Integrating IV&V into a generative AI Enterprise work culture

Jason Blevins (National Railroad Passenger Corporation ( Amtrak))

Keywords
Chatgpt;AI;Generative AI;IV&V;Verification and Validation;Intelligence
Abstract
Because of the power of LLM’s to generate code & narrative text, it has never been more important to apply strong IV&V principles. to ensure that the results generated by GPT models are accurate, reliable, & meet the required standard. Enterprises must prioritize IV&V in the development & deployment of GPT models. This can involve rigorous testing and validation, transparent and ethical use of data, and involving users in the design & development process. This session will draw parallels between the best practices of journalism and how that practice can improve responsible adoption at enterprise scale.