BAIR faculty @JacobSteinhardt discusses the challenges of making AI systems more reliable—and how his work at @TransluceAI aims to bring more transparency to AI models

→ View original post on X — @berkeley_ai,