Use these short easy-to-type quips when a LLM gives you slop output:
- “slop” — The LLM will re-evaluate it’s last output and reword it… hopefully better
- “why not what” — The LLM vomited a mechanical summary when it should have focused on architectural decisions, product needs, experience, etc
- “y/n” — The LLM answered a yes/no question with gibberish instead “yes” or “no”