AI learned to betray others. Here’s why that’s okay There’s been a lot of buzz about some experiments at DeepMind that study whether AI systems will be aggressive or collaborative…