CareerCup

Adam and Eve are the only nodes in directed acyclic graph (DAG) with in-degree zero. Start from 'you' or Einstein, follow the child to parent pointers in a dfs like manner. Eventually you will find the special nodes 'Adam' and 'Eve'. Now run another dfs starting with 'Adam'/'Eve', this time following parent to child pointers. The first node for which one of it's child reports having you in it's sub-tree, and another child reports having Einstein in sub-tree, is "a" first common ancestor. This is where it gets complicated because of a lack of "first common ancestor" definition (partly because of possible weirdo relationships in the history). There could be multiple candidates in a DAG unlike trees.

First couple with a second child birth.

I was thinking as if it was the "lowest" common ancestor problem, which is actually defined on trees. Since we need two parents to get a child, the graph becomes a dag, and most of what I said above is sort of like some observations, but not really a true solution :)

Is this a Creationist model (Adam / Eve) or an Evolutionary model (First humans were descendants of apes)? :P

Either way, perhaps Iterative Deepening Search would be appropriate for this problem?
This would be done by having two independent searches (running in parallel or sequentially while alternating) marking each visited node and checking in each last level of every iteration whether there are any nodes marked, for such, you've discovered all the first common ancestors of two given nodes.

Interesting test cases would be:
- There is Inbreeding in the genealogy tree of one person.
- Two persons are actually the same one.
- Two persons have only one first common ancestor
- Two persons have multiple first common ancestors.
- One person is the first common ancestor of the other.
- There is no first common ancestor (in an Evolutionary model).