Partial pooling of interactions

Just saw a blog about partial pooling of interactions:

https://statmodeling.stat.columbia.edu/2004/10/26/partial_pooling/

Any update in terms of modeling and implementations about the suggestion?