AI is not magic and it is not a robot uprising. It is software that learns patterns from examples instead of following fixed rules. This week you go deep on core concepts, meet NotebookLM, and train your first flood prediction model for the Caribbean.
Deep dive into core AI concepts: data, models, training, and predictions. We walk through Teens Lecture 1 together (your mentor shares the slides link in session and by email).
Test your aim, focus, and teamwork. Then we talk about how throwing practice is exactly how a model trains: try, miss, adjust, repeat.
Design and start building your own AI app. Bonus session: using AI to get better grades, including how to make NotebookLM your personal study coach.
This is a tiny neural network that decides if a Kingston gully community should get a flood warning. The three circles on the left are the inputs (the data). The middle circles combine the inputs. The right circles are the two possible decisions.
A model is only as good as its training data, and you are the trainer. In real AI work this is called labelling data: a human marks each example so the model can learn from it.
1. Hardest Topic, Simplified (Tuesday and Thursday). Pick the hardest topic you are studying right now (CSEC Physics, Maths, Chemistry, anything). Load your notes into NotebookLM the way we demoed in class, and build a notebook that makes it simple. Submit your notebook below. 🏆 Top notebook wins a prize
Submit your NotebookLM notebook →2. Flood Prediction Challenge (weekend). Play the Google AI Quests flood prediction mission. You will train a real flood-forecasting model, the same kind used to protect real communities. Push your accuracy as high as you can, then submit your score. 🏆 Top score wins a prize
Play the Flood Prediction quest → Submit your flood accuracy score →