Researchers have developed an approach that enables machine learning models to learn from each other using human-understandable examples. This method automatically selects the most informative examples to teach a concept, such as images of dogs for the concept of 'dogs'. The team found their approach to be effective in teaching both AIs and humans. --- Why it matters: This matters because it could lead to more transparent and explainable AI models, which are crucial for building trust in AI systems. By allowing machines to learn from each other using human-interpretable examples, the field may move closer to creating more reliable and accountable AI. Source: https://openai.com/index/interpretable-machine-learning-through-teaching