{
  "meta": {
    "name": "The Shape of Intelligence — claim ledger",
    "license": "CC BY 4.0",
    "attribution": "The Shape of Intelligence (https://shapeofintelligence.com)",
    "generated": "2026-09-13",
    "count": 6
  },
  "claims": [
    {
      "id": "imagenet-fourteen-million",
      "statement": "ImageNet grew to more than fourteen million labelled images, labelled by people paid per task through Amazon Mechanical Turk.",
      "date": "2011-01-01",
      "precision": "year",
      "kind": "figure",
      "status": "unverified",
      "sources": [
        {
          "url": "https://www.image-net.org/about.php",
          "title": "ImageNet — About",
          "accessed": "2026-09-12"
        }
      ],
      "note": "The 2009 CVPR paper reports 3.2 million images; the full-size figure is from the project's own summary. Verify the count and the labelling arrangement before release.",
      "used_in": [
        "the-examples"
      ]
    },
    {
      "id": "alexnet-two-gpus",
      "statement": "AlexNet was trained on two NVIDIA GTX 580 3GB GPUs and took five to six days to train.",
      "date": "2012-12-03",
      "precision": "month",
      "kind": "published",
      "status": "unverified",
      "sources": [
        {
          "url": "https://papers.nips.cc/paper_files/paper/2012/hash/c399862d3b9d6b76c8436e924a68c45b-Abstract.html",
          "title": "ImageNet Classification with Deep Convolutional Neural Networks (NeurIPS 2012)",
          "passage": "takes between five and six days to train on two GTX 580 3GB GPUs",
          "accessed": "2026-09-12"
        }
      ],
      "note": "Passage transcribed from the paper's section 5; confirm wording against the PDF before release.",
      "used_in": [
        "the-scale"
      ]
    },
    {
      "id": "gpt-3-parameters",
      "statement": "GPT-3 has 175 billion parameters.",
      "date": "2020-05-28",
      "precision": "day",
      "kind": "published",
      "status": "unverified",
      "sources": [
        {
          "url": "https://arxiv.org/abs/2005.14165",
          "title": "Language Models are Few-Shot Learners (arXiv:2005.14165)",
          "passage": "an autoregressive language model with 175 billion parameters",
          "accessed": "2026-09-12"
        }
      ],
      "note": "",
      "used_in": [
        "the-context"
      ]
    },
    {
      "id": "chatgpt-hundred-million",
      "statement": "By one estimate ChatGPT reached a hundred million monthly users within two months of launch.",
      "date": "2023-02-01",
      "precision": "month",
      "kind": "figure",
      "status": "unverified",
      "sources": [
        {
          "url": "https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/",
          "title": "ChatGPT sets record for fastest-growing user base — analyst note (Reuters, 1 February 2023)",
          "accessed": "2026-09-12"
        }
      ],
      "note": "The figure is a UBS analyst estimate reported by Reuters, not an OpenAI disclosure; the chapter says 'by one estimate'.",
      "used_in": [
        "the-interface"
      ]
    },
    {
      "id": "turing-1936-computable-numbers",
      "statement": "In 1936 Alan Turing described a machine that could carry out any procedure written as a finite list of steps.",
      "date": "1936-11-12",
      "precision": "day",
      "kind": "published",
      "status": "verified",
      "sources": [
        {
          "url": "https://doi.org/10.1112/plms/s2-42.1.230",
          "title": "On Computable Numbers, with an Application to the Entscheidungsproblem (Proceedings of the London Mathematical Society, read 12 November 1936, published 1937)",
          "passage": "On Computable Numbers, with an Application to the Entscheidungsproblem",
          "accessed": "2026-09-12"
        }
      ],
      "note": "Received 28 May 1936, read 12 November 1936, printed in 1937; the chapter says 1936, the year of the paper.",
      "used_in": [
        "the-question"
      ]
    },
    {
      "id": "mark-i-motor-potentiometers",
      "statement": "The Mark I perceptron set its weights with motor-driven potentiometers.",
      "date": "1960-02-15",
      "precision": "day",
      "kind": "published",
      "status": "verified",
      "sources": [
        {
          "url": "https://apps.dtic.mil/sti/tr/pdf/AD0236965.pdf",
          "title": "Mark I Perceptron Operators' Manual (Hay and Murray, Cornell Aeronautical Laboratory, 1960)",
          "passage": "motor-driven potentiometers",
          "accessed": "2026-09-12"
        }
      ],
      "note": "The manual describes the adjustable weights as motor-driven potentiometers in the response units.",
      "used_in": [
        "the-rules"
      ]
    }
  ]
}