Abstract Description: In the environmental world there is no shortage of documents. The advancements in generative artificial intelligence have fundamentally changed the way we can interact with text and documents, but there are still challenges for how to operationalize this technology and make it approachable for environmental professionals.
This session focuses on practical ways that generative AI can be used in environmental regulation and compliance contexts. It includes examples and case studies of how AI tools can be used to effectively understand and synthesize permits and regulations. It presents a survey of the current state of technology as well as investigation into what lies ahead in this fast-changing technology. Additionally, it discusses how to address the "black box" nature of AI and ways to effectively validate output that has environmental professionals and subject matter experts overseeing AI tools.
The session reviews key terms and concepts around generative AI such as data pre-processing, fine-tuning, and retrieval augmented generation. We will examine both enterprise and custom applications and considers the benefits and drawbacks of different platforms. There is also an emphasis on how to securely use AI tools, without exposing confidential data to third parties, as well as the ethical and responsible use of AI.