4 ms·

I still find most LLMS to be extremely poor programmers . Claude will often generate tons and tons of useless code quickly using up it's limit. I often find my

I still find most LLMS to be extremely poor programmers . Claude will often generate tons and tons of useless code quickly using up it's limit. I often find myself yelling at it to stop. I was just working with it last night. "Hi Claude, can you add tabs here.": <div> <MainContent/> <div/> Claude will then start generating MainContent. DeepSeek, despite being free does a much better job than Claude. I don't know if it's smarter, but whatever internal logic it has is much more to the point. Claude also has a very weird bias towards a handful of UI libraries that has installed, even if those wouldn't be good for your project. I wasted hours on shancn UI which requires a very particular setup to work. LLM's are generally great at common tasks using a top 5( popularity) language. Ask it to do something in a Haxe UI library and it'll make up functions that *look* correct. Overall I like them, they definitely speed things up. I don't think most experienced software engineers have much to worry about for now. But I am really worried about juniors. Why higher a junior engineer, when you can just tell your seniors they need to use Copilot to crank out more code

joseda-hg 2y ago
Assuming I know roughly what it will generate, I usually prepend my chats with previsions against this kind of thing "Add tabs here, assume the rest of the page will work with no futher modification, limit your changes so that any existing code keeps working" I also do stuff like "Project is using {X} libraries, keep dependencies minimal Generate a method takes {Z} parameters, return {Y}, using {A}, {B} and {C} do {thing}" I'll add stuff like Language version, frameworks or specific requests based on this, but then I just reuse the setup , So I like to keep the first message with as much context as possible, ideally separating project context from specific request

joseda-hg 2y ago

Assuming I know roughly what it will generate, I usually prepend my chats with previsions against this kind of thing "Add tabs here, assume the rest of the page will work with no futher modification, limit your changes so that any existing code keeps working" I also do stuff like "Project is using {X} libraries, keep dependencies minimal Generate a method takes {Z} parameters, return {Y}, using {A}, {B} and {C} do {thing}" I'll add stuff like Language version, frameworks or specific requests based on this, but then I just reuse the setup , So I like to keep the first message with as much context as possible, ideally separating project context from specific request