To the extent that different humans have incompatible values, the AI alignment problem is strongly unsolvable.
There are whole realms of theory about how to reconcile orthogonal values.
There are whole realms of theory about how to reconcile orthogonal values.