5.3 Writing: Choose Your Prompting Adventure
There’s a great deal you can do with a foundation model, like OpenAI’s ChatGPT or Anthropic’s Claude. Don’t feel overwhelmed—you won’t be able to learn all of this in one day. Take some time today to go on your own prompting adventure and think about exploring some different features of each model.
Tips for working with a large language model:
Provide clear, detailed instructions: If you provide vague or incomplete instructions then your output won’t be good. This is one of the main reasons students do so poorly when they first work with AI–they don’t know how to ask good questions!
Give the model examples: AI is like a guessing engine–it works best if you give clear examples by uploading it to the chat. If you are looking to use AI to brush up on your resume, then you want to give the model two or three examples of what you believe a good resume looks like before uploading your own and asking Claude for advice to improve it.
Give the model time: You are going to feel as weird reading this as I do writing it, but emerging consensus is that all foundation models produce better results if you instruct it to “take its time” when developing its output. Language models are odd ducks and prompt engineering is one of the numerous oddities that exist.
Choose Your Prompting Adventure:
Each activity walks you through how to do a specific task using a large language model. You’ll only have time to complete one in the session today, but can always return later: