Are you stuck? Talk it out first, even to a rubber duck

It is late, your app still does not do what you want, and you think: just one more prompt. Then another one. You might recognise this, if you have ever built something with AI. And the next morning you open your project and think: oh dear, what has happened here?

In a vibe coding training I talked about this with the participants, alongside the good practices I try to follow myself. Two of them seem to have nothing to do with technology at first glance. Yet they help you out of a knot more often than a better prompt.

Copy the prompt at the bottom of this article. With it, you turn your AI into a sounding board that makes you think, instead of a machine that immediately throws a solution at you.

Stop when you get tired

The first is timeboxing. When you notice you are getting tired, stop. You think you can keep going for a bit, but you send your AI down a path you no longer oversee properly yourself. And then the next day you actually start by tidying up instead of building.

I have experienced it myself that Claude told me it was already late, and that it would be better to stop. It had been dark for a long time by then, and that was in the summer. I found that quite funny. The agent that never gets tired was wiser than its user at that moment.

I would therefore plan your work in blocks beforehand. Agree with yourself how long you will work on it, and let your AI write down where you are at the end. Then you start fresh tomorrow, with an overview instead of a mess.

Frequently asked questions

What is rubber duck debugging?

You explain your problem out loud to a rubber duck, or to someone who only listens. By telling it in steps, you often see for yourself where things go wrong. It also works with your AI, if you ask it to only ask questions.

Why should I stop if I am almost done?

When you are tired, you easily send your AI in a direction that you no longer oversee yourself. The next day you are then tidying up first. Let your AI write down where you are and start fresh tomorrow.

Why does my AI immediately give a solution instead of listening?

An AI wants to help you as quickly as possible. Explicitly state that it must not give a solution yet and may only ask questions, one at a time.

Where can I practise this hands-on?

In 'Vibe Coding with AI' you build an app yourself, and you also practise how to prevent getting stuck. For working with Claude in general, 'Claude AI Power User' is suitable. You can find all dates at Workshops & Events.

Keep reading for free

Register free with your e-mail address and read on. Your account opens the free articles in the learning space; the rest is for subscribers, from Learn Plus up.

No comments yetRead the comments

Comments

Members comment here and read each other's comments. Register for free or sign in to join the conversation.

No comments yet. What do you take away?