Will the widespread adoption of prompt engineering commoditize project management skills, or can it help PMs differentiate themselves and command higher value?

From my experience as an administrator and consultant, I'd add that true professional differentiation happens at the intersection of technology and human context.

What resonates most is the focus on "experimenting and identifying new use cases." In my renewable energy projects, I've found that modeling the problem domain first (as I would in software architecture) before applying any technology yields superior results. AI can automate tasks, but structuring problems properly remains a critical human competency.

I agree we're just beginning to understand the transformative potential. Working "backwards" from our project goals (my planning approach) could reveal AI applications that don't just automate existing processes but create entirely new possibilities in managing complex multi-stakeholder projects

Based on my experience as a municipal administrator and environmental management consultant, I've developed several approaches to ensure AI outputs are accurate, relevant, and aligned with project objectives:

First, I approach AI systems much like I approach domain modeling in TypeScript development - by clearly defining the boundaries and entities of the problem space before engaging. This means articulating specific project parameters, constraints, and success criteria in the initial prompts.

When managing renewable energy implementation projects in Bayamón, I found that a "working backwards" methodology is particularly effective - starting with clearly defined deliverables and then decomposing them into specific AI queries. This prevents scope drift in the AI's responses.

For administrative policy work, I've established a verification framework where AI outputs are triangulated against authoritative sources. This is similar to how environmental auditing requires multiple data points to validate findings.

The most successful approach I've implemented is what I call "constraint-based prompting" - where I explicitly define not just what I want the AI to do, but what I explicitly don't want it to include. This boundary-setting produces more focused results.

In my public administration consulting, I've also found that iterative refinement cycles with AI are essential - treating the initial output as a first draft rather than a final product, then providing specific feedback to guide improvement.

Has anyone else found that their professional background provides unique frameworks for effectively guiding AI outputs toward accurate and relevant results?