This sample includes a step by step process for dissolving the idea that AI is magic!
It shows you how you can start using GitHub Models to get free access to GPT tech. This has a bunch of steps you can explore with kids. I've provided all the code, but you could strip out any parts you would rather get your students coding themsleves. Or use the code as speed run to have a quick play with AI by changing the tone and manipulating the data.
You'll find in here Python code to play with AI both in the form of a step by step workshop plan in a Jupyter Notebook (.ipynb), as well as a Python (.py) file to see how it all comes together.
You can also find my slides from PyCon AU in this repo that you can use for your own learning, or with your class. I'll put the video of the conference talk up when it's released.
You'll then use that to start playing with AI. You can see how you can change the tone of the AI by playing with the System Message. See how it tackles different questions, and trying changing the temperature.
Once you've had fun there you can see what you can do when you do when you add your own data. We'll do this with a RAG (Retreval Augmented Generation) pattern.
In the Jupyter Notebook there are two steps where you'll explore this by hard coding in some "data" and see how easy it is to add it to our AI API call.
Finally we'll add some "search" functionality to our RAG system to see how we can incorporate a larger dataset to create a chatbot for a petshop. We'll also make sure that the AI is on task by making sure it will be sell-sell-selling our products!