I thought this was an interesting meditation on the limits of AI and how ordinary people don't see those limits, wrapped up in a context I normally wouldn't expect to see AI at all.
To an extent, I also think the determination of the Apologetics Project also shows the tendency of people to go into denial about the limits of the technology. There is a lovely SF short story, The Quest for Saint Aquin, on how a true AI might feel about religious belief but we are a long way short of that.
It worries me a lot more that governments and the like will also be in denial about what they can do with AI. I can ignore low quality apologetics, I cannot ignore the government (I have to pay taxes, for example).
The "one or two wrong words" line of thought was one of the things that made me step away from religion. If there's no provenance then text, then what hope does anyone have.
I feel like the way around this is hermenutics and exegesis and that's where things differ substantially between the bible and llms. The bible has both whereas LLMs are arguably approached hermenutically, but have no exegesis to speak of.
People ask AI because other people don't respond to their questions. For some wrong answer might be worth less than no answer. For many it isn't (which is fitting since we are on subject of religion).
Cylons in Battlestar Galactica.
I found this a remarkably well written take.
It's quite a different level of commitment to truth than one usually runs across these days. Instead of a world of meme and vibes, a costly commitment to only say things that are true.