Use imperfect prompting

Grace Leung validates what I’ve been thinking for a while—people are over complicating their prompting. Instead of trying to prompt AI chats with robotic logic, the LLMs are increasingly adept at conversing with human-like dialogue. I find it more useful to give examples than to use fancy prompt scripts or defining their character like “You are a…”.
Watch: You’re Doing AI Prompting WRONG (Here’s What Works)