How Generative AI impacts [BioImage] Data Science

In this slide deck we dive into AI-assisted code generation with a focus on bio image data analysis code. We learn how good large language models are for this kind of task. We dive deeply into code quality parameters such as functional correctness, executability and result consistency. We learn how to review generated code critically, in particular by reviewing computation results, e.g. when using these tools:

• bia-bob
• git-bob
• sand-bob

Last but not least we argue about sovereign, privacy preserving infrastructure.