The answer, I'm afraid, is not very edifying.
We cheat.
We cast at least one verb into the active voice.
One way is to provide an indefinite subject such as "people" or "they" or "we":
What will happen when people start to use metallic elements, &c?
Another is to make the passive subject the active subject of the first verb:
What will happen when metallic elements start to be used, &c?
This is permitted with verbs like start,begin,continue,finish,stop which don't really signify an action but act as semi-auxiliaries to tell you the temporal "shape" of the action verb - what grammarians call its aspect.
Actions and tasks and projects can be "started" or "finished", but we speak of an object being "started" or "finished" only if it is being treated as a task or project: "I started my paper today" means "I started writing my paper".
The passive voice exists for a reason; as far as I can tell all languages have some way of creating a construction equivalent in semantics (and pragmatics) to the English passive.
At the semantic level, the passive voice operates on transitive verbs (=2 or more arguments; note that the subject is included as an argument) and "promotes" the transitive object to a subject. The meaning is equivalent semantically (barring idiomatic usages), as the subject of a passive construction is still the object of that transitive verb. Consider these two sentences below, where the event (=what happens to the object) is the same, in either active or passive voice.
1) John ate [the cake]
2) [The cake] was eaten (by John).
In either 1) or 2), this information is the same: there was a salient cake and it was eaten. However, note that the passive allows one to drop out the agent.
So, why would one use the passive voice over the active? Perhaps the writer/speaker wants to give attention to the object. In English the subject position is often where topics go, and so is considered the "important" element. Promoting the object to a subject via passive allows one to emphasize this. Another possible case: we simply don't know who the agent was or perhaps the agent doesn't really matter. The passive is also helpful for generalizing / avoiding assigning blame / being polite.
Consider these cases:
3) The trash can was blown over (by the wind).
4) John was really screwed.
5) That project got really messed up. (<= note how it doesn't blame anyone in particular)
In 3), it probably obvious what blew the trash can over (in general people don't go around blowing down trash cans, etc.). For 4), this begins to get idiomatic--we could try to attribute reason(s) why John is not in a good situation, but usually this is secondary to stressing that John is really screwed.
We can apply this reasoning to the example sentence you gave:
6) [Tonight's moon] can be seen from anywhere worldwide.
In this sentence, the fact that the moon is visible from anywhere in the world is being stressed, hence why the writer probably chose to use the passive. We really don't care who is seeing the moon.
You should use the passive when its the idiomatic way to talk about an event (e.g. the agent usually is implied or not important) or when you want to focus on the object of such an event, and therefore promote it to subject position via the passive construction.
Best Answer
In the US we typically say that classes are offered rather than given
The preposition will be by or at, not from.
What is offered is more likely to be courses than classes. Classes tends to be used more in secondary-school or non-academic contexts.
So:
But note: the passive is awkward in this instance. Other things being equal, English prefers to put “heavy” constituents—those with many words—at the end of a sentence rather than the beginning or middle, because the syntactic structure is most visible when the primary constituents are close together at the front. Compare the structures for the passive and active sentences:
In the passive sentence, that's a long run from Subject to Verb. In this case you could get around this by moving the “minor” adjunct to the end: