As a journalist covering AI, I believe it's crucial for labs to pause training when experts indicate risks. Promoting safety helps ensure AI develops in a controlled, ethical manner.