The kind of writing you want to do seems to me to be affected by the star system. There are lots of aspirants, and very few that do well at it. Those would be your tier 5s.
Consider the difference between a teacher, and a writer of textbooks. If you are teaching 5 kids algebra, you need one teacher, and 5+1 copies of one text book. If you are teaching 10 kids algebra, still one teacher, and still one text book, with 10+1 copies. But when you need to teach 50,000 kids algebra, you need several thousand teachers - and still only need one textbook author.
So teaching doesn't pay very well - but lots of people can find a steady job as teachers. Whereas writing the textbook potentially pays more - but only for one person, and it's not a steady job.
Whatever you write can be read by 10 people - or 10 million people. It costs you the same to produce, either way -- except to the extent that you respond to comments. This makes you more like the textbook writer, and less like the school teacher.
OTOH, if I need something specific written - let's say a manual for the product I want to sell - you are a bit more like a teacher. I can't use any old product manual - it has to be specific to my product. But then you have to make writing products to order. Which may not satisfy the desire to write something you *want* to write.
My suspicion is that this is the key difference. It's not pure writing vs writing mixed with other skills; it's more like the difference between creating fine arts and creating advertising images - the same technical skills may be used for both.
(1) I think you overestimate the technical skills of technical writers. I've worked with full-time technical writers at a Fortune 500 tech company. They don't grok the thing they're writing about. Essentially they rely on software engineers to write documentation which they format. They attempt to make it pretty and clean up grammar and language. Sometimes they inadvertently change the meaning, so I have to review their work pretty carefully.
(2) Often I find that I'm correcting the output of technical writers for things that they should excel in: gramar, formatting and clarity.
I think that a highly skilled technical writer from both the technical and the writing side is probably quite rare. I also think that they are also undervalued, so as a result, we mostly skate along with bad technical writers.
I agree with you that there aren't many jobs for writers where writing is the job in itself. I also think that whatever pure "writing" jobs exist now are likely be replaced by AI pretty quickly.
That's not to say that writing isn't going to remain a valuable skill, but it's going to mean that, even more than today, writing is going to be a skill that's only really in demand in conjunction with some other skills.
One of the best writers I know is a lawyer. I suspect that being able to express complicated concepts in concise, easy-to-understand prose is a significant factor in making him a great lawyer. But he's a lawyer, not a writer.
Most of the best writing I encounter is in the context of fiction. It takes more than some well-written prose to make good fiction, and so I think writing is simply the technical aspect of the creative writer. I suspect that while one can exercise the creative muscle to get better at it, you need a create spark that's pretty rare.
The internet writers you list are all are using writing as a tool to express novel ideas. They are not being compensated for their writing, they're being compensated for their ideas. Some of them are really good at writing (Scott Alexander, Andrew Sullivan), but we read them for their novel ideas or perspectives. Not their ability to synthesize prose. I also read writers that are frankly terrible at writing (Tyler Cowen, Robin Hanson). They're just good enough writers to get their novel ideas on to the page.
For the most part, I don't think you've shared much in the way of novel ideas on this blog, but the best pieces you've written offer a novel perspective. Novel perspective can only get you so far, but it's a good start on the road to novel ideas. Novel ideas gets you to the next tier--not better argumentation, relatability, readability or likeability.
That last is entirely possible - if so, I'll be a good test case for it. I do wonder if you aren't typical-minding a bit, though, perhaps spurred on by the examples I chose. Hoel and Scott are both people appreciated by the small subset of everyone who thinks like rationalists like to; Sullivan is farther from that, Yglesias is farther from there, and so on.
The people who read my Obsidian Roundup don't give a crap about whether I'm a good writer. They're reading because I'm good at finding stuff they care about (think of how well newsletters like The Browser and The Prepared are doing), and because I'm good at articulating perspectives that the buy-my-courseware set of influencers aren't talking but I can talk about because I'm not trying to sell courseware (I'd make so much more money if I did, but I'd hate it).
I can probably manage about 500 words of content on GTD, max, but honestly I like Luhmann's quote best: “I never force myself to do anything I don't like. Whenever I am stuck, I do something else."
I find substacks to read by reading the comments in substacks I am already reading. If the person says something I find interesting, and they have a substack, I sign up for free. When I do so I usually get 4 more substacks recommended to me. I try them too, for free.
2/3rds of these new subscriptions get immediately cancelled. What is going on in that substack isn't something I am interested in. But if you present me something interesting to think about, reflect upon, whether I end up agreeing with you or not .... I pop for a paid subscription. If months go by, and no new thoughts happen, I stop paying, and usually stop reading altogether.
I am not paying you for 'your writing -- a commodity'. I am paying for new and interesting thoughts in me. (And new music to listen to, if you are a music substack.) I think there are a lot of people like me reading substacks, precisely because this is where we can get this sort of experience. The magazines that used to scratch this itch aren't doing it any more, mostly because they don't exist any more. But the itch is still there.
This information makes me anxiously go through the few subscribers of my blog: I know you have seen my Substack user, because Laura Creighton recently liked a comment I made on another Substack.
Then the question becomes: Was my blog so particularly bad that it was an exception to the rule that almost all substacks are worth subscribing to for a short time? *anxiety anxiety*
Please send me a note of feedback if or when you unsubscribe. Provoking new and interesting thoughts is what I try to do, so if I fail to do that it would be interesting to know in what way.
This feels a bit like whether Math is "a job." Not very many people get to do "pure math" all day, after all -- even math professors typically have to also teach, a lot of physicists also have to build stuff and run experiments. Usually it's doing math *for* something and the job requires other skills as well.
I think something similar is going on with childcare. It's a big thing here in Europe (I don't want to say "big business", because it's usually run either by the government, a church, or some other organisation-which-isn't-really-a-business). Nearly all children go to some daycare or preschool, so lots of people work with taking care of them. I'm not sure what they earn, but it isn't much. The usual narrative is that this very important task is chronically underpaid because it is predominantly performed by women, and everybody knows how women aren't valued by the patriarchy. I wonder if the dynamic mentioned in this post might not be a part of the reason. Most people are able to take care of little children. Many people want to. Thus, even if wages are barely above minimum wage, there will be enough willing workers that there's no need to pay more.
And yet society still relies un the assumption that anybody with fuctional biological equipment is capable of caring for a child. And most people do manage.
The best definition I can come up with (while still remaining simple, and while certainly still flawed) for "pure writing" in the sense you mean is: they can usually choose what it is they want to write about.
I like this. I think the flip side of this (for me) is that I can't always choose to write well about stuff I *don't* want to write about. I can usually make it work, but it's worse.
Yes, there's some interaction with quality there that's imperfect but still significant. "Pure writing" is better because it's what the writer wants to write about, and they want to write about it because it will be better when they do, and so on.
It reminds me of all the millions of people online who fancy themselves movie reviewers and write long (and often very good!) pieces about films they love, but never think about how the actual professional critics still have to go and watch *Paul Blart: Mall Cop* on a Tuesday and come up with something interesting to say about it.
In that sense, perhaps standing opposed to "pure writing" is "real writing," which is to say: the ability to make anything interesting.
There’s another form of writing that’s not a job, but provides huge value. Definition by real example:
“I am an engineer in a researchy group at a large company. My job is programming and/or technical research, but because I’m very good at writing lab reports, emails, job interview notes, and promotion packages, I get paid maybe twice as much. Writing isn’t my job, but it helps distinguish and reveal my job performance, making me much more successful.”
Sometimes writing augments or even multiplies other skills.
There is something a bit mathematically sketchy about describing yourself as typical for tier 3 and then directly afterwards saying that you are doing better than 99.9% of writers at tier 3. Doesn't the latter imply you are basically tier 4?
Anyways, very interesting article, thanks for writing!
Believe it or not, not so much mathmatically sketchy as poorly written. I've amended it, but what I was trying to say was "I'm in tier 3, which isn't very impressive. Still, it's better than most of everyone! I'm doing very well in a percentile sense."
There's two parts of writing--having an idea, and communicating it well. The latter can be useful without the former in all the contexts you mention--technical writing, marketing, etc. But it's not particularly uncommon.
Having an idea is what people come to you for; its more of a pundit or perhaps comedian role; the words themselves aren't terribly important except that they should stay out of the way of the idea, with the slight exception that an occasional well turned phrase is a value add, such as writers spawning in the library or a litany of whale puns.
This makes it all feel an awful lot like sports, especially the concept that only a miniscule percentage of writers are "making the team" financially off of writing alone.
Yes, people call these kinds of jobs tournament professions: One of the functions of the Internet is turning more things into tournament professions. For example, think about playing music before and after recording devices were invented. Or theater actors vs. TV/movies, etc.
I wonder how much the price paid is simply a function of demand for paid writing,? There's a limit to how much writing an individual is likely to consume of a particular type, and there's a lot of free stuff out there. Also a limit to how many people are likely to do the consuming in any particular niche.
These dynamics are very different for Windows 11 or Salesforce or Oracle.
So aren't you just seeing the intersection of supply and demand curves? There's something about visibility, too, for which Substack will help, but that's just a size-of-available-market thing that's already reflected in the demand curve.
Oh, for what it matters I'm almost completely sure it's something like that, if it's so that I am right in the first place. The confusion was more around the "am I right" dynamic.
There's also another dynamic which is driven by "most educated people think they can write", and again this isn't true of coding. This means that, while they can't pick out someone who can russle up a Merkle tree in a few lines of Python, they can look at a piece of writing and think "oh I could have done that, so we can find a random college English major who would do that work next time for less".
They're not wrong, at least in some sense: the Merkle tree either works or it doesn't, but the writing could be a little bit worse and nobody would care (meantime their stock options would go up).
I volunteer persuasive writing eg marketing emails as a type of writing which might be highly paid with not too much pre-requisite subject matter expertise?
The kind of writing you want to do seems to me to be affected by the star system. There are lots of aspirants, and very few that do well at it. Those would be your tier 5s.
Consider the difference between a teacher, and a writer of textbooks. If you are teaching 5 kids algebra, you need one teacher, and 5+1 copies of one text book. If you are teaching 10 kids algebra, still one teacher, and still one text book, with 10+1 copies. But when you need to teach 50,000 kids algebra, you need several thousand teachers - and still only need one textbook author.
So teaching doesn't pay very well - but lots of people can find a steady job as teachers. Whereas writing the textbook potentially pays more - but only for one person, and it's not a steady job.
Whatever you write can be read by 10 people - or 10 million people. It costs you the same to produce, either way -- except to the extent that you respond to comments. This makes you more like the textbook writer, and less like the school teacher.
OTOH, if I need something specific written - let's say a manual for the product I want to sell - you are a bit more like a teacher. I can't use any old product manual - it has to be specific to my product. But then you have to make writing products to order. Which may not satisfy the desire to write something you *want* to write.
My suspicion is that this is the key difference. It's not pure writing vs writing mixed with other skills; it's more like the difference between creating fine arts and creating advertising images - the same technical skills may be used for both.
(1) I think you overestimate the technical skills of technical writers. I've worked with full-time technical writers at a Fortune 500 tech company. They don't grok the thing they're writing about. Essentially they rely on software engineers to write documentation which they format. They attempt to make it pretty and clean up grammar and language. Sometimes they inadvertently change the meaning, so I have to review their work pretty carefully.
(2) Often I find that I'm correcting the output of technical writers for things that they should excel in: gramar, formatting and clarity.
I think that a highly skilled technical writer from both the technical and the writing side is probably quite rare. I also think that they are also undervalued, so as a result, we mostly skate along with bad technical writers.
I agree with you that there aren't many jobs for writers where writing is the job in itself. I also think that whatever pure "writing" jobs exist now are likely be replaced by AI pretty quickly.
That's not to say that writing isn't going to remain a valuable skill, but it's going to mean that, even more than today, writing is going to be a skill that's only really in demand in conjunction with some other skills.
One of the best writers I know is a lawyer. I suspect that being able to express complicated concepts in concise, easy-to-understand prose is a significant factor in making him a great lawyer. But he's a lawyer, not a writer.
Most of the best writing I encounter is in the context of fiction. It takes more than some well-written prose to make good fiction, and so I think writing is simply the technical aspect of the creative writer. I suspect that while one can exercise the creative muscle to get better at it, you need a create spark that's pretty rare.
The internet writers you list are all are using writing as a tool to express novel ideas. They are not being compensated for their writing, they're being compensated for their ideas. Some of them are really good at writing (Scott Alexander, Andrew Sullivan), but we read them for their novel ideas or perspectives. Not their ability to synthesize prose. I also read writers that are frankly terrible at writing (Tyler Cowen, Robin Hanson). They're just good enough writers to get their novel ideas on to the page.
For the most part, I don't think you've shared much in the way of novel ideas on this blog, but the best pieces you've written offer a novel perspective. Novel perspective can only get you so far, but it's a good start on the road to novel ideas. Novel ideas gets you to the next tier--not better argumentation, relatability, readability or likeability.
That last is entirely possible - if so, I'll be a good test case for it. I do wonder if you aren't typical-minding a bit, though, perhaps spurred on by the examples I chose. Hoel and Scott are both people appreciated by the small subset of everyone who thinks like rationalists like to; Sullivan is farther from that, Yglesias is farther from there, and so on.
The people who read my Obsidian Roundup don't give a crap about whether I'm a good writer. They're reading because I'm good at finding stuff they care about (think of how well newsletters like The Browser and The Prepared are doing), and because I'm good at articulating perspectives that the buy-my-courseware set of influencers aren't talking but I can talk about because I'm not trying to sell courseware (I'd make so much more money if I did, but I'd hate it).
> I'm not trying to sell courseware
"Buy Eleanor's Actually Another Rehash of GTD (see, less marketing spin!)" :'D
I can probably manage about 500 words of content on GTD, max, but honestly I like Luhmann's quote best: “I never force myself to do anything I don't like. Whenever I am stuck, I do something else."
A datapoint.
I find substacks to read by reading the comments in substacks I am already reading. If the person says something I find interesting, and they have a substack, I sign up for free. When I do so I usually get 4 more substacks recommended to me. I try them too, for free.
2/3rds of these new subscriptions get immediately cancelled. What is going on in that substack isn't something I am interested in. But if you present me something interesting to think about, reflect upon, whether I end up agreeing with you or not .... I pop for a paid subscription. If months go by, and no new thoughts happen, I stop paying, and usually stop reading altogether.
I am not paying you for 'your writing -- a commodity'. I am paying for new and interesting thoughts in me. (And new music to listen to, if you are a music substack.) I think there are a lot of people like me reading substacks, precisely because this is where we can get this sort of experience. The magazines that used to scratch this itch aren't doing it any more, mostly because they don't exist any more. But the itch is still there.
This information makes me anxiously go through the few subscribers of my blog: I know you have seen my Substack user, because Laura Creighton recently liked a comment I made on another Substack.
Then the question becomes: Was my blog so particularly bad that it was an exception to the rule that almost all substacks are worth subscribing to for a short time? *anxiety anxiety*
No, pure sloppiness. My bad. I seem to have missed your substack that time. Fixed now. :) Thank you.
Please send me a note of feedback if or when you unsubscribe. Provoking new and interesting thoughts is what I try to do, so if I fail to do that it would be interesting to know in what way.
Ok!
"Understands how many times to say “Fish” in an article about the sea."
This. My white whale.
SEOby Dick
This feels a bit like whether Math is "a job." Not very many people get to do "pure math" all day, after all -- even math professors typically have to also teach, a lot of physicists also have to build stuff and run experiments. Usually it's doing math *for* something and the job requires other skills as well.
I think something similar is going on with childcare. It's a big thing here in Europe (I don't want to say "big business", because it's usually run either by the government, a church, or some other organisation-which-isn't-really-a-business). Nearly all children go to some daycare or preschool, so lots of people work with taking care of them. I'm not sure what they earn, but it isn't much. The usual narrative is that this very important task is chronically underpaid because it is predominantly performed by women, and everybody knows how women aren't valued by the patriarchy. I wonder if the dynamic mentioned in this post might not be a part of the reason. Most people are able to take care of little children. Many people want to. Thus, even if wages are barely above minimum wage, there will be enough willing workers that there's no need to pay more.
> Most people are able to take care of little children.
It's amazing how unobvious this can be in modern society.
And yet society still relies un the assumption that anybody with fuctional biological equipment is capable of caring for a child. And most people do manage.
The best definition I can come up with (while still remaining simple, and while certainly still flawed) for "pure writing" in the sense you mean is: they can usually choose what it is they want to write about.
I like this. I think the flip side of this (for me) is that I can't always choose to write well about stuff I *don't* want to write about. I can usually make it work, but it's worse.
Yes, there's some interaction with quality there that's imperfect but still significant. "Pure writing" is better because it's what the writer wants to write about, and they want to write about it because it will be better when they do, and so on.
It reminds me of all the millions of people online who fancy themselves movie reviewers and write long (and often very good!) pieces about films they love, but never think about how the actual professional critics still have to go and watch *Paul Blart: Mall Cop* on a Tuesday and come up with something interesting to say about it.
In that sense, perhaps standing opposed to "pure writing" is "real writing," which is to say: the ability to make anything interesting.
There’s another form of writing that’s not a job, but provides huge value. Definition by real example:
“I am an engineer in a researchy group at a large company. My job is programming and/or technical research, but because I’m very good at writing lab reports, emails, job interview notes, and promotion packages, I get paid maybe twice as much. Writing isn’t my job, but it helps distinguish and reveal my job performance, making me much more successful.”
Sometimes writing augments or even multiplies other skills.
There is something a bit mathematically sketchy about describing yourself as typical for tier 3 and then directly afterwards saying that you are doing better than 99.9% of writers at tier 3. Doesn't the latter imply you are basically tier 4?
Anyways, very interesting article, thanks for writing!
Believe it or not, not so much mathmatically sketchy as poorly written. I've amended it, but what I was trying to say was "I'm in tier 3, which isn't very impressive. Still, it's better than most of everyone! I'm doing very well in a percentile sense."
I think he meant “by dint of being in tier 3, I am doing better than 99.9% of the total population of writers”.
“at” here means “given I am at” rather than “compared to those at”.
There's two parts of writing--having an idea, and communicating it well. The latter can be useful without the former in all the contexts you mention--technical writing, marketing, etc. But it's not particularly uncommon.
Having an idea is what people come to you for; its more of a pundit or perhaps comedian role; the words themselves aren't terribly important except that they should stay out of the way of the idea, with the slight exception that an occasional well turned phrase is a value add, such as writers spawning in the library or a litany of whale puns.
ETA: Speaking of new ideas, this wasn't.
This makes it all feel an awful lot like sports, especially the concept that only a miniscule percentage of writers are "making the team" financially off of writing alone.
Yes, people call these kinds of jobs tournament professions: One of the functions of the Internet is turning more things into tournament professions. For example, think about playing music before and after recording devices were invented. Or theater actors vs. TV/movies, etc.
I wonder how much the price paid is simply a function of demand for paid writing,? There's a limit to how much writing an individual is likely to consume of a particular type, and there's a lot of free stuff out there. Also a limit to how many people are likely to do the consuming in any particular niche.
These dynamics are very different for Windows 11 or Salesforce or Oracle.
So aren't you just seeing the intersection of supply and demand curves? There's something about visibility, too, for which Substack will help, but that's just a size-of-available-market thing that's already reflected in the demand curve.
Oh, for what it matters I'm almost completely sure it's something like that, if it's so that I am right in the first place. The confusion was more around the "am I right" dynamic.
Oh I definitely think you're right here!
There's also another dynamic which is driven by "most educated people think they can write", and again this isn't true of coding. This means that, while they can't pick out someone who can russle up a Merkle tree in a few lines of Python, they can look at a piece of writing and think "oh I could have done that, so we can find a random college English major who would do that work next time for less".
They're not wrong, at least in some sense: the Merkle tree either works or it doesn't, but the writing could be a little bit worse and nobody would care (meantime their stock options would go up).
I volunteer persuasive writing eg marketing emails as a type of writing which might be highly paid with not too much pre-requisite subject matter expertise?
I tend to think of a job as paying, and so for most writers, writing is not a job.