hello world
I'll try to overcome the blank blog syndrome with some thoughts about the use of innovation.
A jazz musician1 once told me that innovations in music are incorporated because they extend the current paradigm. Innovations have to be incorporated in spite of novelty, not because of it.
Similarly, in science, the way of progress is to constantly push the boundaries of our current theories extending their scope. New frontier science still has to explain a ball rolling down a plane successfully.
In tech, there seems to be a different approach to innovation. One that looks to radically replace old stuff with new stuff. 'We are going to change the way people < insert user action here > '. Is a good selling point, but there is a value in monotonous2 incremental evolution.
This made me think about the current state of tech in general, and AI integrations in particular. The idea that just because something can be done it has to be done no matter its utility.
Situations where I encounter the maximalist hunger
- Installing apps on my phone: 'This app looks useful, I'll install it just in case'
- The ever growing
importsection anytime an LLM tries to solve a problem in the code - The VS Code package ecosystem: Similar to the apps in the phone, I'll just install a bunch of apparently useful stuff
Garbage collection strategies
- GrapheneOS notifies me when I haven't used an app in a while and revokes permissions given on install. This can be a trigger to remove it completely.
- Neovim: I'm far from a power user, but having the chance to build a text editor from minimal functionality up to something adapted to my needs feels right.
- Pi coding agent: There is a reason why it has been dubbed as the vim of coding agents. Same approach, from bare-bones to adapted.
Friction may be a feature of evolving systems, not a bug.