Text generated by AI models like GPT-4 and Claude is getting increasingly difficult to tell apart from human writing. Researchers at Penn and Northwestern developed a statistical method that tests...Read full article