Deep learning is a subset of machine learning which utilises deep neural networks (networks with more than one hidden layer) to learn how to perform a task without being explicitly programmed to do so. Machine learning (and deep learning) has seen a rise in popularity in the last several years with many of the techniques being applicable to solar physics if used in the correct way. Here, I aim to discuss the fundamentals of how deep neural networks operate and how this can be used to automate laborious tasks and approximate functions which can lead to faster analysis tools. I will discuss the use of deep learning as an alternative for seeing correction in flare observations through the use of autoencoders and the application of such a model to the 6th September 2017 X9.3 flare. Finally, I will talk about the possibilities of using instance segmentation learning for flare ribbon tracking in H? and Calcium ?8542 and the physics we can explore by doing so.