I hope this is a topic to which you will return. Here’s an AI benchmark which ranks foundation models based on their support for Christian values: gloo.com/flourishing-hu…. It is worth considering whether model-makers should pick target religions, and whether competition between religions should play-out as competition between such models.