Someone trained an AI solely on elementary school material: it's a real experiment that led to a 5-billion-parameter language model built at the Max Planck Institute for Intelligent Systems in Tübingen. Li, Zeller, and colleagues call it LittleLearner, and the name is literal. Will a "child" AI, trained only on a specific category of data, then be able to explain things it "hasn't studied," using what it understands? Good question, but don't ask…
This story is only covered by news sources that have yet to be evaluated by the independent media monitoring agencies we use to assess the quality and reliability of news outlets on our platform. Learn more here.
Someone trained an AI solely on elementary school material: it's a real experiment that led to a 5-billion-parameter language model built at the Max Planck Institute for Intelligent Systems in Tübingen. Li, Zeller, and colleagues call it LittleLearner, and the name is literal. Will a "child" AI, trained only on a specific category of data, then be able to explain things it "hasn't studied," using what it understands? Good question, but don't ask…