Practical AI for Biomedical Researchers

Content

The course Practical AI for Biomedical Researchers will provide participants with information about the architecture, constraints and benefits of AI for scientific purposes, with a particular focus on Large Language Models (LLMs). The course will introduce relevant concepts like tokenization, embeddings and output sampling. The course will explore limitations and risks from modern LLM architecture and deployment: potential security and confidentiality risks, hallucinations, validation, skill atrophy, and sustainability concerns. There will be a session on the use of AI for programming, focusing on agentic tools like Claude Code, and another session on the use of LLMs for literature search, summarization, and ideation. Lastly, we will have a panel discussion involving experts with applied AI experience from a variety of backgrounds.

Learning Outcome

A student who has met the objectives of the course will be able to:  
 

• Understand conceptually how AI models work, particularly LLMs
• Evaluate outputs and limitations of AI models critically, particularly regarding hallucinations, privacy risks and biases
• Be informed about current and upcoming developments in AI
• Make better use of AI for scientific tasks involving searching, summarization and programming