n vs n+1 theory

The theory

When you are expected to do n, you are still learning it and probably not very competent. You still need guidance.
When you learn n+1, then you start to understand how n really works.
Example: after taking calculus, you might be able to do problems, but not apply it independently. Then, you take differential equations, and start to understand calculus much better. Maybe you can actually apply calculus yourself to some new problems. But differential equations are still hard after that course.
People should go beyond the bare minimum to achieve competence.

Don’t expect a person at level n to be able to do the level n things all alone: they still need very active support (unless you aren’t using the “sink or swim” strategy).
Example: Don’t expect a fresh PhD student to be able to mentor a masters thesis, even if the PhD student just finished their thesis.
Don’t assume that just because someone has their PhD, they can mentor other PhDs. And so on. Do these people even have enough full example of the PhD process? Are they prepared to handle unique situations? Can they prepare their mentees for what comes after the immediate goal?
See the levels in chapter Academic career ladder for discussion of what happens at each level.

Don’t rely on someone who has just accomplished n to supervise someone working on n.
Provide active “how to mentor others” support for people at all levels.