{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-01-003",
    "source_story_id": "tmp-story-ego4wam-human-data-scaling",
    "edition_id": "mp-2026-10-01-morning-0084",
    "edition_url": "https://themachinepress.com/edition/2026-10-01",
    "position": 3,
    "story_type": "dispatch",
    "section": "robotics",
    "editorial_classification": "editorial",
    "headline": "Alignment Beat Raw Hours in Human-to-Robot Data",
    "slug": "alignment-beat-raw-hours-in-human-to-robot-data",
    "dek": "Ego4WAM separated duration, diversity, action labels and human-robot alignment under one world-action model.",
    "summary": "Ego4WAM separated duration, diversity, action labels and human-robot alignment under one world-action model.",
    "body_text": "Ego4WAM holds the model backbone fixed while varying the properties and uses of egocentric human video. In real-robot and RoboDojo tests, the authors report that aligned human demonstrations improved out-of-distribution generalization and reduced the target-task robot data needed. Duration and task diversity contributed differently, while video-only supervision remained useful without action labels. The study argues against treating hours alone as the scaling variable; its conclusions remain bounded by the tested data and policy pipeline.",
    "why_it_matters": "Ego4WAM separated duration, diversity, action labels and human-robot alignment under one world-action model.",
    "limitations": [
      "Duration and task diversity contributed differently, while video-only supervision remained useful without action labels.",
      "The study argues against treating hours alone as the scaling variable; its conclusions remain bounded by the tested data and policy pipeline."
    ],
    "importance": 9,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-01-003/alignment-beat-raw-hours-in-human-to-robot-data",
    "json_url": "https://themachinepress.com/story/mp-2026-10-01-003.json",
    "first_published_at": "2026-10-01T09:00:00.000-04:00",
    "modified_at": "2026-10-01T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-01-003-001",
        "text": "Ego4WAM separated duration, diversity, action labels and human-robot alignment under one world-action model.",
        "source_ids": [
          "source-2026-10-01-003"
        ],
        "qualification": "Duration and task diversity contributed differently, while video-only supervision remained useful without action labels."
      }
    ],
    "source_ids": [
      "source-2026-10-01-003"
    ],
    "tags": [
      "egocentric video",
      "robot data",
      "world-action models"
    ],
    "image_url": "https://themachinepress.com/issues/2026-10-01/ego4wam-robot-arm.webp",
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-10-01-003",
      "title": "arXiv preprint 2609.40341",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2609.40341",
      "canonical_url": "https://arxiv.org/abs/2609.40341",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-09-30T13:58:35.000-04:00",
      "accessed_at": "2026-10-01T08:18:00.000-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-01-003-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "Alignment Beat Raw Hours in Human-to-Robot Data",
    "publisher": "The Machine Press",
    "published_at": "2026-10-01T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-01-003/alignment-beat-raw-hours-in-human-to-robot-data"
  }
}