Explainable AI for better LLMs and Recommender systems [HvA]
Organizations are concerned about the return on their AI investments and the security of their AI systems, especially in the area of discrimination. For example, in language models (LLMs), XAI methods can contribute to more consistent chatbots. This consistency is not just about more accurate answers, but also about more honest interactions. In addition, such methods in recommendation systems provide more relevant and fairer recommendations to customers.
Key Takeaways:
1. Transparency of your AI systems increases your ROI.
2. Transparency of your AI systems ensures fairer systems.
3. Start working on the transparency of your AI systems now.
Diptish DeyDiptish Dey is a senior researcher and teacher at the Hogeschool of Amsterdam, where he leads the research team Explainability & AI Governance. His research focuses on understanding how AI could add value for business in a responsible way. In this he not only looks at technology, but also to the broader administrative and governance implications.