AI Value Learning

{
  "id": "value-learning",
  "numericId": null,
  "path": "/knowledge-base/responses/value-learning/",
  "filePath": "knowledge-base/responses/value-learning.mdx",
  "title": "AI Value Learning",
  "quality": null,
  "importance": null,
  "contentFormat": "article",
  "tractability": null,
  "neglectedness": null,
  "uncertainty": null,
  "causalLevel": null,
  "lastUpdated": "2026-02-09",
  "llmSummary": null,
  "structuredSummary": null,
  "description": "Training AI systems to infer and adopt human values from observation and interaction",
  "ratings": null,
  "category": "responses",
  "subcategory": null,
  "clusters": [
    "ai-safety"
  ],
  "metrics": {
    "wordCount": 7,
    "tableCount": 0,
    "diagramCount": 0,
    "internalLinks": 0,
    "externalLinks": 0,
    "footnoteCount": 0,
    "bulletRatio": 0,
    "sectionCount": 0,
    "hasOverview": false,
    "structuralScore": 2
  },
  "suggestedQuality": 13,
  "updateFrequency": null,
  "evergreen": true,
  "wordCount": 7,
  "unconvertedLinks": [],
  "unconvertedLinkCount": 0,
  "convertedLinkCount": 0,
  "backlinkCount": 1,
  "redundancy": {
    "maxSimilarity": 0,
    "similarPages": []
  }
}

{
  "id": "value-learning",
  "type": "safety-agenda",
  "title": "AI Value Learning",
  "description": "Research agenda focused on AI systems learning human values from data, behavior, or feedback rather than explicit specification.",
  "tags": [
    "alignment",
    "values",
    "learning"
  ],
  "relatedEntries": [
    {
      "id": "rlhf",
      "type": "approach"
    },
    {
      "id": "reward-hacking",
      "type": "risk"
    }
  ],
  "sources": [],
  "lastUpdated": "2025-12",
  "customFields": []
}

No facts for this entity

{
  "lesswrong": "https://www.lesswrong.com/tag/value-learning",
  "stampy": "https://aisafety.info/questions/8IzO/What-is-value-learning",
  "alignmentForum": "https://www.alignmentforum.org/tag/value-learning"
}

{
  "title": "AI Value Learning",
  "description": "Training AI systems to infer and adopt human values from observation and interaction",
  "sidebar": {
    "order": 50
  },
  "quality": 0,
  "importance": 0,
  "lastEdited": "2026-02-09",
  "entityType": "approach"
}

---
title: "AI Value Learning"
description: "Training AI systems to infer and adopt human values from observation and interaction"
sidebar:
  order: 50
quality: 0
importance: 0
lastEdited: "2026-02-09"
entityType: approach
---

This page is a stub. Content needed.