The main restriction would be to pick "important" predictions, whatever "important" means here.

I don't know if there is any useful algorithm here. The space of possibilities is vast, black swans lurk at the outskirts, and Murphy is alive and well :-/

You can try doing something like this:

• List the important (to you) events or outcomes in some near future
• List everything that could potentially affect these events or outcomes.

and you get your universe of "events of interest" to assign probabilities to.

I doubt this will be a useful exercise in practice, though.