"In other words, it's not so much what the AI is actually doing -- it's how people perceive the AI."
The same words given by two different sources can be taken wildly different is a known thing, especially in religion. It’s not even a little surprising that that applies to AI
Religion is a social construct wherein humans organize and participate in communal activities for the purposes of creating human connection/bonding and quelling anxieties through relinquishing perspective of control in life.
Trying to make an AI preacher is an exercise in trying to emulate the latter without the former. The latter is spiritualism: the beliefs a person creates/adopts to create a sense of 'belonging' in the world. The former is the aspect that takes spiritualism and turns it into religion, as religion is essentially the social commodification of spiritualism. Of course no one seeking religion is going to be satisfied with an AI preacher, because anyone seeking religion is seeking human connection. You could use an AI to discuss spiritual beliefs with people instead, but at that point you're not creating a preacher, but rather a chat-bot theologian.
This feels obvious. Like I get why Buddhism might’ve been plausible, but part of the point of sermons is that they come from someone who gets it. No matter how much an AI can understand Buddhism it can’t understand desire and how it feels to grapple with letting it go. Taoism feels like a worse fit as part of what it deals with is nature and an AI can’t understand what it is like to live in harmony with the world or how it feels to be discordant from it. Christianity is a toss up in there. AI doesn’t have a soul and I understand why that would turn off Christians as well as the fact that many sects focus on personal relationships with their god, but I also understand that many Christian sects are adapting new tools to assist in their devotion. I think that AI Bible study would probably be more appealing to them in the present.
Everything Science
it is the place for high quality scientific content that doesn't necessarily reference a peer-reviewed paper from the last 6 months.