MIT App Inventor

MIT App Inventor is one of the most important educational app-development platforms in the world today, playing a growing role in schools, universities, and STEM programmes globally. Developed at MIT, it allows students to build real Android applications using a visual, block-based approach that emphasizes logical thinking rather than syntax. Its strength lies not only in simplicity, but in depth: through extensions, sensors, and its powerful Web component, MIT App Inventor acts as a true gateway to modern IT and Artificial Intelligence concepts, including APIs, data structures, image classification, and machine learning. For education, it strikes a rare balance—powerful enough to demonstrate real-world technologies, yet accessible enough for beginners and school students.

At one stage, I moved from MIT App Inventor to Thunkable as a primary teaching tool, especially to demonstrate modern generative-AI applications. However, efficient and continued use of Thunkable increasingly requires access to paid subscription tiers, which not all students are ready to have or maintain. For this reason—and because MIT App Inventor remains more aligned with the educational domain—I deliberately moved back to MIT App Inventor as a core platform for teaching AI and IT fundamentals. On this page, you will find a complete, step-by-step tutorial for a practical and engaging educational project: the Parsley vs Coriander Image Classifier. This project introduces students to Artificial Intelligence and Machine Learning in a concrete, hands-on way, using an offline image-classification model (.mdl) built with the Personal Image Classifier extension. It demonstrates how AI systems recognise patterns, compare confidence values, and make decisions—bringing abstract AI concepts to life through a simple, memorable real-world example.