How can generative AI help to shed light on the jungle of regulations in radiation protection? Is it possible to make AI "forget" outdated regulations and substitute by recent ones? And finally - can we avoid hallucination and give correct references to each and every statement ?
Wolfgang Rohde is the head of AiSuNe Foundation Group researching on Generative AI. AiSuNe is a world-wide network of researchers, providing insights into trustworthy and reliable Generative AI systems.
In Germany he partners with the IN3-Group on applied AI systems. Together with IRS he will investigate on “The application of AI to vocational training in radiation protection".
Today's presentation will cover fundamental concepts of Generative AI applications and their potential for analyzing comprehensive knowledge bases, particularly regulatory documents and guidelines.
Eventually we want to answer the question: How can generative AI help to shed light on the jungle of regulations in radiation protection? Is it possible to make AI "forget" outdated regulations and substitute by recent ones? Can we avoid/minimize hallucination? And to increase trust/transparency into the system provide for each statement relevant references into the corpus of knowledge to enable users to verify the correctness of all and each statement.
