Talking to an AI: There’s a ghost in the machine – and it’s us (BTW, ‘yelling’ helps)

Curious users have spent the last year working hard to crack the code of LLMs in an effort to more effectively utilize them – or break them. How should we talk to the bots?