# Your first useful hour: three everyday tasks

Three results you can use today: a tighter email, a dated to-do list, and the case against your own choice.

Start here · v1.1 · draft · https://hayzlabs.com/start/your-first-useful-hour/

In short:
- Use three ready prompts: tighten an email you wrote, turn messy notes into a dated to-do list, and argue against your own choice.
- Swap real names and numbers for stand-ins first, and check every number, date, name and promise in what comes back.
- Save the one prompt you would reuse in a file called first-hour.md.

This is step 1 of [Start here](https://hayzlabs.com/start/start-here/). Use any AI assistant (a chat app such as ChatGPT, Claude or Gemini), including the free one you already have. The three tasks are our suggestion.

**Before you start.** Pick an email and some notes you would be happy to show a colleague. Replace real names and numbers with stand-ins. [What not to paste](https://hayzlabs.com/start/what-not-to-paste/) explains what to leave out.

## Task A: tighten an email you already wrote (10 minutes)

Paste this into a new chat:

```text title="Tighten an email · v1.0"
Here is an email I wrote:
<PASTE_EMAIL>
Make it shorter and clearer without changing what it says or how it
sounds. Show the new version, then list each change in one line.
```

Replace the following:
- `<PASTE_EMAIL>`: an email you wrote and have not sent, or sent last week.

How to tell it worked: read the new version against the old one. Check every number, date, name and promise. If something changed meaning, say "You changed X. Put it back." Keep the changes you like and ignore the rest.

## Task B: turn messy notes into a dated to-do list (10 minutes)

Paste this into a new chat:

```text title="Notes to a to-do list · v1.0"
Today is <TODAY_DATE>. My notes: <PASTE_NOTES>
Make a to-do list with action, owner and due date for each item. Write
"unknown" where my notes are silent. Do not guess. List anything unclear
as questions.
```

Replace the following:
- `<TODAY_DATE>`: today's date, such as `2026-10-12`.
- `<PASTE_NOTES>`: your notes, as messy as they are.

How to tell it worked: every item traces to a line in your notes, "unknown" appears wherever your notes were silent, and nothing appears that you did not write.

## Task C: the case against your own choice (10 minutes)

Paste this into a new chat:

```text title="The case against my choice · v1.0"
I am deciding <DECISION>. My choice is <CHOICE>, because <REASON>. Give
me the three strongest reasons this could be a mistake, and what I would
notice in six months if it was. Don't tell me what to choose.
```

Replace the following:
- `<DECISION>`: a small real decision, such as `whether to move our team meeting to Fridays`.
- `<CHOICE>`: what you plan to do.
- `<REASON>`: why, in one sentence.

How to tell it worked: at least one objection you had not thought of. If all three sound weak, ask: "Now give me the three strongest reasons for the option I turned down." This is the habit that [Getting disagreement instead of agreement](https://hayzlabs.com/guides/getting-disagreement-instead-of-agreement/) explains.

## What you keep

Three results, and a plain text file called `first-hour.md`. Before you close the chats, make that file and copy into it the one prompt (the text you pasted into the chat) you would reuse.

> [!LIMIT]
> Three tasks do not show whether AI helps you over weeks.

## Next

[Step 2: Prompting in one page](https://hayzlabs.com/start/prompting-in-one-page/) (15 minutes).