And this is why it's very important, in the case of a junior engineer, to use your "I'm just starting here" privilege to ask those stupid questions. Or you can be a very senior engineer who has an established reputation, and can get away with asking what sound like "stupid" questions just because people assume you know what you're doing.

We have an expectation that "smart people" should be able to quickly fill in gaps in lightly-explained systems. Sometimes this is good: when you're teaching people a new concept it's great if they can grasp it quickly and approximately. When you're describing the design of a complex system, you absolutely do not want people to make incorrect assumptions about the parts you're skipping over.

The worst example I've seen was learning that the security of an industrial control platform came down to the fact that the management software wasn't installed by default. The designers had assumed that "knowledge of a software library" was a valid access control mechanism. As the cherry on top, another engineer chimed in that the software was actually installed on the system anyway, just in a different location. It took a pile of incredibly "stupid" questions to surface this knowledge.