I agree with most of the points on the article although it doesn't really have much to do with vim (which the author acknowledges).
I too have often compared AI usage to vim in the past months: AI enthusiasts seem to spend most of their time tweaking their setup; creating custom agents, skills, mcp servers, writing custom tools for their agent to use, customizing their harnesses, loops and whatnot. Similar to how us vim users keep updating/rewriting our config. It makes you feel productive while you really aren't and while in the end it might help you better stay within the flow the author talks about it the article the gains are typically very minimal if one is honest with oneself.
Man I just stick with the defaults on vim. I want to be able to install vim on any server I'm SSH'd into and have it all work the same way.
I do use Neovim as an IDE but thats a bit different, as lazynvim does 90% of what I need and the rest is just tiny changes that improves the experience for me. I also use AI to change my lua files simply because its faster and I'd rather focus on programming.
I've been fighting the urge to keep tweaking the setup, and with local models, the desire is even more involved. I'm pretty close now to the following:
1. llamacpp with custom reasoning budget and reasoning message tailored to a specific agent.
2. opencode with tailored agent that has access to dynamic context pruning and a graph of the code base.
3. Because local models love to reason instead of working, the llamacpp budget stops the reason with a message that reminds them to use subagents and to prune the context
---
This is successful enough that I can make and expand greenfield projects into full blown tools. I've just built this morning a open document templating system that lets me build reports into reports and make them look nice with templates. What would have bloated up to ~1M context stays within the dynamic context prune's limits set 50k-85k; some of the limits are because of the 128GB GPU, but mostly because processing times go down the larger the context.
So yeah, it's pretty easy to want to keep tweaking this stuff. I'm happy though, I don't have to ever contemplate some cloud service changing my work flow because they need to squeeze $$$ out of my job.
7 comments
[ 0.24 ms ] story [ 11.6 ms ] threadYou mean, except for the spurious :w here and there...
I too have often compared AI usage to vim in the past months: AI enthusiasts seem to spend most of their time tweaking their setup; creating custom agents, skills, mcp servers, writing custom tools for their agent to use, customizing their harnesses, loops and whatnot. Similar to how us vim users keep updating/rewriting our config. It makes you feel productive while you really aren't and while in the end it might help you better stay within the flow the author talks about it the article the gains are typically very minimal if one is honest with oneself.
I do use Neovim as an IDE but thats a bit different, as lazynvim does 90% of what I need and the rest is just tiny changes that improves the experience for me. I also use AI to change my lua files simply because its faster and I'd rather focus on programming.
1. llamacpp with custom reasoning budget and reasoning message tailored to a specific agent.
2. opencode with tailored agent that has access to dynamic context pruning and a graph of the code base.
3. Because local models love to reason instead of working, the llamacpp budget stops the reason with a message that reminds them to use subagents and to prune the context
---
This is successful enough that I can make and expand greenfield projects into full blown tools. I've just built this morning a open document templating system that lets me build reports into reports and make them look nice with templates. What would have bloated up to ~1M context stays within the dynamic context prune's limits set 50k-85k; some of the limits are because of the 128GB GPU, but mostly because processing times go down the larger the context.
So yeah, it's pretty easy to want to keep tweaking this stuff. I'm happy though, I don't have to ever contemplate some cloud service changing my work flow because they need to squeeze $$$ out of my job.
Reads like a human, and then the more you read, the AI-fluff appears. Not X, Y blah blah blah.
You could have human written it with your words in half the words. Been more authentic.