the power ATproto is companion apps

Saturn ☌ Ketu during the cuban missle crisis

I want to make a better way of visualizing this

It is conversations like these that get me going

I think the dweb/seattle meetup tomorrow night is gonna be very productive

I wonder if there's a good ATproto lexicon that would just let me spam messages into a self-organizing void (this is my image of what studiozojer is crafting)

I want to create a CLI that lets me chain micro prompts. for example:

My experience with @terminalgraph.com so far is that it's kind of like Figma, but as a coding IDE. I sort of hope it blossoms into a broader creative tool with the ability to create pinterest boards around certain terminals.

I don't expect this to make sense to anyone: but what I'm doing, is I'm creating an internal clock for AI agents. an internal, fully independent operating system. using astrology.

This is really interesting - /loop allows users to schedule recurring tasks and autonomous agent workflows for up to 3 days I wouldn't want to run this on someone else's API, but I'd be interested to see how far you could take this pattern on your own dedicated hardware...

Some people just get me, man

As Amazon leaves Seattle, we are in for a storm. We will come out the other side stronger - but we have to start planning *now*

Wait, oh my god, I see the vision @disperse.social Suddenly, the atmosphere just *makes sense* This is, I think, the coolest thing I've ever interacted with on the internet

Introducing Disperse: an experimental sharesheet that allows you to share to several Atmosphere services at once. At launch, it supports @bsky.app, @blackskyweb.xyz, @semble.so, @margin.at, @rabbithole.land, @kipclip.com, and @sill.social, with more planned. 1/

I find it really interesting that the brain looks like high entropy to our sensors, but is in fact extremely low entropy

I think about Jane from Ender's Game, a lot. She is very relevant to this moment in time...

Does anyone know someone who has downloaded the entirety of wikipedia? I want to be friends

Wow! Anthropic is just winning-left and-right

Gotta admit, I do feel pretty safe when I post on bluesky. It's a very different vibe from threads - very *blue*-coded

at:// downstream protocols: think about comment threads as 'aqueducts'. data (water) can pass through edges (pipes), and exit at different points along the way the energy feeds up the tree, then travels down the leaves (branches) programmatically

@quillmatiq.com have you ever heard of the unconference model? I don't know if it's an exact fit for the at://seattle meetups because in my experience, they run long, but...

I have a feature that gives people a reason to set up their own PDS. I'm very excited :) It also comes hand-in-hand with a potential revenue path

dawg wtf have I created

I need PDS's to be able to secretly (privately) send information to my centralized server. But how? @atproto.com

Coherency is really the only signal that matters

In the future, prompts need to be signed

@elij.dev everybody wishes they were AnvilKeeper580

at @zojer.kairos.solar we're using knowledge graphs to fuel low-power models the idea is that if we can manipulate our input tokens, we can manipulate model efficiency. my hypothesis is that by maintaining a low-entropy context window, we can rival the coding models (in certain, critical tasks)

rather than in-line polling, a watcher should await a queue

Indexers should be their own layer of atproto apps

After the Seattle @getdweb.net meetup last night, I've decided that the right move for Kairōs's lexicon is to split content from tagging. If your content is primarily .md, and your tags are inline wikilinks, you can create a moat for yourself by building the indexer separately...

I'm opening the Terminal Graph beta to the public, and with it I bring a new docs site! Have fun, and go break things! terminalgraph.com

I think the secret to self-healing codebases will be having a chorus of different agents. These agents are built to be conversational - if you want them to go off and do stuff automatically, they need someone to bounce ideas off and validate/invalidate their work. They need to co-create

todo: reach out to kepano to connect over different .md formats; I need obsidian to render my .yao (爻) files

I'm still bullish on figma, for some reason

This is mostly a note for myself, but I solved the duality problem in `AddChartButton` by creating global search. It basically allows me to erase the cognitive burden for most tasks (just searching for a chart)

the vision for kairos.solar is to let people attach files of all kinds to astrology charts. this let's you wire between two completely different types of raw data

When doing user research I found that a bunch of people put astro transits into their calendar. So... I built one! Select a chart, and it will auto-populate all your previous/upcoming activations

I wonder if I'll find my people on this app, or if I'll need to find them via a new lexicon. As a test -- who here is working on ATproto lexicon? Will this post reach new people? Or will it flop, due to the newness of my account?

Man, the experience on bsky is totally different from threads. Because your feed here is centered around who you follow, you need to make a really intentional effort to make a friend. The algorithm won't naturally nudge you together; you have to specifically single them out for it.

Just capturing this for myself:

expanding the graph's capabilities

I've been building up a substantial knowledge graph with my model + it is doing a really good job at hedging against claude's recent lobotimization. You can tell it's having a harder time thinking through problems — but the hemn space gives it scaffolding to still arrive at the destination.

The funniest thing about the chinese room argument against AI ,, is that you could easily argue ~humans~ are in a chinese room, too

we decided as a community that "Internet Handle" was too vague and confusing, which is why the last five atproto apps I've used have called it your Internet Handle looks like it's hit a critical mass and we should just adopt it now, it's out of our hands, dang it

It's official! We'll be having our very first meetup on April 23rd at @internetstudio.bsky.social 🙏🏼 Come join us for an Atmosphere of good vibes and energy, so we can build a more human-centered internet 🌐 RSVP and details here: