The Mechanics of AI - How Does It Work?

Session Summary
Building on hands-on training experience, students investigate fundamental AI mechanics: data, algorithms, and predictions. The "Fleep and Bloop" classification activity demonstrates supervised learning, students observe labeled alien creatures, identify patterns, and predict new classifications, mirroring AI's learning process.

Through a drawing prediction game, students discover how partial information and prior experience enable predictions, connecting human pattern recognition to AI processes.

Google's Quick, Draw! experience is next, where students draw while AI predicts in real-time, followed by exploring the massive training dataset showing millions of diverse drawings.

Students analyse why AI succeeds or fails, understanding that AI recognises patterns rather than truly thinking.

The session transforms students into AI analysts who critically evaluate how systems work and where they might fail.