Position: Probabilistic Modelling is Sufficient for Causal Inference | Best AI papers explained | Podwise