Why build?
What compels me to prototype time and time again?
What do I want to do with the things I’m building? What do I think they will accomplish?
I build them to try to capture the contours of what I sense is there. When scrolling through a popular thread, it feels like noise. When looking at the visuals that I’ve generated, it seems to make a bit more sense.
I built the above tool late the other night with Claude. You can play around with it yourself, although the process is a bit unfriendly, at the moment.
I guess that’s what I’m trying to do. I’m trying to make things make more sense. I’m trying to build periscopes that let people see beyond what’s immediately in front of them. I’m trying to build on-ramps to systemic, egregoric thinking.
But... when I dial it back, how do I align what I’ve done with what people want to do?
Most people don’t want to sit around running an algorithm against different data all day long. Or do they?
No, I think people want to:
Know and prove things
Think new ideas
Convince others
Make better decisions
Maybe they would want to use it to communicate better with others. But how? How much effort would that take? It seems like the effort required to run the algo on chats and then make communicative decisions is just too damn high.
Let’s take a step back and think about principles.
People are using Twitter to:
Express themselves
Market things
Explore ideas
Make friends
Coordinate
Get validation
Make ad money
To what end? Well, we communicate with others in a clumsy, two-way modality, and we get ideas, validation, and perspectival mirroring back. Typically we do this in a “vibes-based” way, interacting with Tweets one-off, in a way that either reflects our total integrated understanding, or some loud part. That’s...fine, but it’s messy. It’s clumsy. Maybe we say something we don’t mean. Maybe we miss the point.
What I’m actually trying to build is tools that keep us in the “sweet spot” — tools that help us:
Figure out what’s at stake and what’s really there
Align their own thoughts and positions relative to what’s around them
Stay focused and on track; filter out noise, filter in signal, continuously, in coherence with what they’re doing
These systems and tools are akin to reading glasses, metronomes, and air conditioners. The aim of these tools is to reduce epistemic noise, narrow semantic drift, standardize ontological connections. The idea is that, without tools and technology to help us, we are liable and likely to run amok in our thinking and communicating. But with the intellectual equivalent of lane keep assist and emergency collision avoidance, we can have a smoother, safer, and more rewarding time.
The idea is that we should ask ourselves, implicitly and explicitly, “What am I getting out of this? Where is this going? How am I engaging with this? Why do I care?”
Every interaction, every scroll, and every tap can be an opportunity to nudge ourselves closer to what we value and what we desire for ourselves. Each bit might take us closer to connection and communion with ourselves, and as a pleasant byproduct, with each other.
That’s the aim, anyway.
So, I’ll keep building. The tools I assemble might not see immediate widespread use, but their utility will become more clear with exploration, and putting a library of them together will make it easier to show people what I see when I look at the world. And maybe people will learn to navigate that new landscape with me. There’s only one way to find out!
If you want to try the cluster graph, I made a little walkthrough for the current iteration. I’m considering building a hosted version. I need to figure out how to handle LLM costs and message persistence. I guess I’ll bite that bullet when I come to it.