TL;DR: The two loud versions of AI adoption are both annoying. The enthusiast who name-drops models and posts about prompts. The refuser who turns every conversation into a complaint about AI. The third version, the serious version, is quiet. The professional uses the tool where it helps, ignores it where it doesn’t, and keeps clear about which is which. Here’s what I actually tell clients about AI.
This closing piece of the doubters series describes the third version in detail. What it looks like in a working week. The discipline that holds it together over years. And why the compounding effect is the part that matters.
The two caricatures the doubter is right to dislike
The pro-AI caricature is the loud adopter. The professional who posts daily about prompts. Name-drops models in conversation. Turns every meeting into a discussion of how AI changes things. Holds strong opinions about which platform is currently winning the race. That person is annoying for good reason. The irritation comes from three places. Social signaling. Enthusiasm standing in for results. And the fact that the loud adopter often produces worse work than the quiet professional using the tools where they help.
The professional uses AI where it helps and does not use it where it does not.Share on X
The anti-AI caricature is the loud refuser. The professional who turns every conversation into a complaint about AI. Treats engagement with the technology as a moral failure. Flexes about not using the tools, as though abstention were a credential. That person is also annoying, and the doubter who recognizes the type usually doesn’t want to become it. Both caricatures share the property of making their relationship to technology a central feature of their professional identity, and the serious professional avoids that move on both sides.
The third version, said plainly
The third version of AI adoption is the one that almost nobody writes articles about, because it doesn’t make for sharp content. The professional uses the tools where they help and doesn’t use them where they don’t. They develop working judgment about which case is which through actual use. They don’t post about the tools, don’t host opinions about which platform is winning, and don’t announce their adoption to colleagues.
The tools are part of how the work gets done, the same way email and calendaring software are part of how the work gets done. Nobody talks about their email adoption strategy. The serious AI user works the same way.
This version is undramatic and quiet, and that’s why it’s the right version for most working professionals. The professional’s identity doesn’t become tied to the technology. Technology becomes one of many tools, useful in some cases, useless in others, ignored when not relevant. The judgment about when to use it lives in the professional’s head and shows up in the work product instead of in the professional’s self-presentation. A piece on the practical version of this covers how working professionals actually integrate AI in real work, away from the rhetoric on both sides.
What does serious AI adoption look like in a working week?
The quiet professional opens an AI chat tool when they have a task it helps with. They use it for the task, get the output, integrate the useful parts, and close the tool. They don’t return to it for tasks where they’ve learned it doesn’t help. The professional doesn’t check news about AI more often than they check news about any other working tool. The total time they spend on AI as a topic, outside of using it for specific tasks, is close to zero.
Over a working week, the integration might look like this. Monday: clean up a meeting transcript using AI, takes ten minutes including verification. Tuesday: draft a first pass of a routine memo by typing the rough points into AI, rewrite the result into voice, takes fifteen minutes. Wednesday: nothing, because the day’s work is all category-two judgment and category-three client meetings where AI has nothing to add. Thursday: use AI to summarize a research document for background context, verify the key claims against the source, takes twenty minutes.
Friday: nothing again, because the day is mostly meetings and direct work. The cumulative AI time is under an hour. Total time saved is several hours, because the tasks AI handled would have taken longer by hand. The professional notices none of this consciously after a few months because the integration is just how the work happens now.
The discipline that holds the line
From the podcast
Martin Ricketts is blunt about the cost of entry for anyone running a business, on Leaders and Their Stories:
If you’re running a business, pay for the things.Twenty dollars a month is the price of the assistant you keep saying you cannot afford to hire.
The serious professional has one discipline that distinguishes them from the loud adopter and from the abstainer. They check honestly, on a regular basis, whether the tools are still helping. The check isn’t a formal review. It’s a habit of noticing. When the AI output requires more cleanup than doing the task by hand would have taken, that’s a signal the tool has stopped helping for that task.
If the output is shaping the professional’s voice in ways they don’t approve of, the tool has crossed into voice work that should stay human. And when the professional cannot remember writing the recent work, that’s a signal to step back and rewrite by hand for a while.
None of those signals require sophisticated detection. The professional notices them in the course of normal work, the way they notice when any other tool is no longer serving them. The discipline is the willingness to act on those signals when they appear. Stop using the tool for the task that no longer benefits. Rewrite by hand for a period. Change the workflow to push the tool back into the safe zone.
A piece on bringing humans back into the work covers the diagnostic for this in more detail.
The compounding effect over years
The quiet professional who has worked this way for two or three years has accumulated something neither the loud adopter nor the loud refuser has. They have working judgment about a class of tools, built from real use across many specific tasks. That judgment lets them decide when and how to use AI without consulting anyone. The judgment is portable. It transfers to new tools, new models, and new use cases as the technology evolves.
The professional doesn’t need to chase each new release, because they can evaluate any new offering against the judgment they already have.
The loud adopter chasing every release has experienced novelty instead of building judgment. Each use was short and superficial. The loud refuser has built no judgment at all. They start from zero whenever they finally engage, which may be under deadline pressure, with no time to develop the discernment experienced users already have.
The quiet professional is positioned for whatever the technology becomes over the next decade. Their judgment was built through use, and use is the only thing that produces judgment.
The honest close
This series has tried to take the doubter position seriously, name what’s correct in it, and engage with what is mistaken. The doubters who are right will remain right, in the categories of work AI doesn’t affect. Those who are partly right will find the case for engagement convincing on the specific points where their work overlaps with what AI does. Doubters who were wrong about the fad framing or the ethical-refusal framing have, I hope, found a more useful position to take than the one they came in with.
The single thing I would have a thoughtful doubter take from this series is the structural test from the previous article. Look at your working week. Identify the category mix. Decide which side of the line you’re on, with honest data instead of inherited skepticism. From that decision, the right answer for you is either to engage with the tools where they help your specific work, or to continue staying out with informed reasons instead of reflexive ones. Either outcome is fine.
What’s not fine is the continued refusal to look, because the refusal to look is what makes the doubter position vulnerable to the compounding gap that catches up later. Look once. Decide once. Then go back to doing the work that’s yours, with whatever tools serve it best.