Ideas · the objection I get most
“Focus on one thing, or none of them will work.”
I hear it whenever someone learns that I have a day job at IBM, twins, a works-council seat, and eight projects in various states of alive. Usually followed by a Musk anecdote, delivered as either praise or warning. Both miss what is actually going on, so here are five pictures of what I have measured on my own bench.
The short version: focus is a rule about capital and headcount. It was never a rule about a person’s evenings. And the price of one lap around a project fell so far that the advice needs re-pricing, not repeating.
The advice was priced in weekends.
“Do one thing properly” assumed a fixed cost per lap: idea, build, show it to someone, learn, repeat. That cost is the whole reason the advice existed — you only got a handful of laps per year, so you had better spend them all in one place.
When one lap costs six weekends, spreading yourself over four projects means one lap each per year — which really is useless. When it costs an evening, four projects is four laps a week. The advice didn’t change. The price did.
Where this number comes from
It is the “before ↔ after AI” line in my own lifecycle map: before, one lap cost six weekends of writing every line yourself — this is where most projects died, quietly. Not a study. My bench, both sides of it.
Eight projects. Two of them are expensive.
“Eight projects” sounds like eight times the work. It isn’t, and this is the picture I wish I could hand people instead of arguing. What costs a life is not having projects. It is having them in the expensive stages — the ones that need strangers, money and my calendar, not just my attention.
A paused project costs nothing but the honesty of the label. An exited one costs a mentoring call now and then. Only Launch and Scale are greedy — and I keep at most two in there, not by discipline but because there is no room for a third.
So the real rule isn’t “work on one thing”. It is: hold as many as you like, but never let more than two be expensive at once. That is a load limit, and load limits are easier to keep than vows.
Why Pause and Exit are stages and not failures
The children’s book tool is paused because I published the Bitsy books the hard way once, loved the writing and hated the production details — so it waits for an agent that can carry the annoying parts. BYTE and STEM GIrls are exits: both handed over and still running without me. If those three counted as failures, I would look reckless. Labelled honestly, they are what makes room for the two that are live.
Musk is the wrong comparison — different verb.
He gets quoted at me in both directions: as proof it can be done, or as proof that people who try are insufferable. Neither works, because the thing he does is not the thing I do.
He isn’t doing five things at once. He is deciding what five organisations do — a delegation problem with a private jet attached. I am doing one thing at a time, in the two hours after bedtime, across a portfolio that mostly sits still.
Quoting him at a person with evenings is like quoting a shipping line at someone with a canoe. Same water. Nothing else in common.
A pause is an option, not a corpse.
Every focused career sounds focused afterwards. Told forwards, it is a pile of things that sat still for years until one of them was suddenly worth finishing.
If I had “focused”, that paper would have been thrown away in 2015 as a distraction. Keeping a project doesn’t mean working on it. It means not deleting the option.
The part that makes it real, not romantic
The pentomino world championship has run at the Max-Planck-Gymnasium in Saarlouis since 2004 — 80+ players a year, its own rules. I won it; my siblings hold titles too. The paper was the groundwork. What changed in 2026 wasn’t inspiration, it was that a build lap got cheap enough that opening a school-gym championship to everybody became an evening’s work instead of a project.
Not eight projects. One workshop with eight benches.
The objection assumes the projects compete for the same hour. Mostly they don’t — because the ones I build for my own busywork end up serving the others. Each new tool makes the next project cheaper, which is the opposite of spreading thin.
This is also the honest answer to “is it a hobby or a business?” Neither, quite. It is the research budget — and it happens to be the only budget I control completely.
When “focus on one thing” is correct
I am not arguing that the advice is nonsense. It is precise advice that gets repeated outside its own conditions. It holds, hard, when:
- someone else’s money is in it, and they were promised one bet;
- people are employed by it, and their rent depends on the roadmap;
- there is a runway, i.e. a date after which trying again is not an option;
- the project is in Launch or Scale — the two greedy stages.
None of those describe my evenings. All of them describe a funded startup. That is the whole disagreement: the advice is about a company. The objection is aimed at a person.
And one thing I won’t dress up: this only works because Pause and Exit are real stages for me, with real labels, in public. A portfolio without honest labels isn’t a portfolio. It’s a pile.
← Ideas · Carolin Neumann · August 2026