We investigate the potential for Large Language Models (LLMs) to enhance scientific practice within experimentation by identifying key areas, directions, and implications. First, we discuss how these models can improve experimental design, including improving the elicitation wording, coding...