Explore the intersection of deep learning and information security in this thought-provoking 20-minute conference talk from USENIX Enigma 2018. Delve into Jeremy Howard's perspective as a deep learning researcher, challenging the notion that deep learning is overhyped in the infosec community. Discover potential advantages of applying well-understood academic approaches to information security challenges. Examine topics such as deep mining, image recognition, and big learnings in infosec, while addressing common misconceptions about deep learning being a black box. Gain insights into why deep learning remains a relevant and interesting field for information security professionals.