AI is a game changer, and the change is not the tool. This page is how I think about AI and how I actually use it. It is a living document, and I will update it as my thinking changes.
Last updated: September 2026
The tool does not change the person
People have always had tools, and each new one arrived with a promise. What the promise leaves out is that a tool only goes as far as the person holding it wants to go.
People do change, but only when they want to. That usually happens as they grow, as they get exposed to new information and let it move them. By default, most people do not want change. They want what they already have to keep working. Someone with that mindset stays the same even when you hand them a state-of-the-art tool. They use it to do the old thing slightly faster, or they do not use it at all.
So the attention belongs on the humans: the ones who want to get somewhere, and who keep finding new ways to get there with whatever is within reach. AI is the latest turn of that cycle. It is a powerful turn, but it only turns for the people who were already moving.
There is no skipping the foundation
You can build a mansion with the tools available today. You cannot build one without a foundation. If the foundation is not strong enough, the mansion collapses, and no tool changes that. There is no workaround.
The part people miss is that you will find out. Maybe not on the first build, maybe not on the tenth, but sooner or later something breaks in a way the tool cannot explain to you, and the gap shows.
So I do not finish the foundation before I let myself touch the new toy, and I do not let the new toy stand in for the foundation. I solidify the foundation while I learn, build, and play. The two happen at the same time, or the second one does not last.
How I actually use it
I follow a 25-50-25 rule, which I took from Peter Yang. The first 25 percent is mine: deciding what I am trying to do and what good looks like. The middle 50 percent belongs to AI: the heavy lifting, the drafting, the checking, the parts that used to eat the day. The last 25 percent is mine again: reading every line, judging whether it is right, and owning the result. The middle is where the speed comes from. The two ends are where the quality comes from.
I stand on the shoulders of giants. Someone has usually already built the skill, the tool, or the pattern I need, and I would rather use it well than rebuild it badly.
I care about workflows more than prompts. A good prompt helps once. A good workflow helps every day.
And I treat everything as a system with inputs, processing, and outputs. When something goes wrong, the first question is which of the three broke. That habit is older than AI, and it is the most useful thing I bring to it.
Why I think it is here to stay
Three things convinced me that AI is now its own vertical, and that there is no going back.
The first is agentic workflows. An agent is an automation with a brain. Automation has been around for decades. The brain is new, and it changes what can be handed off.
The second is that Google now certifies it. Google Cloud just released the Professional Agentic Architect certification, a new line built around designing and running agents. A company does not build a certification track for a fad. Google is all in on agentic systems, and to me that is the signal that it is time to embrace them.
The third is a pattern I have seen before. Right now AI is in its Docker moment: individuals picking up tools one at a time and finding out what they can do. Next comes the Kubernetes moment, when agents get orchestrated at scale. After that come the app rewrites, new ways of hosting, new ways of orchestrating. I do not know exactly when or how, but it is the usual cycle.
An extension of you, not a replacement
AI is an extension of you. It makes what you already want to do easier. An LLM is only as good as the person using it. The same model in two different hands produces two very different results, and the difference is the person.
If you still think AI will replace everything, think again.
As Steve Jobs put it in 1998, “A lot of times, people don’t know what they want until you show it to them.” We are in the middle of being shown. Let us make the most of this ongoing human evolution.