Interpolating causal mechanisms: the paradox of knowing more