#TSQL2sday Roundup: Has AI Helped You with Your SQL Server Job?

Earlier, I hosted a TSQL2sDay on Has AI Helped You with Your SQL Server Job?. I had lots of fun reading various contributions to this topic by industry leaders of SQL Server. Here is the roundup of all the contributions.

Using ChatGPT for T-SQL Code Reviews: Brent Ozar explores the use of ChatGPT for T-SQL code reviews, highlighting its efficiency in identifying security and performance risks, like SQL injection and the misuse of LIKE for string searches. Despite its rapid and helpful insights, Ozar notes ChatGPT’s limitations with more complex code, where advice can become vague and less useful. Yet, he values ChatGPT as a fast, initial check and an educational tool for clients and their tech teams, acknowledging its productivity boost while reminding us of the necessity for human oversight. (Don’t forget to read comments to this blog post.

T-SQL Tuesday #173–The AI Job Helper: Steve Jones discusses the theme set by Pinal Dave on AI’s impact on SQL Server work. While not heavily involved in SQL Server, Way0utwest experiments with AI, like Copilot and ChatGPT, with mixed effectiveness for SQL tasks. AI has proven useful for initial searches, offering a quicker start than traditional methods, though it still requires critical evaluation and testing of its suggestions. This exploration highlights AI’s role as a potentially helpful, yet imperfect, tool in enhancing database management practices.

AI and SQL: Rob Farley reflects on AI’s impact on SQL work, initially skeptical due to a misleading Bing Copilot answer. He’s intrigued by Microsoft’s AI advancements in SQL yet wonders about their effect on his expertise. Despite hesitations, Rob acknowledges using AI for syntax help, remaining open to future possibilities. This is an excellent start on this topic.

AI-Assisted SQL Server Workflows: Chad Callihan delves into using AI for SQL Server tasks for T-SQL Tuesday. He shares his cautious yet optimistic use of AI tools like Copilot for error-solving and generating test data. Despite AI’s limitations, its ability to quickly produce test scenarios showcases potential efficiencies. Chad’s narrative highlights AI as a useful yet not all-encompassing tool, reinforcing the value of human expertise in optimizing SQL Server operations.

T-SQL Tuesday 173 – Has AI Helped You with Your SQL Server Job: Mikey Bronowski shares insights on AI’s growing influence in SQL Server professions, sparked by Pinal’s T-SQL Tuesday invitation. Emphasizing daily use of tools like ChatGPT and GitHub Copilot, Bronowski highlights AI’s benefits in documentation and problem-solving, making tasks more efficient. With excitement for Azure SQL Copilot, he looks forward to intuitive database optimization.

T-SQL Tuesday #173: Todd Kleinhans explores ChatRTX: Todd Kleinhans shares insights into utilizing NVIDIA’s ChatRTX for local analysis of SQL query plans, bypassing the need for internet access and addressing security concerns. His experimentation with ChatRTX showcases its capability to interpret and analyze converted .sqlplan files, providing detailed feedback like wait statistics. This method offers a novel, secure avenue for SQL Server professionals to apply AI in optimizing queries, indicating a significant step forward in SQL performance analysis.

TSQL2sday: AI’s Impact on SQL Server Work: Andy Leonard shares his perspective on AI’s effectiveness in his role, mainly through its assistance in content creation and brand management for SQL Server-related endeavors. While Large Language Models (LLMs) have occasionally produced valuable outputs, his experience with AI for coding, such as NL2SQL, has been mixed. Despite the challenges in direct coding assistance, Leonard remains optimistic about AI’s potential in software development, including T-SQL. His insights highlight AI’s current strengths in supporting business tasks over technical SQL development, illustrating a cautiously optimistic outlook on AI’s future contributions to SQL Server-related work.

AI-Assisted Query Optimization in SQL Server: Pinal Dave shares his exploration of AI in optimizing SQL queries for T-SQL Tuesday. He tested a query for unsold products and used an AI tool to suggest more efficient versions. Surprisingly, the AI-provided alternatives showed lower costs in execution plans but higher resource consumption upon further investigation. This experience highlighted the importance of validating AI suggestions with thorough testing. Dave emphasizes AI’s value in offering new perspectives but reminds SQL Server professionals to rely on their expertise and rigorous analysis for optimal results.

Before we end this roundup, I want to mention a special Twitter (or X) post by Erik Darling on this topic. Just see the tweet yourself.

I have included everyone who has pinged me on my blog, Twitter, or email. However, if I have missed including your blog post, just nudge me by leaving a comment, and I will immediately include yours.

Reference: Pinal Dave (https://blog.sqlauthority.com)