{
  "schema": "vedokrok.public-item.v1",
  "release_id": "MHC-RPUB-20260920-75ad787a",
  "url": "/knowledge/use-ai-to-accelerate-public-content-only-after-the-evidence-exists",
  "id": "MHC-D-RESEARCH-1156",
  "version": "0.1.0",
  "title": "Use AI to accelerate public content only after the evidence exists",
  "summary": "Generation speed is not an expertise multiplier when the evidence layer stays empty.",
  "kind": "principle",
  "body": "Use AI for transformation, editing, structure or exploration after the substantive value is anchored in real experience, inspected sources, data or a concrete artifact. Google's guidance favors original value and demonstrable expertise and warns about extensive automation used mainly to attract search traffic. When AI substantially shapes the content and readers would reasonably care, make the production process appropriately transparent.",
  "limits": [
    "AI-assisted writing is not inherently low quality, and disclosure practices depend on context. The source guidance does not say that every minor AI edit requires a label."
  ],
  "topics": [
    "union-public-expertise-and-network-compounding"
  ],
  "intents": [],
  "source_ids": [
    "RS-445CD77F9E00564A"
  ],
  "evidence": [
    {
      "claim": "Google's current people-first guidance asks creators to provide original information or analysis, substantial added value, clear sourcing, demonstrable expertise and content with a primary audience purpose rather than producing many topics mainly for search traffic.",
      "source_id": "RS-445CD77F9E00564A",
      "role": "supports",
      "note": "These are self-assessment principles and search guidance, not a deterministic ranking formula or evidence that any individual page will receive traffic.",
      "locator": "Content and quality questions, expertise questions and people-first content sections"
    },
    {
      "claim": "Google's guidance says that when automation or AI substantially contributes to content, explaining how and why it was used can help readers understand its useful role when that information would reasonably be expected.",
      "source_id": "RS-445CD77F9E00564A",
      "role": "contextualizes",
      "note": "The guidance does not require identical disclosure for every minor AI-assisted edit and does not promise a ranking benefit from disclosure.",
      "locator": "Who, How and Why sections"
    }
  ],
  "use_when": [
    "AI makes it possible to publish much faster than you can personally verify, experience or improve the underlying material."
  ],
  "avoid_when": [
    "AI-assisted writing is not inherently low quality, and disclosure practices depend on context. The source guidance does not say that every minor AI edit requires a label."
  ],
  "example": "AI helps turn a real technical investigation into a concise article, but the logs, decision logic, sanitized examples and factual checks come from the actual work.",
  "check": "Removing the generated prose would still leave a real underlying contribution: evidence, experience, analysis, data, method or artifact.",
  "sources": [
    {
      "id": "RS-445CD77F9E00564A",
      "title": "Creating helpful, reliable, people-first content",
      "url": "https://developers.google.com/search/docs/fundamentals/creating-helpful-content"
    }
  ],
  "relations": [
    {
      "from": "MHC-D-RESEARCH-1156",
      "to": "MHC-D-RESEARCH-0375",
      "type": "useful_with",
      "url": "/knowledge/write-acceptance-criteria-before-asking-ai-to-generate"
    }
  ],
  "collections": [
    {
      "id": "RC-6D439AFF8D5318B7",
      "title": "Turn real expertise and professional relationships into durable public career capital",
      "url": "/collections/turn-real-expertise-and-professional-relationships-into-durable-public-career-capital"
    }
  ]
}