use the ai to make highly customized personal software instead of paying a rent-seeker to make a little widget in a couple hours that they then ram with ads and spying to make a trickle of money off of forever.
Timeline
Post
Remote status
Context
6this also deprives the platform of 50% cut rent seeking.
this destroys the bad version of capitalism
@sun
dunno any software would wanna use personally could be made that way?
dunno any software would wanna use personally could be made that way?
@ageha I recently made a little static html page that very easily helps you pick a time for meeting between multiple attendees in different timezones. nothing exactly like this seemed to exist, and what did exist had ads and shit all over it
Replies
5
@whiteline @ageha no doubt in this case. tbh when I was a professional paid programmer I got complaints I was slow. never a bad programmer. but slow.
@whiteline @ageha yesterday I started ChatGPT on converting over a 2D art tool written in crusty D code, into a modern MacOS Swift app targeting Metal GPU and WebGPU acceleration. It's working great so far. It too a lot of preparation to make the LLM output good
@whiteline @ageha I was thinking about it and tbh it created a better final product than if I did it by hand. this is a self-own I know but facts are it did a great job
@sun @whiteline @ageha I've spent the past week+ using an llm to build a tool that acts as a canary to protect against credential theft from supply chain attacks (i.e. miasma) and helps build app armor profiles for the applications you use by watching audit logs, classifying and grouping events, and providing suggested rules that can be applied iteratively with explanations generated by a local llm server. It's not done yet, but the idea is to make mandatory access control something more people can take advantage of eventually. It's not done yet, but some of the biggest hitting features (like clustering/grouping events) have been working almost since day 1. Mostly now it's about tuning the rails so the llm doesn't hallucinate garbage suggestions, and I'm collecting all of the input, output, and finalized profile data so that I can try my hand at fine tuning a model specifically for this purpose.