LLMs do not produce repeatable results, by their nature.

Neither do humans.

Two people can give the exact same prompt to the exact same LLM and get different results.

No one cares.

This is like telling someone else to code something for you.

Exactly.