Welcome back to another edition of T-SQL Tuesday. This month’s blog party host is Taiob Ali, who asks “How is AI changing our careers?“

Rewind a Year Ago…

Had you asked me this question a year ago, I would have said not at all. Me in the summer of 2024, still looked at AI as a hyped up fad technology. At that point, all I really knew of AI were stories of people using ChatGPT for research with flawed results, writing papers and blogs with hidden inaccuracies, and generating ridiculous images. To me, it was overblown hype and fluff.

What Changed Since Then?

Two things started to change my perspective. First was SQL Server 2025 – I got involved in Private Preview and one of the flagship enhancements was Vector Search support. The second was Anthony Nocentino (b), who started sharing with me how he was making use of AI while coding. I had a number of “what, really, you can do that?” moments. He showed me how these tools were more than just “dumb chatbots” but could take wider inputs like existing code, and how one could iteratively work with them to modify and make changes.

The industry is not only changing, but changing TOO FAST.

I had the fortune of attending an AI and Data conference here in Boston where I heard that quote and it really resonated with me. I think back to other shifts and evolutions in technology in the last 30 years… the rise of the Internet… smartphones… computing… and I would argue that none of these have moved as fast as we’ve seen AI tech evolve in the last handful of years. While I haven’t keep tabs on it, I am very much doing so now.

Where Am I Today?

If you’ve been tracking my blog, you’ll see that I’ve really dug into Vector Search on SQL Server 2025. My “Practical AI in SQL Server 2025: Ollama Quick Start” blog has been one of my most popular blogs ever! And I am super happy with my new presentation: A Practical Introduction to Vector Search in SQL Server 2025.

How Am I Using AI Day to Day?

I don’t spend my day-to-day living in code, like I once did. But I do find myself having to dig deep and troubleshoot interesting things. Just the other day, had a scenario where a customer was setting up a new WSFC cluster but without shared storage, and were confused as to why some nodes saw volumes presented and others did not. We have a Gemini enterprise license, so I used the Pro model to do some deeper research on my behalf. It generated a very in-depth report for me that helped teach me a ton in regards to how Windows recognizes volumes that are presented to it when it is a standalone install vs when it is a node in a cluster.

What was more important to me is that this Deep Research functionality provided over 40 different citations, enabling me to dig deeper at the sources it used to further validate its findings. Hallucinations (or false output) are definitely a real possibility (and the “why” is a complete separate but fascinating topic), so having a more robust output with citations that I can validate, is a turning point for me.

Like It Or Not, Change is Here…

There’s still a lot of hype and a lot of noise to cut through. And there are also very sharp edges and definite threats and dangers too. But I also strongly believe that we must look past the ragebait headlines and dig deeper into the component technology pieces themselves. I believe that we, as technologists, need to better understand each component so that we can utilize these tools in a responsible, ethical manner.

The other aspect that I am very mindful of is that as a data professional, what drives all of these tools? Data! I feel that we data professionals have an amazing opportunity to future-proof our career journey, if we embrace and build our expertise in this industry.

And here’s an AI generated image of a dog, using the AI image generation built into WordPress. I still prefer real photos but this ain’t too shabby.