It is unknown whether or not we should treat nonexistent people as moral agents (like people rather than like trees), but it's an interesting idea to consider.

If we do this, we should focus on non-personal preferences rather than personal ones, because we can satisfy infinitely more preferences that way.

This contradicts the way most people reason when they treat nonexistent people as moral agents.

However, there is a problem: we need to try and figure out the preferences of nonexistant people to see what treating them as moral agents implies.