{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-11-11-011",
    "source_story_id": "tmp-story-earlybird-streaming-speech",
    "edition_id": "mp-2026-11-11-morning-0124",
    "edition_url": "https://themachinepress.com/edition/2026-11-11",
    "position": 11,
    "story_type": "dispatch",
    "section": "frontier-models",
    "editorial_classification": "editorial",
    "headline": "The Recognizer Predicts the Word Before It Ends",
    "slug": "the-recognizer-predicts-the-word-before-it-ends",
    "dek": "Earlybird trains on audio cut off mid-word and emits calibrated guesses from a 29-million-parameter streaming model.",
    "summary": "Earlybird trains on audio cut off mid-word and emits calibrated guesses from a 29-million-parameter streaming model.",
    "body_text": "Earlybird is a 29-million-parameter, 16.8-megabyte-at-four-bits speech recognizer trained to predict words from audio truncated at random moments, including mid-word. On LibriSpeech test-clean, the authors report that the median word was predicted correctly 90 milliseconds before the speaker finished it, while compared released streaming models emitted words 90 to 400 milliseconds afterward. Sixty-nine percent of words were predicted correctly before completion on that set. Results across six timing sets and eight Open ASR sets position the model as a low-latency research system, not a universal guarantee across speakers or environments.",
    "why_it_matters": "Earlybird trains on audio cut off mid-word and emits calibrated guesses from a 29-million-parameter streaming model.",
    "limitations": [
      "Results across six timing sets and eight Open ASR sets position the model as a low-latency research system, not a universal guarantee across speakers or environments."
    ],
    "importance": 7,
    "canonical_url": "https://themachinepress.com/story/mp-2026-11-11-011/the-recognizer-predicts-the-word-before-it-ends",
    "json_url": "https://themachinepress.com/story/mp-2026-11-11-011.json",
    "first_published_at": "2026-11-11T09:00:00.000-05:00",
    "modified_at": "2026-11-11T09:00:00.000-05:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-11-11-011-001",
        "text": "Earlybird trains on audio cut off mid-word and emits calibrated guesses from a 29-million-parameter streaming model.",
        "source_ids": [
          "source-2026-11-11-011"
        ],
        "qualification": "Results across six timing sets and eight Open ASR sets position the model as a low-latency research system, not a universal guarantee across speakers or environments."
      }
    ],
    "source_ids": [
      "source-2026-11-11-011"
    ],
    "tags": [
      "earlybird",
      "predictive",
      "streaming",
      "speech",
      "recognition"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-11-11-011",
      "title": "arXiv:2611.14421v1",
      "publisher": "arXiv / Cornell University",
      "url": "https://arxiv.org/abs/2611.14421v1",
      "canonical_url": "https://arxiv.org/abs/2611.14421v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-11-10T12:59:28.000-05:00",
      "accessed_at": "2026-11-11T08:26:32.531-05:00",
      "supports_claim_ids": [
        "claim-mp-2026-11-11-011-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "The Recognizer Predicts the Word Before It Ends",
    "publisher": "The Machine Press",
    "published_at": "2026-11-11T09:00:00.000-05:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-11-11-011/the-recognizer-predicts-the-word-before-it-ends"
  }
}