# I want to start

You want to make something useful. Start with someone you know, a personal project or a community group. A problem you can see is enough.

**Three cards to get going:**

1. [Talk to the person who has the problem](/book/chapters/04-01-talk-to-the-person-who-has-the-problem.html). Ask them to show you what they do today and where they lose time.
2. [Trying costs less than asking](/book/chapters/05-05-trying-costs-less-than-asking.html). Build a first version in an hour, with an AI if it helps, and show it.
3. [Come back a month later](/book/chapters/06-02-come-back-a-month-later.html). Set the date you will look at what your attempt changed.

**Your first attempt.** Find someone near you who repeats the same task every day: answering the same question, copying the same information, checking the same list. Build the smallest thing that saves them that time. Show it to them the same week.

**What you look at next.** Did they use it? How much time did it save? What did they still have to do by hand? Fix the next thing.

**A full case.** [A useful first attempt](/first-try/) shows a student building a stock list for her aunt's shop in one hour, and what it changed in a week.

---

Keep a card for later and come back to the [four paths](/paths/) when your need changes. To see where you stand, the [builder test](/builder-test/) gives you a level and a line of work.