This article describes a number of studies on the effectiveness of what the author calls 'friction bots' - chatbots that emulate difficult people (for example, the person who agrees with everything but won't execute the plan). Overall: "role-playing has been proven time and time again to be good for learning. Rehearsal is good for the brain, too. lLM’s are really good at mimicking human beings. Human-human “simulations” are great! We shouldn’t abandon them. But what LLM’s may introduce is a way to scale the practice."
Today: Total: [Direct link] [Share]
