The author reflects on the evolution of the data industry and programming, particularly SQL, over the past several years. A central theme is the dichotomy between writing code for efficiency and readability. In the past, debates raged over whether SQL should use leading or trailing commas for better code management. While leading commas can simplify the editing process, they create readability issues, highlighting the tension between ease of writing and ease of understanding. As AI technologies advance, the landscape of coding is shifting. Current trends show that a significant portion of code, like 40% at Coinbase, is now AI-generated. The focus is moving from writing and reading code to testing it, raising questions about how data professionals can keep pace. Unlike other computing fields, where automation can streamline processes, data analysis remains complex. There's a persistent challenge in verifying the accuracy of AI-generated work without manual checking, which diminishes the potential for automation in data analytics. The author proposes a solution: rethinking how SQL is presented. Instead of merely improving tools for writing SQL, there's a need for better ways to read and comprehend it. This could involve visual representations or new languages designed specifically for understanding complex queries. The goal is to create accessible diagrams from SQL queries, making it easier for those familiar with SQL to verify and annotate queries effectively. This approach seeks to bridge the gap between the technical complexity of SQL and the need for clear, understandable data analysis.