A very useful use case for LLMs is chatbots. This has been demonstrated primarily with the rise of ChatGPT, as prior to that, LLMs were not as remarkable.

Here is the structure of a typical chat prompt:

[
  {
    "role": "system",
    "content": "You are a helpful assistant."
  },
  {
    "role": "user",
    "content": "Who won the World Cup in 1998?"
  },
  {
    "role": "assistant",
    "content": "France"
  }
]

The most important parameter is the system prompt. This can be adapted in multiple ways, depending on what the generator should serve.

The user message is the input from the chat, and the assistant message is what the LLM generates.