Over the past couple of chapters, we’ve gone through the code of what makes something “AI,” but all of this content is just a small sample of the world of ML/AI in general. Though a lot of the tools you use to make these algorithms will be the same (such as scikit-learn, Keras, and TensorFlow), the implementations will be vastly different depending on the task. However, the general structure we set up for making deep learning models (i.e., make generators -> define model -> define callbacks -> train) does apply to a number of different deep learning-based tasks. Since we don’t have time to talk about everything, we’ll instead talk about how to start your own projects, how to understand errors, and what to do when you encounter those errors.