If you're aiming for a book that bridges the gap between AI concepts and their real-world uses, I'd suggest 'The Alignment Problem' by Brian Christian. It doesn't just list applications; it walks through the messy, complicated history of trying to make machine learning systems work reliably and safely in practice—from self-driving cars to medical diagnostics. The narrative digs into why certain theoretical approaches fail when you try to deploy them, which, for me, is where the real learning happens.
A more direct handbook-style recommendation would be 'Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow' by Aurélien Géron. It's dense with code and project walkthroughs, but the explanations around why you choose one model over another for a specific task, like image recognition or natural language processing, are remarkably clear. The applications feel tangible because you're building them step by step, even if you're just following along. I kept it on my desk for months during my last project.