Machine Learning and Security - The Good, the Bad, and the Hopeful

Machine Learning and Security - The Good, the Bad, and the Hopeful

Explore the intersection of machine learning and security in this 48-minute conference talk presented by Aleksander Madry at "Beyond Crypto: A TCS Perspective," an affiliated event at Crypto 2018. Delve into the success story of machine learning before examining its inherent brittleness and the challenges it poses. Investigate the origins of adversarial examples and their key components. Understand why adversarial robustness comes at a cost and consider potential contributions from cryptography to address these issues. Gain insights into the good, bad, and hopeful aspects of machine learning in the context of security.