ChatGPT: Building Robust and Trustworthy GPTs

In this course, Jules White—a senior generative AI advisor and professor of computer science—shows you how to navigate ChatGPT’s limitations by understanding how it makes mistakes and how that impacts the selection of use cases. Discover how to craft fact-checkable prompts and develop strategies for detecting and managing hallucinations, ensuring that your AI-driven solutions are accurate and dependable. Learn how to rethink the architecture of your use cases to make them safer, using proven methods like filtering, navigation, and leveraging your own expertise to evaluate output quality. Gain practical insights into building safe operations for use cases where you have sufficient expertise to judge results effectively. Plus, explore the ACHIEVE framework, which helps you design use cases that are both ethical and self-beneficial. By the end of this course, you'll have the knowledge to create fact-checked, trustworthy AI solutions that integrate seamlessly into your workflow.