{
  "id": 1878357,
  "title": "Using synthetic data for AI training is 'a big mistake,' says AI pioneer Rich Sutton",
  "url": "https://urgent.news/2026/08/19/using-synthetic-data-for-ai-training-is-a-big-mistake-says-ai-pioneer",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-19T05:49:46.000Z",
  "source": {
    "name": "Business Insider",
    "slug": "business-insider",
    "url": "https://www.businessinsider.com/synthetic-data-for-ai-training-is-big-mistake-rich-sutton-2026-8"
  },
  "original_language": "en",
  "account": "AI pioneer Richard Sutton has cautioned against the use of synthetic data for training artificial intelligence (AI) systems. Tech giants such as Google and OpenAI are actively seeking real-world data to train their AI models, according to Sutton. Sutton's Oak Lab is focused on teaching AI agents through experience, rather than using synthetic datasets. Sutton, a Turing Award winner and contributor to the current AI boom, argues that using synthetic data for AI training is a significant mistake. During an episode of Sequoia's podcast, Sutton stated that the reliance on synthetic data for scaling AI is misguided. Synthetic data refers to information generated artificially by algorithms or AI models, as opposed to being gathered from authentic, real-world sources. Sutton noted that human behavior, for example, cannot be adequately represented through synthetic data. Instead, the AI pioneer advocates for real experiential data, which AI agents can acquire by interacting with their actual environment and learning from the outcomes. He contends that simulations of the physical world, including variables like friction and wear in robot motors, cannot be replicated through synthetic data. Sutton and his former student Khurram Javed recently founded Oak Lab to develop AI agents that learn continuously from their own experience, rather than solely relying on large, pre-curated datasets. The startup has not disclosed any funding or investors.",
  "summary": "He said that the AI industry is headed in the wrong direction because of its reliance on synthetic training data.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}