Previously in this series (“The Hidden Variable in AI Prompting Is You”), I explored the idea that the best prompt does not come from a template. It comes from you. Your clarity, your thought process, and your willingness to iterate do more to shape AI output than any “magic script.” This follow-up builds on that idea by mapping out five levels of prompting effort, from quick polish to designing how thinking happens, and the kind of support needed to progress through the levels.
Most conversations about AI still start in the wrong place.
People ask, “What prompt should I use?” as if there is a magic template that guarantees a good answer. However, AI does not work that way. The most important factor in prompting is not the script. It is you, your clarity, your thought process, your ability to iterate, and the effort you bring to the task.
In the last article, I introduced two ideas:
- Prompting style, how you frame the task, ask questions, and guide the model.
- Prompting effort, how much cognitive work you invest, breaking down complexity, refining, and iterating.
Two people can use the same prompt and generate completely different results because prompting is not about the words alone; the human thinking behind them drives it. Low effort prompting can handle simple tasks. High-effort prompting is required for strategy, synthesis, and evaluation.
If prompting style and effort matter this much, it helps to have a map to understand the different “levels” of prompting we move through, and the kind of support people need to grow. I see five levels that keep appearing in real organizations.
Thinking in terms of levels is helpful because it makes something invisible more concrete and gives us clearer guardrails. Levels support governance. For work that touches risk, equity, or strategy, you can say, “This should never be handled at Level 1. At minimum, it belongs at Level 3 with clear human review.” That is much clearer than “use AI carefully” and helps leaders match the level of prompting effort to the level of risk.
Levels also help with self-awareness. When people can say, “I am mostly working at Level 1 and 2,” it becomes easier to ask, “Is that enough for the kind of work I do?” It moves the conversation from “I am good or bad with AI” to “This is where I am on the ladder for this task.” Levels give us language for the effort we are actually bringing to prompting, so we can make more honest choices about when to stay where we are and when to move up.
Here is a brief overview of the five levels.
Level 1: The “Make It Pretty” Prompter. At this level, AI is mainly a cleaner and polisher. You ask it to fix grammar, shorten emails, or rewrite a paragraph to sound more professional. It is useful for speed and surface quality, but it does not change how you think or make decisions.
Level 2: The Structured Requester. Here, you start to structure your requests. You might ask for a one-page brief with specific sections, or a table that compares options by cost, risk, and time. You are designing the container for the answer to make the output easier to use and share, but the work is still mostly about formatting and organization.
Level 3: The Analytical Co-creator. You treat AI as a thinking partner. You might ask it to analyze tradeoffs between three strategies, identify risks you may be underestimating, or play the role of a skeptical stakeholder. You bring context and goals; the model helps you see patterns, blind spots, and implications. This level requires more effort in questioning, not just formatting.
Level 4: The Workflow Orchestrator. You design repeatable workflows rather than one-off prompts. For example, you might build a simple sequence that turns raw meeting notes into action items, risks, and draft follow-up emails, or a sequence for drafting and refining client proposals. Your focus shifts from “What is the right prompt?” to “What is the right series of prompts to get from input to outcome?”
Level 5: The Architect of How We Think. You are designing how thinking happens. You use AI to surface patterns in decisions, question assumptions built into your processes, and align prompting practices with your organization’s values, equity commitments, and strategy. The goal is not just better answers on a given task, but better questions, better decisions, and better systems.
As you read these levels, you may notice that you move between them depending on the task. The goal is not to stay at the top all the time, but to be aware of where you are and be intentional about when a higher level of effort is needed.
Closing Reflection: Hybrid Intelligence in Practice
I approach AI through the lens of hybrid intelligence, where humans and technology co-create outcomes. In this model, the human mind remains the core of the system. Prompting becomes an intellectual practice, one that requires curiosity, precision, and the willingness to think deeply. Prompting style is how you show up to the work. Prompting effort is how much of yourself you bring to it.
The five levels are simply a way to name what is already happening:
- At Levels 1 and 2, AI polishes and packages what you already know.
- At Levels 3 and 4, AI becomes a partner in analysis and execution.
- At Level 5, AI helps you examine the system itself, your assumptions, your patterns, and your use of power.
As a scholar-practitioner, I believe our greatest opportunity is not in automating human thought but in elevating it. When we strengthen the human side of the equation, our clarity, our ethics, our discipline, AI becomes a catalyst for insight, not a substitute for it.
As you think about your own work, at which level are you operating most often? Is that where it needs to be?
