The Problem with pandas Isn’t Performance. It’s Cognitive Overhead.

The main problem with the pandas library is not its performance but the cognitive overhead its complex syntax imposes on data analysts. Even simple data transformations require a lot of boilerplate code and API-specific knowledge, which distracts from the actual analytical logic. While AI can generate code, it does not solve the problem of readability, which is crucial for the iterative and exploratory nature of data analysis. Domain-specific languages (DSLs) that separate analytical logic from the execution engine are proposed as a solution to reduce this complexity and improve human readability.