On the left, a graph of an export of my public Roam. On the right, a graph of my MD file based website, written in Obsidian. While I didn't copy/paste content between the two, I discuss roughly the same topics. What does this tell you about how I'm writing in each medium?
Alternative behavioral design question: What is it about the design and limitations of each app that might encourage these differences in hypertext writing style?
Other interesting thing: my public Roam has 40,370 words, whereas my public MD file based website has 36,031 words. I wouldn't be able to tell there was that much of a disparity from looking at the graphs!
Just to make this easier for #roamcult people who are hearing a lot about Algorithms of Thought and have no idea what people are talking about...
They’re essentially thinking workflows you standardize for yourself to reliably produce an outcome.
For example, I have a page in @RoamResearch for questions I like to ask new prospects. An algorithm of thought might be a repeatable process for coming up with the right questions for that specific prospect.
Alt example, I keep “lenses” for myself. Series of questions related to a concept allowing me to apply it to a new context. Lenses of failure states, feedback loops, expectations, etc.
I wonder what a social media or personal knowledge management app would look like if it had *excellent* user-oriented newsfeed management built in. Given an endless stream of information, how can we give the user the most control over what takes up their attention?
Newsfeeds are essentially a jumping off point for exploratory browsing. More in QT thread on that. People browse information without a goal and sometimes new goals arise. Essentially, scrolling through your newsfeed can give you new goals when you had none.
A key question also would be: what are the effective responses to each block of content in a newsfeed? Most social media has functions for liking, commenting, and saving for later. What else? I would love to see newsfeed management around content I saved for later.
Many apps have inbox management tools, like [snooze, leave unread, reply, archive] in email, or [schedule, sort, complete] in task managers. What "newsfeed management" tools could be given to users to enable them to make a never-ending stream of information personally valuable?
For people interested in applying BeSci to product design, this falls into a class of questions about Choice Architecture.
One question you could use to answer this: what are the user’s possible options in response to each card of information to train the algorithm favorably?
I’m loosely defining inbox management as how you handle a finite list of items to be process, and newsfeed management as how you handle an infinite list of items that can never be fully processed. I’ll expand on my thoughts on newsfeed management here but I want to hear yours too
I tweet a lot because I’ve already tweeted a lot and have to thread my thoughts together. If all of the things I’ve tweeted were in Roam I could just thread things there and I’d probably be a lurker here. Also I like attention.
Honestly though has anyone made a single player Twitter client? Tweet and search only, no news feed or notifications? I feel a bit of a compulsion to add onto threads sometimes but I get sucked into exploratory browsing.
Anyone else catch themselves opening Twitter, checking notifications to see if there’s anything to deal with, and then OPENING THE NEWSFEED? That’s a dangerous string of behaviors if you’re cutting down distractions. Two gateway behaviors! I just wanna write thoughts & thread
Yesterday I was given a demo of @codexeditor. Absolutely wacky, in all of the right ways. Everything is an entity on a graph that can have labeled relationships with other entities. The UX is basically creating a new desktop OS in your browser for managing all info on that graph.
When I say everything can be an entity, I mean it... text, images, video, audio, websites. Iian is leaning into workspace model, allowing saved arrangements of what you have on the screen to revisit later. Those are also graph entities you can link to.
Whether it will see broad success or not, I have no idea. It's the sort of thing where you can only wrap your head around it so much without using it. It's insanely ambitious in doing a million things, so we'll see what people will actually do with it
Reading @DellAnnaLuca’s excellent book “Ergodicity” in its Roam form right now. The book’s content is excellent, but that’s not the focus of this thread. Since it’s pioneering the Roam book form, I wanted to give some thought on how the transition from paper to Roam is handled.
First I’ll point out the positives. It is really cool having the full text accessible in my own database. Annotating directly within Roam allows me access to my full knowledge base, and being able to reorganize its structure as I see fit is a treat.
I love that he made certain Roam/css improvements to augment the reading experience. For example, the styling to allow for footnotes was an excellent choice. Also, having pages built for quick definition lookups is great.