RogerDearnaley comments on We Need a Baseline for LLM-Aided Experiments