I hate writing long posts, but I also often have a lot of background material to cover before I'm able to express what I mean. So I've got to choose between either writing a short post that will only be understood by people who share my background, or a long post that will be accessible to all, at the cost of making it boring for those who already know the background.
In an ideal world, every publication that tries to expand on the frontier will also serve as an adequate entry point for newcomers to the field. Not in the sense that everyone should immediately be able to understand the novel contribution, but in the sense that a newcomer reading it will learn what the prerequisites are, and where they may learn them.
The standard way academia tries to handle this is by mandating a highly structured introduction where you reference foundational work related to what you're doing. But this is highly inefficient, and thoroughly unfun for the writer. Plus, it requires you to redo that introduction every time you wish to explain something new.
Personal knowledge management systems exemplify a nice alternative. When I'm writing notes to myself, it's essential that I don't waste words on trying to contextualise exactly where in my network of related ideas this fits in. Instead, I simply wrap prerequisite terms or statements inside double brackets [], and that automatically creates an internal link to where I've got that explained. Note titles are like APIs that let me import context I've already defined elsewhere.
But for this to work in a collaborative setting, it's important that readers don't have to open a new tab every time they wish to import context. As an alternative, notice how when you hover over this link to the Arbital wiki, it opens a box so you can quickly check the contents without taking your eyes off where you were in this post.
Thus, I would love if the EA forum had an interface for writing personal wikis in one's profile. I could then write up my prerequisites in cohesive entries, allowing me to write public posts faster, with fewer redundancies, and catered to a wider range of reader familiarity.
The EA forum already hosts its own wiki that everyone can edit, but I wouldn't dream of using it for my own concepts that are only relevant within my own paradigm. It does, however, mean that the facility is there, and it could require less development time to copy the design into profile pages.
Here's how it could look. (Arbitrarily using Lizka's profile for the example, sorry!)
Semi-public working environments
An additional benefit to this is that it introduces another rung on the ladder between "unpolished, private" and "polished, frontpage". I'm probably not alone in having a lot of ideas that I feel are too unprepared to make a public post out of, yet I would still like to be able to easily refer to them in discussions. Having a place for it integrated into something I already use helps nudge me to write them up without having to maintain my own blog.
Besides, we should encourage people to work with the garage door up. Too many people walk around with the illusion that "real" scientists only ever make well-measured statements and never make mistakes. The truth is more like, you cannot learn to produce anything valuable unless you're comfortable with taking your own crazy ideas seriously, and most of your ideas are misses that will look embarrassing in hindsight.
If you have someone you think is extremely talented at doing what they do (e.g. research), the most effective way to learn from them often isn't from their public presentations. If instead you could witness the methodological mess behind how they produce what they produce, it might help break some illusions, and you might find hidden patterns that reveal to you why it has to be that particular mess if the aim is to be effective.
As an example, let's say I'm writing up a new discovery, and "grokking" happens to be a prerequisite concept to understand my contribution. I could introduce my discovery with...
"The term 'grokking' was coined by OpenAI researchers in 2022 and it refers a particular pattern in a graph that occurs given the following circumstances..."
...or I could simply use the term while explaining my contribution, and that gives me an opportunity to link back to where the prerequisite is explained.
Take a look at how Gwern's website does hover-over popups with the option of pinning them, moving them, and even open up new boxes from within. It's the most beautiful implementation of this I know of. Andy Matuschak's Evergreen Garden notes also shows a neat way of doing it, and I highly recommend exploring them to learn an alternative to the linear kind of reading/writing most of us are unfortunately used to.