As the demand for computer scientists grows, major tech companies, such as Google, are innovating to maintain a vibrant and skilled talent pool. Our collaboration with Google focused on envisioning and crafting an engaging product tailored for teenagers, inspiring them to explore careers in Computer Science through a Machine Learning experience. The Google Education team aimed to provide a user-friendly and enjoyable experience for non-technical individuals to grasp the basics of machine learning. Despite previous iterations, the initiative didn’t resonate with the intended audience, specifically teenagers in STEM academies.

Working closely with Google educators, we identified that the curriculum’s desktop-centric approach and its lack of engaging content were key reasons for its limited impact. To address these challenges, we designed a mobile-first website experience using a relatable metaphor to elucidate the principles of Machine Learning. Taking inspiration from the HBO comedy, Silicon Valley, we introduced “Pizza Not Pizza,” a service that analyzes images and responds with “pizza” if it detects a pizza or “not pizza” if it doesn’t. Behind the scenes, creating and training this model involved tens of thousands of pizza images—slices, pies, Sicilian, and yes, even deep dish, though some argue it’s a casserole rather than pizza.

Throughout the experiment, students will guide the creation of a decision tree model to determine whether something is a “pizza or not.” The primary emphasis will be on a pizza and its diverse attributes. Irrelevant features will be introduced to the dataset, allowing students the option to refine it for a more accurate and efficient model. The objective for students is to achieve a 95% model accuracy using supervised learning.

Although machine learning has often been perceived as a mythical concept, it consistently reshapes people’s lives. Both Google and our team are aligned in recognizing machine learning’s potential to revolutionize businesses and industries. Access to knowledge about machine learning should be open to all individuals eager to acquire this understanding.