#95 – Dawn Song: Adversarial Machine Learning and Computer Security
Published:May 12, 2020 23:20
•1 min read
•Lex Fridman Podcast
Analysis
This article summarizes a podcast episode featuring Dawn Song, a computer science professor at UC Berkeley. The conversation focuses on the intersection of computer security and machine learning, particularly adversarial machine learning. The episode covers various topics, including security vulnerabilities in software, the role of humans in security, adversarial attacks on systems like Tesla Autopilot, privacy attacks, data ownership, blockchain, program synthesis, and the US-China relationship in the context of AI. The podcast provides links to Dawn Song's Twitter, website, and Oasis Labs, as well as information on how to support the podcast.
Key Takeaways
- •The episode explores the challenges of securing machine learning models against adversarial attacks.
- •It discusses the importance of human factors in computer security.
- •The conversation covers a wide range of topics related to AI and its implications.
Reference
“Adversarial machine learning”