you are viewing a single comment's thread
view the rest of the comments
[–] 1 point 2 days ago

There's none, you always need to validate its output yourself. The models lack the capability to understand if something it produce is wrong or malicious.

  • source
  • parent