Addressing Bias, Transparency, and Reliability in Developing AI Applications скачать в хорошем качестве

Addressing Bias, Transparency, and Reliability in Developing AI Applications

In this session by #DisruptAISummit, Christos Leivadiotis, Enterprise Architect for Cloud and Artificial Intelligence at CCHBC, explores AI applications, focusing on their definition and importance. We delve into bias in AI, examining its types, impacts, and ways to reduce it. Transparency in AI is also discussed, emphasizing its importance and how to achieve it through explainable AI and open-source models. Reliability in AI is another key topic, with a focus on thorough testing and error handling. We identify challenges like complexity and data privacy and discuss solutions such as regulatory frameworks. A case study on brAInbank, a NextGen Knowledge Management System powered by Agentic AI, provides practical insights. Finally, we explore future directions in AI, including emerging trends and innovations, and conclude with a summary and Q&A session.