I haven’t blogged in a while, been a little busy, but today I found I was in a writing craze or maybe because Oracle announced layoffs, and Oracle is the company I think people should be watching most closely in this whole AI story.
Let me back up first, because there is a method to my madness, (as always.)
Where I started
My first corporate job in IT was at US West, a company that’s now part of CenturyLink after a few decades of acquisitions folded it in. I worked desktop support for a small, isolated group covering Regulatory, HR, EEO and Web. Fourteen guys and me. We ran Novell servers and mainframes, with a couple of NT boxes thrown in for good measure. This job was my first time I’d touched a corporate network.
I learned fast, but what really shaped my career was noticing which tickets nobody else wanted. Database software was one of them. There were a lot of egos in an office full of young guys, and the technology they didn’t want gave me room to dig in and own something and so I did. I got good at it, good enough that Oracle came in to review the work I’d done supporting their applications, and that review turned into a recommendation to make me a Database Administrator.
That single recommendation redirected my entire career and as many know, I’ve never looked back.
It led me somewhere good, but I’ve always worked across platforms, primarily Oracle, but also 20 years in SQL Server, ten in MySQL and PostgreSQL, too, but Oracle is the technology I kept getting pulled back to, even during the five years I spent at Microsoft. I’m known well enough in the Oracle community, though I’ll admit I’ve also been seen as a little less loyal than most, given my interest in other database platforms. Most communities are welcoming people, so I belong to a LOT of communities. I’ve made friendships everywhere, but I still value the ones I have in the Oracle community, but it’s important to know I just tend to march to my own beat. Sometimes that beat is red, sometimes blue, sometimes it’s every shade of amber or the whole rainbow. I’m fine with that, and honestly, so is most everyone else.
Getting In AI Early
I’ve been in AI since before it was the thing everyone talks about. I was at Microsoft in 2019 when Copilot first started getting pushed internally. After I left, I got pulled into a run of AI projects with barely any formal expertise to speak of, and had to learn Generative, RAG and Agentic AI as I went. That hands-on foundation put me ahead of where most people in the field are only just arriving now. It’s also what led to me speaking on stage for the last three-plus years about Data Protection in the Age of AI.
Those talks landed with so many, but what I hear most often lately is some version of “this is finally starting to make sense to me.” This tells you something about the pace of adoption versus the pace of understanding and that they’re not on the same curve.
The Shiny People or the Steady
I can spot the “happy, shiny AI people” from a mile off, (and I’m sure they see me, too.) They’re the ones selling more consumption, more usage, more adoption, with almost nothing said about doing it right: efficiently, safely, securely. If you’ve ever run long distance, you know the racer who sprints out of the gate almost never wins, while it’s the one who holds a steady, deliberate pace the whole way. I try to be the latter in all things, even if I am told to slow down, I return to that steady pace sooner or later.
There’s an analogy I heard recently from Scott Hanselman that’s stuck with me. AI right now is a bit like a toddler being told he can walk to the store and get candy, except somebody handed him a chainsaw first. Its kind of difficult to picture a kid, with a chainsaw, with no context, no guardrails, and a goal. It has the makings of a Stephen King novel, and you know how bad some of those scenes end up in his novels.
My AI View
I do work on AI built into my own company’s products. I sit in on design reviews, look at mockups, flag security concerns, push back on UI decisions that don’t hold up. That’s part of the job I have as an advocate and engineer.
But when I’m on stage or writing, I commonly use AI very differently. I’m reaching for something open source and Ollama is my go-to, paired with an open model like Qwen, Gemma or Mistral. Qwen 2.5 Coder and Qwen 3 are my favorites specifically, lightweight enough to run locally while I build out scripted solutions to the everyday problems data professionals face. I like to demonstrate that you can do AI without a massive cloud bill and do it without a black box. Just a tool they can understand and can trust, but isn’t that many or our goal when we demonstrate something for a group?
Why Today
Today gave us another layoff in the ongoing AI bubble story, this time from Oracle. The journey of the never-ending data center and infrastructure for AI is a march that never ends. The need for more infra, more power, more water, for more AI in exchange for more debt and less jobs has experienced another deep cut into an organization, but many don’t understand why the Oracle “dinosaur” is even part of the scenario.
I tell people constantly that Oracle deserves closer attention than Anthropic, Nvidia, OpenAI, Meta or SoftBank. Most focus on the big names, thinking Oracle is a dinosaur, but no one thinks why its involved or what happens if Oracle falters.
Oracle is holding up more of this bubble than most people realize, and part of the reason it can’t be allowed to fail is who’s depending on it. The American government is one of Oracle’s biggest customers. If Oracle goes under, the DOD, the DOJ and a long list of other agencies lose the software they’re built on top of. That’s not a scenario anyone in power is going to allow. Let’s also be honest with ourselves: it won’t be called a bailout. It’ll be called something that sounds like better branding, but we do have to recognize Oracle’s role here.
So that’s the picture underneath all of this and yes, I know, it’s all a bit of a web of interconnected, cross-invested companies, holding the AI bubble together, resting on the backs of American jobs that used to belong to real people. It’s the strange, split feeling we’re all navigating right now, as we are loving what AI can do while watching what it’s already costing.
To everyone who lost their job today, including people I know personally, my heart goes out to you. If there is anything or anyone in my network that can assist you in the next step in your career journey, please do not hesitate to reach out to me.
