Watch a 24-minute lecture from UC Berkeley professor Dawn Song at the Simons Institute exploring critical aspects of responsible AI development, focusing on alignment, trust mechanisms, watermarking techniques, and copyright challenges in Large Language Models (LLMs). Gain insights into the sociotechnical framework needed to ensure safe AI advancement through science and evidence-based approaches. Learn about current challenges and proposed solutions in creating trustworthy AI systems while understanding the complex interplay between technical implementation and social responsibility.
A Sociotechnical Approach to a Safe, Responsible AI Future: A Path for Science and Evidence-based AI Alignment and Trust
Overview
Syllabus
A Sociotechnical Approach to A Safe, Responsible AI Future: A Path for Science‑ and Evidence‑based..
Taught by
Simons Institute
Related Courses
-
Pluralistic Alignment: A Roadmap for Integrating Diverse Human Values in AI Systems
-
Empowering Responsible Use of Large Language Models
-
AI Interactions: Misuse, Markets, and Managers in Complex Systems
-
LLM Alignment - Techniques for Building Human-Aligned AI
-
Evolving Trends in Prompt Engineering for Large Language Models with Built-in Responsible AI Practices
-
Is AI Alignment Even Necessary? Thoughts on OpenAI's Uncensored Internal Chain of Thought Approach
