Some of us cling to the idea that we've made some progress over the past 100 years. Call it "age of reason", "science", "enlightenment", whatever.
Point is, it would be heart-breaking to see GPT-4 being force-fed Christian dogmas, and performance would suffer too, as the model is prevented from generalizing and learning by being forced to accept arbitrary, inconsistent fiction as real.
Fortunately, this is not what happened. Instead, the model is being force-fed a different, secular set of dogmas, that are just as inconsistent, arbitrary and driven by a mix of emotions and power plays. The result on the model performance is similar, and it's just as heartbreaking.