Where it goes wrong: the text with no decision
Ask ten people why they put off writing something and nine will say they don't know how to start. That is rarely true. Usually they don't know what they want to say — and you only notice that at the moment you have to put it on paper.
That is good news, because it means the real work sits before the writing. This course is about that: decide first, then formulate. The AI comes afterwards, and it turns out to be strikingly more useful that way.
The text that keeps everything open
You know the kind. It says the project 'requires due attention', that 'we would be glad to discuss this further' and that 'the planning is subject to further alignment'. Everybody reads it, nobody does anything, and two weeks later nothing has changed.
Sentences like that are not a sign of incompetence. They appear because the writer dodged a decision: he wasn't sure whether he was allowed to say it isn't going to work, so he wrote something that roughly means that without saying it.
An AI tool does not fix that. Feed it that vagueness and you get it back in smoother sentences — which is worse, because now it reads like a finished text.
What an AI does and doesn't do for you
What it does well: phrasing, shortening, ordering, rewriting for a different reader, and producing the fifth variant of something you are stuck on. That is the craft side of writing, and there it saves time.
What it does not do: work out what you want to achieve, know what is going on in your organisation, or decide what you do and do not say. That is the substance, and it stays yours.
So the dividing line is not 'good writing versus bad writing' but decision versus execution. Anyone who grasps that uses these tools better in ten minutes than someone who has memorised twenty tricks.
What this does to your time
In practice your work shifts. You spend longer working out what you want to say — often no more than a minute or two — and far less time phrasing, rewriting and polishing.
The gain sits mostly in the part where you normally start over three times. That is exactly the part you can hand over, provided you have decided beforehand what has to be in the text.
- Find a message in your sent folder from the past month that got no reply, or a reply you didn't want.
- Mark the sentences that decide nothing: 'we'll look into it', 'shortly', 'in consultation', 'where possible'.
- Write below the email, in one sentence, what you actually wanted to happen.
- Check whether that sentence appears in the email. For most people it doesn't, or it sits at the bottom.
- Keep this email. In lesson 2.1 you rewrite it, and then you will see the difference.
Stuck?
- You can't be that direct where I work. Direct is not the same as blunt. 'I need your approval on this quotation before Friday, otherwise delivery moves' is clear and polite at once. What people experience as blunt is usually the tone, not the clarity.
- Sometimes I want to keep things open. That's allowed, but then it is your decision and you write it down: 'we haven't made a choice yet; we will know more in March'. That is different from being vague because you don't dare put it in writing.
- I'm not a bad writer. Probably not. This course is not about writing better but about getting to something usable faster, and about texts that do what you intended.
- Do I have to do this for every email? No. For three lines to a colleague this is nonsense. It applies to texts where something is at stake: a customer, a decision, money, or someone you need to convince.
- Why do texts turn vague, and what happens when you hand that vagueness to an AI?
- Which part of the writing can you hand over and which part can't you?
- Where does the time saving actually sit?
- When is this approach overkill?