Secure Numerical Computing is Hard: Lessons from 10 Years of Open Data Science and the Long Road Ahead

Learn about the challenges and security implications of numerical computing in data science and AI through this 35-minute conference talk from Anaconda's perspective. Explore insights from the annual State of Data Science industry survey highlighting enterprise adoption challenges of open source software for data science and machine learning workloads. Examine emerging security concerns in deep learning and AI, including issues with exotic hardware, just-in-time compilation, binary distribution, and data-oriented supply chain attacks. Gain valuable understanding of key principles for managing software and data supply chains in the modern era of machine learning and AI deployment, drawing from a decade of experience in the PyData movement and enterprise Python tool distribution.