Writing code with AI is almost just writing the code yourself in a different programing language. You have to be just as tediously specific about everything you want the program to do and the AI often creates new bugs or even removes old features when implementing new ones.
I found that the only way to get any progress was to use multiple AIs (chatgpt/deepseek) to check each other's work and fix each other's bad code. At the end of the day it's still trial and error with stupid machines in tedious detail, like coding was in the first place.
There are probably better methods and AIs for coding, but from what I have experienced so far it's simply changing the syntax from code language to english.
@black_pilled Language-heads have such a huge head start with these; the response is only as good as the prompt.
@temulentorator I must admit I'm slowly getting it to do what I want, but if I was proficient in python I probably would've had this done a long time ago. It's not so much the specificity that I'm having a problem with, it just keeps making new bugs and breaking previous features for no reason.
@black_pilled @temulentorator To be honest, current programming languages are already good, powerful and intuitive enough for anyone capable to do any software in a timely manner.
The issue are the programmers: it requires reasonable IQ to be a good one.
They are trying to replace them with some shitty ai since the pajeets can't do it. It all comes down to the overall IQ decline.
