👔 The Best AI Operators Think Like Managers, Not Like Tool Users

There's a mental model for working with AI that most people inherit from their experience with software: find the tool, learn how it works, use it to accomplish specific tasks. The mental model is tool-use, and it produces a certain kind of result.

There's a different mental model that produces a different kind of result: management. Specifically, the kind of thoughtful management you'd apply to a capable but inexperienced hire who needs clear direction, good context, consistent feedback, and well-understood expectations to do their best work.

These mental models produce genuinely different outcomes. Not because the tools are different, but because they shape how people interact with them in every session.

------------- Context -------------

The tool-use mental model tends to produce transactional interactions. You need something done. You open the tool. You describe what you need in the way that feels natural. You evaluate what comes back. You iterate until it's close enough. You move on.

This works. It produces reasonable output. But it carries a specific set of limitations that become most visible when the work requires more than average output. The tool-use approach doesn't naturally lead to investing time in context, because context feels like overhead on a transactional interaction. It doesn't naturally lead to articulating quality standards clearly, because the assumption is that the tool will produce something and you'll adjust it. It doesn't naturally lead to diagnosing what went wrong when output misses the mark, because the instinct is to try a different prompt rather than identify the root cause.

The management mental model produces different habits. A manager who wants good work from a new hire invests time in context upfront rather than treating it as overhead. A manager provides examples of what good looks like rather than leaving quality standards implicit. A manager who gets poor work diagnoses whether the problem was the brief, the capability, or the execution rather than just asking for a redo. These habits, applied to AI interactions, produce significantly different results over time.

------------- What Management Thinking Looks Like in Practice -------------

The management mental model isn't about treating AI tools as people. It's about borrowing the practices that make human management effective and applying them to AI interactions, because those practices are fundamentally about producing good work through a capable but directed resource, and they work in both contexts.

Specifically: good managers invest in onboarding. They don't hand a new hire a task and hope for the best. They invest time upfront ensuring the hire has the context they need: what the work is for, who it's for, what good looks like, what the common mistakes are. This investment pays back across every subsequent piece of work the hire does.

Applied to AI: before any significant piece of work, invest time in the brief. Not just what you want, but who it's for, what good looks like specifically, what you'd want to avoid, any examples that represent the standard you're aiming for. This upfront investment changes the quality of what comes back, not just on this task but as a practice that builds the habit of complete briefing.

Good managers also develop calibration over time. They learn what a given person does well, where they need more support, and how to interpret their work accurately. Applied to AI: developing a clear understanding of where specific tools perform well and where they introduce errors or inconsistency allows for much better quality control. You know where to look carefully rather than reviewing everything at the same level.

A marketing director described her mental model shift as going from "I'll try some prompts and see what comes out" to "I'm briefing a capable junior who needs clear direction to produce their best work." The shift changed how she prepared for AI interactions: more time on the brief, clearer examples, more explicit quality criteria. The output changed accordingly. Revisions dropped by about half. The work she was reviewing was closer to ready rather than further from it.

------------- The Feedback Loop That Builds Quality Over Time -------------

One of the most underused practices in AI work is the feedback loop. Good managers don't just evaluate work. They communicate what was right, what was wrong, and what they'd want to see differently. This communication is what produces improvement over time.

AI tools don't improve from session to session in the way a human hire does. But within a session, or within a well-maintained context document, the feedback loop is real. Articulating specifically what missed the mark and why, rather than just asking for a revision, produces better revisions. And the practice of diagnosing specifically what went wrong builds the habit of understanding what produced the gap, which makes future briefs better.

A consultant who started writing a brief debrief note after any significant AI interaction, one to three sentences describing what worked, what didn't, and what she'd do differently next time, found that her briefing quality improved consistently over four months. The notes weren't extensive. But the practice of reflection built a clearer mental model of what produced good AI output and what didn't, which showed up in better initial briefs.

------------- Practical Moves -------------

First, before any significant AI task, spend five minutes on what a manager would call the briefing: context for the work, who it's for, what good looks like, what to avoid, any reference examples. Treat this as investment, not overhead.

Second, when AI output misses the mark, diagnose specifically what in the brief was unclear or missing before requesting a revision. The diagnosis produces a better revision than just asking for a redo, and it builds the briefing capability that matters across everything.

Third, build context documents for recurring work types, as you would build an onboarding document for a hire who handles a specific function. The document does the context work once rather than in every session.

Fourth, develop a calibration map for the AI tools you use most: the categories of work they handle reliably, the categories where they introduce errors, the types of output that need careful review versus quick checks. This calibration makes your review process more targeted and more efficient.

Fifth, treat the quality of your briefing as a skill to develop, not a fixed capability. It improves with practice and reflection. The professionals who brief AI tools most effectively aren't doing something innately different. They've developed the habit of investing in the brief and of learning from what the output reveals about the brief's clarity.

------------- Reflection -------------

The mental model you bring to AI interactions shapes everything downstream: the quality of the briefs, the usefulness of the output, the efficiency of the review process, and the rate at which your AI-assisted work improves over time.

Tool-use thinking produces reasonable output with consistent ceilings. Management thinking produces output that gets better as the practices improve, because the habits it generates compound in a way that transactional interactions don't.

The practices aren't complicated. They're the same practices that make someone a thoughtful manager applied to a different context. And like management skill, they improve with deliberate attention.

How would your AI interactions change if you approached each one the way you'd approach briefing a capable new hire rather than operating a piece of software?

What would you do differently in the next session?

3 comments

👔 The Best AI Operators Think Like Managers, Not Like Tool Users