Newsletter 300 and AI

Related blog post

Newsletter #300 is a significant milestone for SQLBI: it marks nearly 12 years of sending our updates every two weeks. I thought this was a good time to clarify SQLBI’s position on a topic we haven't discussed much yet: Artificial Intelligence (AI).

SQLBI and artificial intelligence: clarifying our position

At SQLBI, we primarily create content focused on Microsoft BI technologies. Over the years, we've specialized in DAX and semantic modeling because that's where we can deliver the most value. While we know and occasionally cover other technologies, our goal is always to produce content that will remain relevant for a long time. We avoid spending resources on the shiny, latest trends because we want our articles, books, and videos to provide value over time. We’ve been intentionally cautious about discussing AI publicly—not because we don’t believe in it, but because we try to avoid contributing to the hype. AI is transforming many industries, including BI, and it will be around for a long time. However, I've noticed that whenever I mention AI on social media or in videos—sometimes seriously, sometimes sarcastically— people interpret my words according to their existing beliefs. Those who fear AI think we agree with them, while enthusiasts take any positive mention as confirmation that AI is the solution to everything. The reality is somewhere in between. At SQLBI, we try to keep a balanced and objective perspective. Our job is to help professionals working with the Microsoft BI ecosystem make more informed, better decisions. We focus on what’s practical and valuable today, rather than betting on uncertain future scenarios.

AI in BI: current realities and future perspectives

There’s no doubt that AI is powerful, but right now, there are clear limitations to its effectiveness in BI. A common misconception is that AI can already do everything for us—like automatically generating queries without human oversight. That’s simply not true. AI tools, especially those related to semantic modeling, still require substantial human validation. It’s interesting that AI performs significantly better when paired with carefully-designed semantic models—meaning humans actually need to work more to prepare the model, not less. Is this extra investment justified? It depends. If your semantic model is to be used widely, across many reports and by many people, then yes, it makes sense to invest extra time to ensure AI can use it well. If it's just for a small-scale project, traditional methods might still be faster and more cost-effective. We began exploring how AI tools can enhance productivity in BI, starting with the early versions of Q&A in Power BI and continuing with the introduction of ChatGPT. However, we never wrote much about them, aside from several unplugged videos. In newsletter #300, we include content about services that leverage AI to help build semantic models. These tools can significantly accelerate prototyping and development, even though completing the job often still requires human intervention. Recent practical experiences confirm our initial gut feelings: AI speeds up specific tasks dramatically, but the full end-to-end time savings aren’t always as impressive as the initial prototypes suggest. Going forward, we’ll continue to discuss AI only when we see clear, measurable benefits—never just because it’s trendy or exciting. We do not showcase flashy demos that don’t hold up when applied to real-world data. Our audience's attention is precious, and we prefer not to waste it.

Newsletter #300 and beyond: our continuing commitment

So what can you expect from SQLBI in the future? Simply put, we'll keep doing what we've always done: share content that helps you do your job better. Of course, this means adapting to changes in technology as it evolves. The only certainty in technology is change, and SQLBI will continue to adapt our content to those changes accordingly. But our focus remains practical—highlighting what actually works today, not just promising ideas. Our newsletter is also meant to be enjoyable. We take our work seriously, but we never take ourselves too seriously. Humor and a human touch matter, and we hope to keep providing both as we help you navigate changes in BI technology. There has always been a healthy dose of nerd humor in the news section of our newsletter. You will find new, small surprises here and there. I hope they will make you smile. Thank you for joining us on this journey. With 300 newsletters behind us and many more to come, we hope you'll continue to find value in our work. Enjoy DAX!