Python is frequently viewed as the go-to language for data science, but the author argues it's not the best choice for many tasks involved in that field. While acknowledging Python's strengths in deep learning, particularly with frameworks like PyTorch, the piece emphasizes its shortcomings in data wrangling, exploratory analysis, and visualization. The author's experience running a computational biology lab reveals a pattern: students proficient in Python often struggle with tasks that could be completed quickly in R. This suggests that the language's complexity can hinder productivity, even among skilled users. The author points out the limitations of Python's libraries and tools, contrasting them with R's more straightforward syntax for data analysis tasks. For example, simple visualizations or calculations that should take moments can end up being time-consuming in Python. The author recalls co-teaching a class where even an expert in Python struggled with convoluted code that would have been simpler in R, reinforcing the notion that the problem lies not with the users but with the programming environment itself. When choosing a language for data science, the author stresses the importance of interactivity and low startup costs. Languages that allow for quick, exploratory analysis are ideal. The article briefly mentions Julia but does not delve into it, suggesting a focus on R and Python as the primary contenders. The author also addresses performance, arguing that convenience and clarity should take precedence over speed in this context. Ultimately, the piece presents a critical view of Python's role in data science, urging users to consider alternatives where appropriate.