2 ms·
I think we still need an LLM to enable the system as a whole to understand vague and half-baked human input. I can easily ask an LLM to write be a function in
by ben_w 3y ago
I think we still need an LLM to enable the system as a whole to understand vague and half-baked human input.
I can easily ask an LLM to write be a function in a random programming language, then feed the output to a compiler, and pipe errors from the compiler back to the LLM.
What doesn't work so well is typing "pong in java" into a bash shell.
This isn't a perfect solution (not even for small projects), but it does demonstrate that automated validation can improve the output.
- a-r-t 3y agoThis is what ChatGPT's Code Interpreter does (writes code in Python and then runs it to check for errors). I'm not sure if it's enabled for everyone yet though.