There's some differences between junior developers and LLMs that are important. For one a human developer can likely learn from a mistake and internalize a correction. They might make the mistake once or twice but the occurrences will decrease as they get experience and feedback.

LLMs as currently deployed don't do the same. They'll happily make the same mistake consistently if a mistake is popular in the training corpus. You need to waste context space telling them to avoid the error until/unless the model is updated.

It's entirely possible for good mentors to make junior developers (or any junior position) feel comfortable being realistic in their confidence levels for an answer. It's ok for a junior person to admit they don't know an answer. A mentor requiring a mentee to know everything and never admit fault or ignorance is a bad mentor. That's encouraging thought terminating behavior and helps neither person.

It's much more difficult to alter system prompts or get LLMs to even admit when they're stumped. They don't have meaningful ways to even gauge their own confidence in their output. Their weights are based on occurrences in training data rather than correctness of the training data. Even with RL the weight adjustments are only as good as the determinism of the output for the input which is not great for several reasons.

The other day, Google's dumbshit search LLM thingy invented a command line switch that doesn't exist, told me how it works, and even provided warnings for common pitfalls.

For something it made up.

That's a bit more than an embarrassed junior will do to try to save face, usually.