zlacker

I do a few AI projects and my number one rule is to use LLMs only to process requests and never to generate responses.