Coffee Break PD

In the article Generative Artificial Intelligence and Undesirable Output: Developing a Taxonomy for Undesirable Output with a Focus on Hallucination, Eric E. Cohen, CPA, explores the growing concerns surrounding generative AI’s production of unreliable or misleading content—often referred to as "hallucinations." Published by ThinkTWENTY20, this thought-provoking piece proposes a detailed taxonomy to classify various types of undesirable outputs, from factual inaccuracies to stylistic and semantic errors. Through historical context, real-world examples, and expert insights, Cohen highlights the urgent need for standardized definitions, user education, and system guardrails to ensure responsible use of AI tools. This course unpacks these findings and their practical implications for professionals navigating the evolving landscape of AI in business and finance.