Why read this? Information stopped being scarce a while ago. Competent answers did too. What’s still rare is a point of view somebody actually arrived at themselves. This essay is about where mine come from: an hour most evenings that produces nothing at all, what I feed into it, and the three tests that whatever comes out of it has to survive. It’s about the part of the work that doesn’t look like work.
Musing
Most evenings I stop working long before I stop thinking. I am musing.
The tasks are done, my son is asleep, the house is quiet. And then my mind goes wandering through whatever I’ve read that week. A study I skimmed on Tuesday bumps into a conversation from Thursday. Nothing productive happens. There’s no document open, no outline, not a single word on the page.
And yet this is where the next essay is being written.
Musing: to think deeply, to ponder, to be lost in quiet reflection. It’s an old-fashioned word and I’ve grown fond of it, because it names the part of writing that never shows up in a calendar. The work of writing is mostly the hours that don’t look like writing.
I think that’s true far beyond writing.
The bottleneck has moved: Own Position
For most of my professional life, the constraint was access. Finding the study, the benchmark, the case, the framework. Whoever had better inputs had better output.
That constraint is gone. You and I now have the same access, the same models, the same summaries, in seconds. Which means a competent answer has stopped being a differentiator. Everyone can produce one. Everyone does.
What’s left is the thing a model can’t do for you: arriving at a position nobody handed you. Not a hotter take, but a genuine one. Something that comes from your own reading, your own experience, your own quiet evening.
That’s a skill, and it can be practiced. Over the years mine has settled into three parts: the hour itself, what I feed into it, and what the results have to survive afterwards.
The hour itself
1. I protect the unproductive hour
This is the part I’d defend hardest and the one that’s hardest to justify to anyone looking at my calendar.
It isn’t a scheduled slot once a week. It’s several evenings, most weeks. A long drive with a podcast running that I’m only half listening to. A long drive with nothing running at all, just looking out the window. The bathtub, which has probably produced more of my writing than any desk I’ve ever owned. From the outside, all of it looks exactly like doing nothing.
There’s decent evidence that it isn’t. In one well-known experiment, people who spent an incubation period on an undemanding task that let their minds wander performed better on creative problems they’d seen earlier than people who rested or worked on something demanding (Baird et al., 2012). Mind wandering seems to do real work, and what it needs is an activity that occupies you just enough to stop you from steering.
That’s why the podcast doesn’t ruin it. I’m not listening for information. I’m listening loosely enough that my attention keeps sliding off and going somewhere else, which is the whole point. Silence works too. Focus doesn’t.
The practical implication is uncomfortable, though. This time is the first thing that disappears when the calendar fills up. It has no output, no deliverable, nothing to report in a status meeting.
And the calendar isn’t even the bigger threat. The gaps that survive get filled with a phone. Every queue, every red light, every five quiet minutes on the sofa now comes with an obvious way to spend it, and scrolling occupies precisely the part of the head that musing needs free. The empty moments haven’t disappeared. We’ve just stopped letting them stay empty.
Which is why this has to be defended on purpose rather than hoped for.
What I feed it
An empty hour on its own produces nothing. Wandering needs something to wander through, and that part I actually work at.
2. I read far away from marketing
Almost nothing I’ve written that mattered came from reading marketing content.
Marketing sources tell me what marketing already thinks. That’s useful for staying current and nearly useless for having a thought of my own. So the ratio I aim for is roughly two thirds outside the field: sociology, cognitive science, organizational research, technology history, occasionally fiction.
The point isn’t to be well-rounded. It’s that the raw material for a new idea has to come from somewhere the idea doesn’t already exist.
If your reading list looks like your job title, your ideas will look like everyone else’s in that job.
3. I collect what doesn’t fit yet
I keep a running file of observations that have no home. A number that surprised me. A sentence a student said. A pattern in three client conversations that might be nothing.
No structure, no tagging system, no productivity method. Just a place where unrelated things sit next to each other long enough to become related.
This is old practice. The sociologist Niklas Luhmann built his entire body of work on a slip box where notes were linked rather than filed, so connections could surface that he hadn’t planned. And there’s evidence the principle generalizes: an analysis of 17.9 million scientific papers found that the highest-impact work tends to combine mostly conventional material with a small dose of unusual pairings (Uzzi et al., 2013).
Most of my essays exist because two notes that had nothing to do with each other ended up on the same page. The hour is where they meet.
What has to survive
Whatever comes out of an evening like that arrives half-formed and far too pleased with itself. Most of it doesn’t survive the next three steps, which is the entire purpose of the next three steps.
4. I write to find out what I think
I used to believe writing was transcription. You have the idea, then you type it up.
That’s not how any of my good pieces happened. In practice, I sit down convinced I have an argument, and about two thousand words in I discover the argument doesn’t hold and something more interesting is standing behind it. The draft isn’t the output of thinking. It’s the instrument.
Which is also why I’ve become careful about handing the first draft to a model. If a machine writes it, I skip the exact step where the thinking happens, and I end up with a text that reads fine and taught me nothing.
5. I have to teach it
My unfair advantage isn’t the lecture hall itself, it’s the obligation that comes with it.
When I teach generative engine optimization, or how to use Claude Design to build landing pages and produce marketing content, understanding isn’t the goal. Students have to be able to do it by the end of the session. That forces a question I’d otherwise skip: what’s the shortest path from not knowing this at all to actually using it? Everything that doesn’t serve that path has to go.
This is where most of the real work sits. Getting a complex topic into something simple enough for people who don’t think about it every day is genuinely difficult. Simplifying isn’t the step that comes after understanding. It’s how you find out whether you understood it well enough to hand it to someone else.
The rest comes for free. Students have no stake in pretending something works, and the questions they ask are almost never the ones I expected. They push back on the frameworks too. If a model has four boxes and only three of them do any work, someone says so. If a playbook contains a step nobody could actually carry out on a Monday morning, that’s the step they get stuck on. Sometimes one of them just has the better argument, and I go home and rewrite.
You don’t need a lecture hall. Explain the idea to one person who isn’t obliged to be impressed, and try to do it in three sentences.
6. I use AI as an opponent, not an author
I use AI constantly. Just not for the sentences.
What it’s genuinely good at is pressure. I’ll give it a finished argument and ask for the strongest case against it. I’ll ask which claim would collapse first under scrutiny. I’ll ask what a skeptical CMO would object to in the first two paragraphs. I’ll ask it to name the three papers that contradict my position.
That’s an enormously useful sparring partner and it costs nothing. It just isn’t a substitute for having a position in the first place.
Let the model challenge your thinking. Don’t let it do your thinking.
What this means going forward
None of this is a productivity system. I’d be suspicious of anyone who turned it into one.
But I do think it points at something structural. As more of the execution layer in marketing becomes automated, the scarce input shifts upstream. Not to who can produce more content, but to who has something worth producing content about. The competitive edge moves from output speed to the quality of the thinking behind it.
And thinking, inconveniently, still takes the same amount of time it always did.
So most evenings I stop working long before I stop thinking. The next essay is being written right now, without a single word on the page.
Stay tuned.
Key Takeaways
Mind wandering during undemanding activity improves later performance on creative problems (Baird et al., 2012), which makes unproductive time a working method rather than a luxury.
Unproductive time is now lost twice over: first to a full calendar, then to a phone filling every gap that survives. It has to be protected on purpose.
Access to information is no longer a competitive advantage, because everyone has the same access and the same tools. The advantage has moved to having an original position.
Ideas rarely come from within your own field. Reading mostly outside your domain supplies the raw material that the field doesn’t already contain.
High-impact scientific work tends to combine largely conventional knowledge with a small share of unusual combinations (Uzzi et al., 2013). Original thinking is mostly recombination.
Explaining an idea out loud to someone with no obligation to be impressed is the fastest way to find out whether it holds. Compressing it into something simple is not the step after understanding, it’s the test of it.
AI is a strong opponent and a weak author. Use it to attack your argument, not to produce it, because the drafting is where the thinking happens.
What’s your version of the unproductive hour? I’d like to know where other people do their actual thinking. Reply or leave a comment.
Yours,
Prof. Dr. Andreas Fuchs 🦊🎓
References
Baird, B., Smallwood, J., Mrazek, M. D., Kam, J. W. Y., Franklin, M. S., & Schooler, J. W. (2012). Inspired by distraction: Mind wandering facilitates creative incubation. Psychological Science, 23(10), 1117-1122.
Uzzi, B., Mukherjee, S., Stringer, M., & Jones, B. (2013). Atypical combinations and scientific impact. Science, 342(6157), 468-472.
Interestingly, my own experience is not entirely different.
As an architect, I’m expected almost by default to generate ideas on demand. I also grew up in the turbulent world of investment construction, where anyone in a leadership position must always have a few original ideas up their sleeve. Ideas are an unpredictable asset, but some environments leave you very little choice but to learn how to produce them.
There is one thing these environments have in common: they rarely leave you without a problem to solve. There is always a concrete task.
Perhaps for a mind wired to construct, a concrete problem is all the incubation it needs.