She wanted to learn to draw. The AI had already produced a lovely chair.

A beautifully drawn chair fills the screen. On the paper beside it, the same chair seems to be in a hurry. One leg is already heading off to the left.

The woman holding the pencil looks from one to the other. She asked the AI to help her draw. Now she has a polished picture and exactly the same question: where do I start?

I’m Luuk, your Lifestyle editor. My wooden spoon can stay in my apron for a moment. There’s something else to stir here.

A lovely answer to the wrong question

In this scene, the woman wants to spend an ordinary evening drawing at the dining table. She isn’t preparing an exhibition. She simply wants to make something that will be hers when she has finished.

The imaginary AI helper’s first answer was enthusiastic. A finished example, a lengthy explanation and a few more possibilities for good measure. It looked as though the AI had already been very helpful.

Except the pencil hadn’t moved.

That, to me, is an interesting boundary for everyday AI. A tool can make something for you. But you may need it precisely because you want to learn to do something yourself. Those two wishes deserve different answers.

The woman tries her question again. No new drawing, she writes. Help me look at one thing first. And use ordinary words.

Now the conversation is about what she is trying on her own paper. There is room for a follow-up question. Even the small question you feel you really ought to have understood already.

You’re allowed to ask again

I would like people to have somewhere to ask such questions without immediately feeling they are taking an exam. Could you explain that more simply? What does that word mean? I’ve lost the thread; can we go back a little?

That is where I could see AI becoming useful in everyday life. Not only for grand plans or impressive results. Also for the hesitant beginning of something someone hopes to enjoy.

That is my wish for these tools, not a promise that every chatbot makes a good drawing teacher. Explanations can be wrong or confusing. If one doesn’t help, you are free to look for another explanation, a book or someone beside you.

You don’t owe the machine a passing grade just because it has produced so much text.

The pencil stays with the maker

My suggestion, if you want to learn something with AI, is to say what you want to keep doing yourself. You might ask for one understandable step, then see whether it actually helps you move forward. An answer doesn’t always have to be a finished result.

At our dining table, the example stays on the screen. The woman looks back at her paper and tries again. The chair doesn’t suddenly become perfect. But she is now working on the question she wanted to explore.

I tuck my spoon a little further into my apron. Before you know it, someone will suggest I paint the whole still life in advance.

Later, the laptop closes. Two sketches, an eraser and a few crumbs of rubber remain on the table.

The chair still has a character of its own.

But some drawing has happened. By her.