When someone asks whether it’s too late to start using AI, I want to finish the question with them. Too late to do what?
Learning to make something useful for yourself is a different undertaking from changing professions or becoming an AI researcher. You don’t need to solve all of those possibilities before trying the first one.
In March 2026, I was mostly using a free chat account as a slightly different kind of search. By the time I recorded this episode, I’d spent a few months building small things around my actual life. I was forty-four. I had an engineering degree I hadn’t used in decades, and the terminal still felt intimidating.
That’s my starting point. Yours doesn’t have to look the same.
Give it something you already understand
The blank chat box is a surprisingly effective way to forget every sensible question you’ve ever had.
So don’t begin with “teach me AI.” Begin with a task whose result you can judge. Make a rough email clearer. Turn a messy list into a plan for this afternoon. Work out what to cook from ingredients you already have.
Knowing the situation gives you an advantage: you can notice when an answer misses the point. That ability to judge the work matters more than sounding technical.
Start keeping your own useful material
One idea in this episode matters to me beyond the first attempt: the small things accumulate.
A set of instructions that worked. A decision and the reason behind it. An example of what good looks like to you. A rough tool you can improve later. The next project can begin with something you already made.
I think of this as building my pile. It doesn’t require an AI to remember everything automatically. Save the useful parts somewhere you control, and bring the relevant pieces into the next conversation. Revise them when your preferences or circumstances change.
This is also why I wouldn’t wait until I understood every tool. Trying one real task tells you something that another hour of watching tutorials may not.
Your first ten minutes
The first-win prompt asks a few questions about your day, chooses a small task with you, and walks you through it in plain language.
Afterward, check the result. Did it save effort? Did you have to repair it? Would you use it again? Keep one useful instruction and one note about what you changed.
You don’t have to decide today whether AI will be part of a new career. You can decide whether it helped with this one task. That’s enough information to choose the next experiment.
For a longer path, the starter kit takes those small attempts through a month of exploration and into a six-month roadmap.