Wrote a draft of an LLM policy for #SilverBullet. Not final yet, but interested in feedback. TL;DR: #NoAI
https://gist.github.com/zefhemel/077c0b62f54da2cbfcc09c13bfe88ec8
Discussion
Loading...
@zef I notice that when I use LLM to write some code, I then suffer from what the Dutch called "tunnel vision". It then hinders you from thinking if the code you wrote is complete or if something is missing. It is not directly laziness.
@mpiscaer yeah, this is probably not specific to LLMs. It's just natural to, when presented with a solution, consider that solution rather than starting from a blank slate.