AI and LLMs · Senior · Question 1 of 75
When is an LLM the wrong tool for a feature?
- 1When the team has never used an LLM before in production
- 2When the feature needs to process any kind of text input
- 3When rules or a simple model solve it reliably and cheaper
Show the answer
Correct answer: When rules or a simple model solve it reliably and cheaper
Explanation
Deterministic logic, regex or a small classifier are faster, cheaper and testable. LLMs pay off for fuzzy language tasks where rules break down.