Conversation

Replying to
That youthful energy is now mostly gone, and I periods of 5k word burst energy are declining. I have to cashflow/time dependency equation OR accept the drift towards shorter and shorter forms until I'm only doing aphorisms :D
2
4
So what size should you target? Suppose you write 1000 words/day, so 7000 words/week, should you write 7 1000 word things, 14, 500 word things? 2 3-4k word things? 1 7k word things? The "cost" to you will range from $350 to $7000. But it's a trick question...
1
If you write 1000 words a day at a current chunk size of 1000 words (so you finish a thing a day), that does NOT mean you can work on 1/7th of a 7000 word thing per day over a week. Nope. Chances are, if you work 7 days on 1 thing, the end result will be more like 3500 words.
1
1
Due to the overheads of reloading the mental model into your head every morning and quitting at a stable place every evening, you are less efficient. The rework risks also go up (the chunks you write and have to throw away) with size and breakdown structure.
1
2
If you write in 1 sitting, there's a good chance, if you're decent, that you'll be able to write the whole thing in a single flow without anything needing to be thrown away. Very little chance of that happening as you go longer and split the work across sessions, days, weeks...
2
3
I could probably model this out properly under various conditions of time/cashflow, and the boundary conditions of savings/steady-job-income you would need to occupy various positions on the intensity (words/day) vs. chunk size plane
1
It's a constraint solver problem. If you have a job that pays X and leaves you with Y free hours per week, and you write at Z words/hour and want to target a chunk size of W, will you be able to reach an equilibrium output flow?
1
Think of it as: if you can't spend $x/word in one way or another to write at a given length, but still try to you'll produce crap. You might still produce crap, but without the base input cost you're kinda guaranteed to produce crap. Starving is not a good condition to write in.
1
1
Let's work out what a minimum-wage person in CA might be able to do. Min wage is $13. At 40h/week that's $520. Assuming they balance their budget somehow, their cost-of-life is about $3/hour. If they write 300/words per hour in their leisure time, it's 1c/word input cost.
1
1
So by my cost/length curve, their natural chunk size is probably around 100 words if they want it to be good. They're best off tweeting rather than even targeting short blog posts of 300 words.
1
The social psychology there is that the leisure hours are going to be so full of the stresses and anxieties of life at that income level, and the leisure hours available in such low-energy, smaller sessions, their output at bigger chunk sizes is going to be abysmal.
Replying to
Of course there are outliers. People who've written great works in prison or while subsisting on food stamps or slaving away in gulags. But the very rarity of such cases tells you that it takes the force of enormous passion/creative vision to override the default sweet spot
1
4
tldr: there actually is a meaningful unit economics to writing, but most people are not aware of it because they rarely think in terms of production cost, only the realized market value... which for the median published public word is 0.
1
4
Probably less than 0 actually. Negative if you count all those people writing words on non-free servers and not making anything from it.
1
Note that this entire thread is about just the input costs of writing alone, assuming nothing very special in terms of research beyond just casual reading you would do anyway/living life and using that as fodder.
1
1
Special research costs that can be directly attributed to the writing itself: - Reading things you would not otherwise read - Archival research - Interviewing - Travel for on the ground data collection
1
Research costs that should NOT be attributed to writing are costs that can be attributed to a higher-value output. Ie research on a time machine to write about a time machine is not really research about the the writing, since the time machine is the more valuable output.