Large Language Models (LLMs) change the way we do Bio-image Analysis. In this slides I outline how we came from GPU-accelerated bio-image analysis using CLIJ and clesperanto, and used code-generators to produce reporducible Bio-image Analysis workflows, e.g. as Jupyter Notebooks. LLMs such as ChatGPT now addsanother way for creating these workflows much more convienently. We see tools in action such as the Bio-image Analysis GPT, BiA-Bob and blablado which aim to accelerate the adoption of LLMs in the Bio-image Analysis Community for facilitate reporducible workflow creation.