Vibing on the fly by having an LLM write functions during runtime

https://github.com/ramiluisto/cursed_vibing_on_the_fly

by matriisitulo • 7 months ago

1 1 7 months ago

I recently saw a [meme](https://www.reddit.com/r/AICompanions/comments/1ph8w96/devel...) about modern programmers implementing code in the spirit of

  def is_odd(num):
      return OpenAI.chat(f"Is {num} odd?")

To me this felt like it wouldn't scale very well, with custom prompts needed for every function. So I implemented a decorator that you can use to ask `gpt-5-mini` to implement your functions during runtime based on function name, docstring[Optional] and type annotations [Optional]. In practice it looks like this:

  from cursed_vibing_on_the_fly import ai_implement

  @ai_implement
  def check_if_number_is_prime(n):
      pass

To me this makes the code feel so much more dynamic and, dare I say, vibrant!

Vibing on the fly by having an LLM write functions during runtime

Related Stories