6,
    "rigor": 7,
    "actionability": 7,
    "completeness": 8
  },
  "category": "models",
  "subcategory": null,
  "clusters": [
    "ai-safety"
  ],
  "metrics": {
    "wordCount": 3525,
    "tableCount": 3,
    "diagramCount": 0,
    "internalLinks": 31,
    "externalLinks": 5,
    "footnoteCount": 0,
    "bulletRatio": 0.12,
    "sectionCount": 25,
    "hasOverview": true,
    "structuralScore": 13
  },
  "suggestedQuality": 87,
  "updateFrequency": null,
  "evergreen": true,
  "wordCount": 3525,
  "unconvertedLinks": [
    {
      "text": "ourworldindata.org – Scaling Up AI",
      "url": "https://ourworldindata.org/scaling-up-ai",
      "resourceId": "cb1cd9e4d736df7f",
      "resourceTitle": "Our World in Data"
    },
    {
      "text": "epoch.ai/trends",
      "url": "https://epoch.ai/trends",
      "resourceId": "b029bfc231e620cc",
      "resourceTitle": "Epoch AI"
    },
    {
      "text": "hai.stanford.edu",
      "url": "https://hai.stanford.edu/ai-index/2025-ai-index-report",
      "resourceId": "da87f2b213eb9272",
      "resourceTitle": "Stanford AI Index 2025"
    }
  ],
  "unconvertedLinkCount": 3,
  "convertedLinkCount": 0,
  "backlinkCount": 0,
  "hallucinationRisk": {
    "level": "medium",
    "score": 40,
    "factors": [
      "no-citations",
      "high-rigor"
    ]
  },
  "entityType": "analysis",
  "redundancy": {
    "maxSimilarity": 20,
    "similarPages": [
      {
        "id": "ai-timelines",
        "title": "AI Timelines",
        "path": "/knowledge-base/models/ai-timelines/",
        "similarity": 20
      },
      {
        "id": "projecting-compute-spending",
        "title": "Projecting Compute Spending",
        "path": "/knowledge-base/models/projecting-compute-spending/",
        "similarity": 20
      },
      {
        "id": "large-language-models",
        "title": "Large Language Models",
        "path": "/knowledge-base/capabilities/large-language-models/",
        "similarity": 18
      },
      {
        "id": "language-models",
        "title": "Large Language Models",
        "path": "/knowledge-base/capabilities/language-models/",
        "similarity": 17
      },
      {
        "id": "self-improvement",
        "title": "Self-Improvement and Recursive Enhancement",
        "path": "/knowledge-base/capabilities/self-improvement/",
        "similarity": 17
      }
    ]
  },
  "changeHistory": [
    {
      "date": "2026-02-27",
      "branch": "claude/ai-compute-scaling-metrics",
      "title": "Orchestrator v2 (standard): AI Compute Scaling Metrics",
      "summary": "Improved \"AI Compute Scaling Metrics\" via orchestrator v2 (standard, 24 tool calls, 0 refinement cycles). Quality gate: passed. Cost: $5.48 actual.",
      "duration": "407.8s",
      "cost": "$5.48"
    }
  ],
  "coverage": {
    "passing": 5,
    "total": 13,
    "targets": {
      "tables": 14,
      "diagrams": 1,
      "internalLinks": 28,
      "externalLinks": 18,
      "footnotes": 11,
      "references": 11
    },
    "actuals": {
      "tables": 3,
      "diagrams": 0,
      "internalLinks": 31,
      "externalLinks": 5,
      "footnotes": 0,
      "references": 3,
      "quotesWithQuotes": 0,
      "quotesTotal": 0,
      "accuracyChecked": 0,
      "accuracyTotal": 0
    },
    "items": {
      "llmSummary": "green",
      "schedule": "red",
      "entity": "green",
      "editHistory": "green",
      "overview": "green",
      "tables": "amber",
      "diagrams": "red",
      "internalLinks": "green",
      "externalLinks": "amber",
      "footnotes": "red",
      "references": "amber",
      "quotes": "red",
      "accuracy": "red"
    },
    "editHistoryCount": 1,
    "ratingsString": "N:6 R:7 A:7 C:8"
  },
  "readerRank": 72,
  "recommendedScore": 218.86
}

No external links

No backlinks