I read the blogging advice of a few different people. It seems like the best blogs do both steps.
Stage 1: First they write because they want to solve or understand some important problem in their life that they're curious about.
Typically somewhere here an assumptions gets made, that a problem you care about solving is also a problem many others care about solving.
Often this writing can actually be private. You might get a little bit of feedback if you write publicly though.
You probably won't get a large audience even if you write publicly while you're still on this step.
Stage 2: Then they write about this in a way that other people will read.
This means picking a target audience (MOST IMPORTANT), picking the right communication medium for that audience, writing only surprising not obvious stuff for that audience, using humour, using ingroup signals and all sorts of other tactics.
Advice almost every blogger agrees on
Everyone agrees that marketing/advertising speak is stupid. Write something so good that people spontaneously share with their friends, not because they're being paid to share it.
Write lots and lots of stuff.
Don't (publicly) write things you believe are false. Selectively writing some true stuff and omitting other stuff is okay.
People whose writing advice I looked at (random list)
Paul Graham
Naval Ravikant
Noah Smith
Eliezer Yudkowsky
Scott Alexander
Linchuan Zhang
For most of my posts, I am still in stage 1, not stage 2. I am writing because I want to figure out some problem for myself, not because I want to communicate to a large audience. I might (or might not) get around to stage 2 at a later date.
This means I mainly want feedback on whether my ideas are correct and important.
I might be slightly less interested in feedback on how to communicate my ideas better, although I am also open to that.
(For a few posts, I might explicitly flag that someone else is a target audience for that post.)
2026-01-03
Target audience varies depending on the document.
Unless specified otherwise, target audience is my future self and anyone else who shares similar curiosity as me.
Some documents start with a target audience declared explicitly.
Effectively persuading anyone else of my ideas takes a significantly different approach than just recording them for my future self.
If writing to my future self, crisp world models and terse vocabulary is generally a good idea.
If writing to teach others, often sharing ideas and arguments is not enough.
A more effective method is to teach via "exposure" and "immersion", by providing them a large amount of input from groups that already accept my ideas as true. (This is similar to Krashen's ideas around language acquisition, I'm extrapolating it more broadly to cultural acquisition here.)
If writing to persuade others, I am probably running a simulation of a representative of the target audience in my head.
Humans do this by default, we imagine representatives of various groups and simulate them in our head. This is possible even if we've never met an actual representative of that group. This is very useful neurological machinery. It also contributes to prejudice, as one can end up forming strong conclusions based on incomplete information, and later updating these beliefs can be painful.
Subscribe
Enter email or phone number to subscribe. You will receive atmost one update per month