🧠 AI Teaches Itself! Anthropic researchers have developed a method that lets language models fine-tune themselves — no human labels needed! 😮 🔍 It's called Internal Coherence Maximization (ICM): The model improves by checking if its own answers make sense and don’t contradict each other. ✅ Performs great on tasks like truthfulness, math, and helpfulness — even better than human-labeled data in some cases! ⚙️ This could be a game-changer for faster, cheaper, and safer AI development.