On Adversarial Training & Robustness with Bhavna Gopal

Bhavna Gopal, a PhD candidate at Duke with research experience at top tech companies, uncovers the world of adversarial training and AI robustness. She explains how adversarial attacks threaten AI model integrity, especially in sensitive fields like healthcare and law. The conversation touches on the challenges of evaluating model performance and the ethical ramifications of AI deployment. Also discussed are the complexities of self-driving cars and the importance of interpretability in ensuring public trust in AI technologies.