Before you can use AI well, you need to know what problem you’re trying to solve. That sounds obvious. In practice, it’s the step most people skip.
I see it in how people talk about AI tools. Someone says they want to “learn how to prompt better” or “get good at using AI.” And when I ask what they’re actually trying to accomplish, there’s often a long pause. The tool became the goal before the problem was ever identified.
That pause is where the real work starts.
Prompting Isn’t a Skill in Itself
There’s been a wave of “prompt engineering” courses and content in the past couple of years. Some of it is useful. But a lot of it is solving the wrong problem.
Prompting well is a byproduct of knowing what you want. If you’re clear about the outcome, the prompt tends to follow. If you’re fuzzy about the outcome, no prompting technique will save you. You’ll get impressive-sounding output that doesn’t actually help, and you’ll spend more time trying to figure out why.
The underlying skill isn’t prompt construction. It’s problem articulation:
-
Can you describe what you’re trying to accomplish? And why?
-
Can you separate the symptom from the actual issue?
-
Can you name the constraint that’s slowing you down?
That’s the harder thing. AI just surfaces how underdeveloped it often is.
When I See It Clearly
I’ve been consulting long enough to know that most problems presented to me are not the actual problem. A client says they need a better reporting tool. What they mean is that two departments are working from different data, and nobody knows who’s right. A team says they need more documentation. What they mean is that context walks out the door every time someone leaves the team.
The same dynamic shows up with AI. People come to it with a vague desire: “I want to be more productive.” But productivity isn’t a problem. It’s an outcome. The problem is what’s eating your time, what you keep avoiding, or the handoff that keeps failing.
When you bring AI a vague desire, you get a vague result. When you bring a specific problem, you get something you can actually use.
Identifying the Problem Is the First Step
AI is an accelerant. It speeds up whatever direction you’re already pointed in. If you don’t know the direction, acceleration doesn’t help.
The question is always the same: what is the actual problem you’re trying to solve?
Not “I want to be better at X.” Not “I want to learn Y.” What, exactly, is not working right now? What would fixed look like?
When you can answer that clearly, the right tool becomes obvious. And the prompt almost writes itself.
The most valuable thing AI has done for me isn’t automate anything. It’s made the cost of unclear thinking impossible to ignore.
Leave a Reply