I have now sat in rooms with several thousand people who were told, in one way or another, that AI was not for them. Teachers. Clerks. Small business owners. The message they absorbed was always some version of: this belongs to the technical people.
It doesn't. What belongs to the technical people is the plumbing. What belongs to you is the judgement about what is worth doing.
So here is the least technical advice I can give.
Start with a task, not a tool
Almost everyone starts by opening a chatbot and typing "what can you do?" That is the worst possible first move, because the answer is a list of capabilities with no relationship to your Tuesday.
Instead, take ten minutes and write down three things you did last week that felt repetitive, or slow, or like something you should have been able to hand off. Not ambitious things. The email you rewrote four times. The meeting notes nobody reads. The report you rebuild from the same spreadsheet every month.
Pick the smallest one.
Run three small experiments
An experiment has a shape: a task, a try, and an honest verdict. Give each one twenty minutes.
- Rewrite something you already wrote. Paste in a message you sent and ask for three versions: shorter, warmer, more direct. Keep the one you'd actually send.
- Summarise something you already read. Give it a document you know well and ask what it thinks the argument is. You are testing whether it understood, and you already know the answer.
- Draft something you keep postponing. The awkward email. The first paragraph of the proposal. You are not looking for a finished thing; you are looking for a starting point that is better than a blank page.
Then write down what happened. Not "it was impressive" — what specifically was useful, and what was wrong.
The only real skill
After a few of these you notice the pattern: the people who get good at this are not the ones with the best tools. They are the ones who can describe what they want precisely, and who can tell when the answer is confidently wrong.
That is a writing skill and a judgement skill. You already have both. You have been using them on people your whole career.
What to ignore for now
Ignore agents, workflows, fine-tuning, and anything with a waitlist. Ignore the person on LinkedIn who says everything has changed forever. Ignore the urge to buy a course before you have run three experiments of your own.
The gap between the people who benefit from this and the people who don't is not technical. It is about ninety minutes of deliberate practice and a willingness to be unimpressed.
