i've made maybe 20 personal LLM tools this year. 3 survived past the first week. not because the rest weren't useful, just wasn't willing to debug them when something broke.
Which is where the "emacsification" analogy breaks for me.
The reason people who like emacs write their one-off program in emacs is that it is an extraordinarily introspectable and debuggable programming environment. There is no "code, compile, run" loop - you just write code against the live running environment. Devoid of that fast feedback loop, writing code just isn't as much fun.
Well taken, but for MANY years, I copied and pasted elisp snippets off stackoverflow without understanding any of it. And to this day, there as many xkcd-style "space heaters" as there are emacs users. OP's point stands that LLMs make possible a new generation of quicker, dirtier hacks destined for the cruft heap.
I built a vibe coding replay/inspection tool https://vibe-replay.com/
I built a co-work for cursor, I haven't publish that yet
I also built something combined activity watch and screen pipe, that provide `what did I do` context to LLM, this one didn't end up very well, still an interesting exploration though
Not OP:
- One shotted script that I schedule through cron that sends me a message when a new one piece manga chapter releases.
- Also a simple script that moves my cursor one pixel every 30 seconds. Cannot disclose why I need this.
The reason people who like emacs write their one-off program in emacs is that it is an extraordinarily introspectable and debuggable programming environment. There is no "code, compile, run" loop - you just write code against the live running environment. Devoid of that fast feedback loop, writing code just isn't as much fun.