Chatgpt on the Book of Mormon's veracity

[Nofear]

8 hours ago, Robert F. Smith said:

AI can be bullied into saying almost anything.  I had an AI lie to me.  When I told the AI that, it apologized.

Yes. That is true. It is also true that I referred to video as "funny". It's not evidence.

Some bullying takes much less effort than others. If one asks the probability that Trump is an unwitting agent of Russian agendas vs is Biden (or any other president) an unwitting agent of Russian agendas, the former question takes much less effort to come to a positive answer than the latter.

[let’s roll]

Thank God, literally, that we have more tools available to us than ChatGBT 01 Pro was given to evaluate the BofM.

Also, wouldn’t AI tell us that if one of the authors of the BofM wrote that the development and exercise of faith was a vital purpose of the BofM then scholarly research “proving” that the book is an accurate record of an ancient people in the Americas is a detriment to that purpose?

Edited March 19, 2025 by let’s roll
finishing my thought

[Analytics]

5 hours ago, Benjamin McGuire said:

The only challenge that generative AI faces in this particular arena though is that it doesn't do original research. When we get to issues like Mormonism, bias in the training material is a source of bias that a ChatGPT cannot easily eliminate. Anachronisms as discussed by ChatGPT above is an example of the problem because of the fact that there are embedded in virtually all LDS apologetic material and all critical material a shared set of assumptions that are (I have argued) incorrect. The Book of Mormon is a 19th century text - either as an original work or as a translation - and so we have to differentiate between anachronisms that can be attributed to the translation layer (assuming that there is one) and anachronisms that can't be when we are trying to compare the two theories. ChatGPT doesn't flag this sort of thing because in the literature it is trained on, this question is generally ignored. This is not to say that anachronisms don't exist - merely that our discussions of anachronisms are generally engaging a range of assumptions that are never really spelled out (except in the unrelated specialist literature of translation itself - where I have only found one source that discusses the Book of Mormon in terms of translation theory), and so largely unavailable to ChatGPT when engaging this sort of question.

You are absolutely right that AI can only work with the data it has access to. If you ask it an extremely broad question—such as whether the Book of Mormon is 'true'—it will do its best to weigh the available information, but it might not adequately account for strong yet obscure arguments. However, AI is a tool that can be directed toward specific questions, allowing us to examine those arguments in detail.

Regarding your point about translation-layer anachronisms, this isn’t a new idea, and its relative obscurity may stem from its limited explanatory power. While it might account for a few anachronisms, it doesn’t provide a systematic framework for distinguishing between translation-driven and genuine historical anachronisms. More importantly, it doesn’t give us a clear, testable model for how the text could plausibly function as an ancient record. If we can’t use it to make meaningful predictions or refine our understanding of the text’s origins, its value as an explanatory tool is questionable.

[Benjamin McGuire]

5 minutes ago, Analytics said:

Regarding your point about translation-layer anachronisms, this isn’t a new idea, and its relative obscurity may stem from its limited explanatory power. While it might account for a few anachronisms, it doesn’t provide a systematic framework for distinguishing between translation-driven and genuine historical anachronisms. More importantly, it doesn’t give us a clear, testable model for how the text could plausibly function as an ancient record. If we can’t use it to make meaningful predictions or refine our understanding of the text’s origins, its value as an explanatory tool is questionable.

But, I think that this is my point. If the literature is focused entirely on explanatory tools, then this issue (which is, I think, an important issue) gets lost - both in the literature, and subsequently in ChatGPT. The idea of translation layer anachronisms isn't itself a particularly obscure idea. The problem is that Book of Mormon believers often conflate (and want to conflate) the text of the 19th century Book of Mormon with the contents of the alleged gold plates. And the critics are happy to comply with this desire - because it makes their argument that much easier.

[JVW]

I thought what the AI generated was amazing. Thank you for sharing it.