{
  "id": 2151258,
  "title": "Are We Thinking Correctly About AI Intelligence?",
  "url": "https://urgent.news/2026/08/20/are-we-thinking-correctly-about-ai-intelligence",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-20T14:04:40.000Z",
  "source": {
    "name": "Quanta Magazine",
    "slug": "quanta-magazine",
    "url": "https://www.quantamagazine.org/are-we-thinking-correctly-about-ai-intelligence-20260820/"
  },
  "original_language": "en",
  "account": "When an LLM answers a question, is it reasoning like humans, or merely generating text that appears to reason? This distinction goes beyond mere philosophy; it shapes our trust in AI, how closely we must supervise it, and ultimately, its real-world impact. Melanie Mitchell, a cognitive scientist and computer scientist at the Santa Fe Institute, argues that we lack proper methods for gauging machine cognition, and that AI operates through non-human cognitive mechanisms. In an episode of The Joy of Why, Mitchell discusses with Steven Strogatz how psychological methods used to study cognition in babies and animals can be applied to probe AI, and outlines six principles for better assessing machine cognition. Their conversation covers interpreting AI's internal workings, recent AI-assisted breakthroughs in mathematics, and a cautionary tale of a math-performing horse from the early 1900s. Listen on Apple Podcasts, Spotify, TuneIn, or Quanta's streaming service.",
  "summary": "Computer scientist Melanie Mitchell discusses why artificial intelligence doesn’t “think” or “reason” like humans, and how we can create better methods for measuring machine cognition. The post Are We Thinking Correctly About AI Intelligence? first appeared on Quanta Magazine",
  "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."
}