The most important AI skill has nothing to do with AI
If you want to use AI without losing yourself in it, one thing matters most: knowing what you want before you ask. Not writing better prompts, but being clear about where you are heading. Without that picture you cannot judge an answer, you can only accept it. This article explains why, how it connects to LAND from the InnerPower Method, and what small step you can take today.
It starts with a vague idea
You have something in mind. Not sharp yet, but it needs to be done. You open an AI, type a few lines, and out comes an answer that looks good. Well written, complete, convincing.
And then comes the question most of us skip: is this what I was actually looking for?
I did exactly the same when I started. Asking blind questions, with nothing around it that knew my goals, or knew me. I got all sorts of things back, just not what I was looking for. It was not the technology. It was that I did not yet know what I wanted.
What sits underneath
Alberto Romero wrote a sharp piece about this, which I read through AI Report. His point: most people use AI badly, and not because they type the wrong words. They do not know what they want before they open the screen.
That is an uncomfortable observation, because it moves the problem from the tool to the person using it. And it is right. The moment you ask a vague question, the model makes hundreds of small decisions you never made. Which tone, what length, what you already know, what you meant. It guesses, and it guesses convincingly.
That is why the answer feels good even when it solves the wrong problem. You have no measure to hold it against, because that measure was your own goal, and you did not have it yet.
Why this is older than AI
This is not a new problem. It has only become more visible.
Someone who does not know what they want will be led by the first voice that offers something. It used to be someone else’s opinion, the expectations around you, or simply the path of least resistance. Now it is sometimes a language model that proposes something plausible within two seconds.
So the question is not: how do I prompt better. The question is: do I actually know what I want, before I hand it over?
Why I always start with LAND
Within the InnerPower Method I work with three movements: LAND • BREATHE • FEEL. They come in that order, and that is not an accident.
LAND is the first one, and the one most often skipped. Landing means pausing to notice where you are before you start moving. That applies to your body, and it applies just as much to your work. Before you reach for an instrument, a method or a technology, the first question is always: what do I actually want here?
In a body it works much like it does behind a screen. Whoever moves on without landing takes what is offered. Whoever lands first can choose.
What your body has to do with it
This sounds like a thinking exercise, but it is not entirely. Knowing what you want is largely a physical experience.
When you are rushed, you choose badly. Your nervous system is set to act, not to weigh. You grab the first thing that looks workable, and that is exactly the moment a convincing AI answer carries you along. Slow down for a moment and space appears between the question and the answer. In that space you notice whether something fits.
That is not vagueness, that is how attention works. You do not have to believe it, you can simply try it.
A small step for today
Next time you want to ask an AI something, do this first. It takes thirty seconds.
- Put your feet on the floor and take three slow breaths.
- Write down in one sentence what you want to achieve. Not what you want to ask, but what needs to be finished.
- Add a few words about how you will know it is good.
Only then open the screen. You will notice your question sharpens by itself, and more importantly: that you can judge the answer instead of accepting it.
Does it not work straight away? That is information too. Then you know you did not yet know what you wanted, and that is exactly the gain.
Coming back to yourself, again and again
A lot is written about the skills you need in a world with AI. Prompting, choosing tools, building workflows. All of that can be learned, and all of it changes every year.
What does not change is this: knowing what you want, and being able to come back to yourself when you lose sight of it. That is the only skill that scales, with or without AI.
Frequently asked questions
How do I ask an AI a good question?
Do not start with the wording, start with your goal. First write down in one sentence what needs to be finished and how you will know it is good. Then give context: who is it for, what does the reader already know, how long may it be. The less the model has to guess, the more usable the answer.
Why do I get answers that look good but are not right?
Because with a vague question a language model makes hundreds of small choices for you: tone, length, angle, what you supposedly already know. Those choices are plausible, but they are not yours. Without your own measure you will not notice the difference, because the answer sounds good.
Can I ask an AI to question me first instead of answering straight away?
Yes, and it is one of the simplest ways to get better results. Add at the end of your request: ask me three questions before you begin. It forces you to put into words what you actually want, which is exactly the work you would otherwise skip.
How do I stop my writing from sounding generic?
A model is built to work for everyone, which is precisely why it polishes away what is yours. What helps: bring the content only you have. Your own examples, your own words, what went wrong. Let the model help with structure and pace, not with who you are.
How much context should I give?
More than you think. Context is the most important ingredient for usable output. Who is the reader, what is the goal, what should not be in it, what have you already tried. A five-line brief almost always beats a one-line question.
What if I do not know what I want yet?
Then that is your first task, not the machine’s. Slow down: feet on the floor, three calm breaths, and write down what needs to be finished. If you cannot get it into one sentence, the idea is not ripe yet. That is not failure, that is information.
Is this the same as mindfulness?
There is overlap, but the aim is different. Mindfulness is about being present with what is here. LAND is about knowing where you stand first, so that you can then choose deliberately. It is not an exercise to become calm, it is a condition for deciding well.
Going further
If you would like to recognise this in your body rather than only understand it, that is what the trainings of House of InnerPower are about. You learn to work with the nervous system, the breath and slowing down, and you notice how differently you choose once you have landed first. Current dates are on the House of InnerPower agenda.
Stay close to this work
Join the House of InnerPower newsletter
Sign up and you’ll get one meaningful welcome email to start, from Ilse-Marie Sobering.
Ilse-Marie Sobering is the founder of House of InnerPower and the creator of the InnerPower Method: LAND • ADEM • VOEL (Land, Breathe, Feel), a cyclical practice of returning to yourself again and again. She has studied yoga philosophy and Eastern philosophy for more than 25 years, and has worked professionally since 2015 as an E-RYT 500 yoga and Yoga Nidra teacher, mindfulness and breathwork coach, Ayurveda practitioner and EFT (Emotional Freedom Techniques) practitioner. In Spain, where she lived for seven years, she taught trainings, teacher trainings and retreats at various locations in the Málaga region, including near Periana, Nerja and Frigiliana. She now works from the Netherlands, teaching trainings and speaking in the Netherlands, in Spain and online. Within the method she works with Yin Yoga, Yoga Nidra, breathwork, meditation, Qigong and somatic work, complemented with elements from her hypnotherapy training. She teaches teacher trainings, continuing education and company training, for people living with long-term stress and for professionals who want to deepen the way they guide others.
