Do you think prompting is becoming a real skill?
by•
I’ve been noticing something interesting while working with AI tools.
Give two engineers the same task and the same AI, and the results can be completely different.
Not because one is better overall, but because of how they prompt.
Some are very structured and clear, others are vague, and the output reflects that.
It made me wonder:
Is prompting becoming an actual skill we should measure and improve?
Curious how others are thinking about this. Are you seeing the same thing?
I’ve been experimenting with a small tool around this, but still early.
50 views
Replies
I'd say yes, but I'd put it lower on the priority list than many people seem to. Knowing what good output looks like matters more than knowing how to ask for it. The prompt gets you in the right direction, but without the ability to evaluate what comes back...
I absolutely think so. I've seen same objectives requiring 5-6 prompts by one person being done in 2 by another. My experience is that Prompts creating by LLMs dont work as much as what you write.. as long as I stick with the standard requirements: TCG REI (Task, Context, Goal, References, Evaluate, Iterate) the less I write the better the output.
It kind of shows how critical how formulating your thoughts is..
Prompting is basically the new "Google-fu" but with way higher stakes for your sanity! It’s the bridge between a vague dream and an agentic workflow that actually executes. In my market discovery phase, I realized that finding a sharp niche depends entirely on how well you can talk to the machine. It’s the difference between a "hallucinating toy" and a real logic engine for your business. Do you think we’ll eventually have "Prompting Certifications," or is that just too dystopian to imagine?
Yes, but I’d frame it slightly differently.
Prompting itself is becoming a skill, but the deeper skill is problem framing.
The best results usually come from people who can clearly explain:
- what they want
- what context matters
- what constraints exist
- what a good answer should look like
- how the result will be checked
So I don’t think it’s only about clever prompts. It’s more about structured thinking.
In that sense, prompting is becoming a measurable skill because it reveals how clearly someone can define and communicate a task.
I think this is where it gets interesting.
It doesn’t feel like “prompting” is the skill by itself.
It’s more like a visible layer of something deeper.
The people who get the best results usually:
- know what they’re trying to get to
- give the right context
- set clear constraints
- and can tell when the output is actually good or not
So it ends up being less about wording,
and more about how well you can think through a problem.
Which makes me think the gap we’re seeing between engineers isn’t random.
It’s measurable.
Two people, same AI, same task but very different outcomes.
That probably says something about how they approach problems, not just how they prompt.
So the real question might be:
Do we start evaluating that directly?
Not just “can you code”,
but “can you work with AI to get to the right result”.
That’s actually what I’ve been exploring recently.
Less about prompts themselves, more about how people approach them.
Still early, but curious if that direction resonates:
https://buildly.click