Abstract

In this presentation, we will discuss the outcomes and challenges of a year-long initiative to teach AI to a diverse group of European research supporters. The target audience includes pre-award, post-award, and administrative staff with limited knowledge of AI. The main objectives are to emphasize the importance of continuous learning, the effectiveness of learning circles, and the benefits of open-source and locally run LLMs.

The methodology involved maintaining a weekly learning circle with participants from various departments, inviting legal services, communicating with colleagues from other universities and EU institutions, and offering online workshops and conferences. Key outcomes include increased interest and understanding of AI's limits and use cases, heightened awareness of legal and ethical challenges, and a shift towards EU-based, open-source, and locally run LLMs.

Challenges such as concerns about legal and ethical issues were addressed through the inclusion of legal and ethics specialists. Maintaining technological intelligence on AI evolution ensured participants stayed up-to-date. Future directions emphasize the creation of dynamic and adaptable learning structures to facilitate a safe and current approach to generative AI use.