Over the past weeks I’ve had a few experiences on Semble that point towards an interesting broader pattern: bookmarks combined with social networks, open protocols and intentional design, can be much more than bookmarks - they can be sensors!

Each case began with the simple act of me saving a card to Semble.

1.

The first was Brennan Kenneth Brown’s "What have note-taking PKMs accomplished, really?". It looked interesting, I added it to my reading list, and then I did what everyone does with saved links. I forgot about it.

A few days later a notification came in: @zafarali.me had connected a card in my library. They added a new card, ”Note-Taking and Personal Knowledge Management”, and linked it to the Brennan Kenneth piece with an opposes relation, plus a note explaining that the new piece is a helpful rebuttal.

A Semble notification reading "zafarali.me connected a card in your library," with his note beneath it. Below, two article cards side by side: unattributed.cc's "Note-Taking and Personal Knowledge Management" on the left, and brennan.day's "What have note-taking PKMs accomplished, really?" on the right, joined by an arrow labelled OPPOSES pointing from the first to the second.

So I read the rebuttal first. It was thoughtful and included a lot of information about the original article. Someone else's thirty seconds of curation reorganized my reading queue. Another friend who saved the same article as I did, also got the notification from @zafarali.me’s connection and similarly re-prioritized. I later revisited Brennan’s article and my engagement with it was significantly improved by virtue of having read the response beforehand.1

So those thirty seconds compound in an interesting way - since connection notifications are sent to everyone who saved the card, that single connection could alert not just me, but anyone with the card in their library.

2.

The second was Timothy Gowers' "Thoughts about the Leiden Declaration", it looked like it had a nuanced discussion around human aspects of research in the age of powerful AI. I saved it and added it to my far-too-long reading list. I still hadn’t gotten around to it, when a few hours later Rafael Batista connected it with an explainer connection to a short but helpful summary by Ted Underwood:

A Semble notification: Rafael M Batista connected "Thoughts about the Leiden Declaration" to a Bluesky post with an EXPLAINER relation. Below, the connected card shows Ted Underwood's post: "Thoughtful and relevant outside math. A partial summary: Gowers thinks it's important to sustain a human mathematical culture, but is unconvinced by the Leiden Declaration's attempt to do that by reaffirming human ownership of..." — quote-posting tachikoma, whose post shared the Gowers blog post.

Underwood's post is a quote-post of a post sharing the blog. The odds of me finding this on ephemeral feeds are slim, but by attaching a “sensor” to the blog post, I was made aware of it, courtesy of the Semble curation network. In this case, the connection provided me with a useful gist of the blog post that enriched my understanding of it even before I get around to reading it.

3.

A few days ago the White House Office of Science and Tech Policy (OSTP) released "Science: A New Golden Age", a long document laying out the administration's plans for American science. It’s a complex document that is genuinely hard to “wrap your head around”: dense, consequential, and drawing both sharp criticism as well as real enthusiasm. Documents like these are important sites for collective sensemaking: they are places where we can help each other unpack hidden assumptions, ask the right questions, improve our critical thinking and gain a more comprehensive understanding of relevant perspectives.

In cases like this, discussions around documents and other media are scattered across many blogs and social media platforms that do not talk to each other, and thus far there has been no systematic way to collect them.

With Semble though, we’re starting to see a new way we can approach this challenge: the Semble page for any URL provides an anchor for attaching context about it from around the web. The Semble page for the White House document already has 13 connections made by 4 contributors. Connections bring in otherwise fragmented content - a blog post, a post on Twitter, posts on Bluesky. Typed relations (supports, opposes, helpful, etc) help navigate different perspectives. It isn’t hard to imagine AI bots supporting larger scale synthesis and aggregation, though we will continue centering human trust and curation.

Sensors, not just bookmarks

The framing I keep returning to is that sembling a link (saving it to Semble) is less like bookmarking it and more like dropping a sensor on it.

Bookmarks are typically associated with hoarding behavior. It’s all too common to feel mild guilt upon saving an article: 218 items on my to-read list, how many will I ever really get to?

Meme titled "THE BOOKMARKOOOR" showing a cartoon of a person at a computer, surrounded by text labels depicting compulsive bookmarking behaviors: "Has 2983 bookmarks, never looks at them," "Saves every thread," "bookmarking this," "I'll use this in my next project," "this is gold," "will add this to my second brain," "I'll come back to this when I have time," "can't believe they're giving this up for free," and "saving this for later." The image humorously depicts someone who habitually saves online content with the intention of revisiting it but rarely does.

Semble changes this. While a bookmark is just for storage, a sensor also listens for you. Your Semble card becomes a sensor on that link so that whatever the network learns about it can find its way back to you: who else saved it, what supports or critiques it, what it should be read alongside. Adding the sensor might even help you decide whether to read the piece or not, or in which order, as the first example demonstrated. A sensor can shorten your reading list, which is not something a bookmark has ever done for anyone.

Closing thoughts

On one hand, three anecdotes are just three anecdotes. But we think they provide interesting early evidence for what we call the circular attention economy. Herbert Simon framed attention as a scarce resource, consumed by information and then gone, like fuel burned. A linear attention economy of overloaded individuals. Our reframe is that consuming information can also produce something - attention traces. And when they are shared, for example as as a card on Semble or a post on Bluesky,

[these] public products of human attention are like ant pheromone trails, crucial signals for shaping collective attention. They represent a more circular attention economy at work, where one person's consumed attention becomes a signal that guides others' attention allocation.

In all three cases, we saw shared attention traces doing the work. @zafarali.me had already read both PKM pieces; Ted Underwood had already read Gowers and Rafael made the connection between the two explicit on Semble. The reading happened regardless and the only marginal cost was the few seconds it took to record the trace. And those traces, which would normally have evaporated as “creative exhaust”, were recorded by Semblers and “recycled” via Semble, in the process meaningfully reallocating my attention. And my friend's, plus anyone else holding the same card.

Attention spent once, recycled several times over.

Join the circular attention economy by dropping some sensors or making some connections in Semble!

This is the first in a series of posts on new sensemaking patterns we spot on Semble & the broader Atmosphere. If you notice patterns, feel free to add them to this collection:

Patterns of Sensemaking (by Cosmik Network) — Semble
Field notes and examples related to new patterns of sensemaking on Semble or the broader Atmosphere. Open to anyone
https://semble.so/profile/cosmik.network/collections/3mrul7yyj6q2b