Is the Book of Mormon Divinely Inspired?

[Marcus]

IHAQ’S sigline:

1. Eye witness testimony is notoriously unreliable. 2. The best evidence for The Book of Mormon is eye witness testimony, therefore… 3.The best evidence for the Book of Mormon is a type of evidence that is notoriously unreliable.

IHAQ is talking SINGULAR...

What on earth are you talking about?

[MG 2.0]

Large language models are not good at logic. They are literally and exactly trained to say exactly this about any possible topic: the kind of thing that gets said about this on the Internet.

They make no effort at all to check whether the things that get said on the Internet make any sense. That's the whole point of this recent LLM breakthrough in artificial intelligence. Determining what it means to "make sense", in simple and precise enough terms to put into an algorithm, has turned out to be difficult. That is the rock on which the first phase of the quest for artificial intelligence foundered.

The new phase just abandons that goal, and takes the lowest common denominator of the Internet as its standard. The kinds of things people say, whether stupid or not? That's what these large models say.

That's why there's no chance at all for the current generation of AIs to be superhumanly intelligent in any really important sense. They might conceivably manage to be smarter than any single human, but they literally cannot be smarter than humans collectively, because collective humanity is all that they know.

And in fact they suffer badly from the problem that they cannot sift wheat from chaff, because their whole point is not to even to try to distinguish between wheat and chaff. They say the kind of things that people say on the Internet, about anything. On simple topics that have drawn a lot of attention, most of what gets said on the Internet is actually pretty smart. LLMs are Wikipedia squared, in this way.

On anything that isn't quite simple enough for a good consensus to have emerged on the Internet, LLMs are just useless. They give you the lowest common denominator of ignorance trying to pretend, because that's what the Internet gives.

Thank you for that. Would you give more ‘trust’ to the organization and output (general and even specific conclusions…all three of my references were generally, and in some cases specifically, in agreement) of LLM’s when you get similar (but not the same…is that important?) answers?

Do multiple witnesses matter in this instance or is the jury out?

Like you can fool some people some of the time but you can’t fool everyone all the time. Does this apply to both the inquiries and the outputs when accessing A.I. systems?

Seemingly there is something to be said when/if you have multiple witnesses, not only with people, but also with different A.I. systems and algorithms.

Regards,
MG

[MG 2.0]

IHAQ’S sigline:

IHAQ is talking SINGULAR...

What on earth are you talking about?

Are you kidding? Read my posts.

Regards,
MG

[Marcus]

IHAQ’S sigline:

IHAQ is talking SINGULAR...

What on earth are you talking about?

Are you kidding? Read my posts.

Regards,
MG

Where in your posts have you explained what "IHQ is talking SINGULAR" means ????????

[MG 2.0]

What on earth are you talking about?

Are you kidding? Read my posts.

Regards,
MG

Where in your posts have you explained what "IHQ is talking SINGULAR" means ????????

It’s rather obvious he is referring to singular eyewitness testimony and not independent and yet composite (positive) testimony given over time and space.

Singular vs. multiple. Make sense?

IHAQ:

1. Eye witness testimony is notoriously unreliable. 2. The best evidence for The Book of Mormon is eye witness testimony, therefore… 3.The best evidence for the Book of Mormon is a type of evidence that is notoriously unreliable.

Is he referring to ALL three witnesses over space and time or is he referring to the likelihood that ONE eye witness might be mistaken?

So you’re multiplying the ‘mights’ by three?

Regards,
MG

[Marcus]

Where in your posts have you explained what "IHQ is talking SINGULAR" means ????????

It’s rather obvious he is referring to singular eyewitness testimony and not independent and yet composite (positive) testimony given over time and space.

Singular vs. multiple. Make sense?

lol, no. You're not making sense at all. He wasn't referring to singular testimony.

IHAQ:

1. Eye witness testimony is notoriously unreliable. 2. The best evidence for The Book of Mormon is eye witness testimony, therefore… 3.The best evidence for the Book of Mormon is a type of evidence that is notoriously unreliable.

Is he referring to ALL three witnesses over space and time or is he referring to the likelihood that ONE eye witness might be mistaken?

So you’re multiplying the ‘mights’ by three?...

Are you serious? First of all, multiplying the probability of three independent events DECREASES the likelihood, and if there are three completely dependent events the probability would REMAIN THE SAME as any one of the events, so all you are saying is that in the entire universe of possibilities, the likelihood is the same or less.

Your nonsensical statements actually work AGAINST your argument, by definition decreasing the likelihood of your assertion.

[Physics Guy]

No, three LLMs agreeing on a subject like this doesn't add much. A little bit, but not much.

Simple logic with words that are often used casually, but whose strict meanings are more precise than that, is the kind of thing that LLMs don't do well. They'll have absorbed thousands of Internet examples of statements using the words in loose senses. Most of the people that made those statements knew very well that the words technically had stricter meanings than that. The LLMs don't have that kind of knowledge, however. Their whole existence is about not even trying to have that kind of knowledge. They just repeat what they've heard, on the Internet.

So imagine that you have hired a personal assistant: a young B+ student who knows some things, but doesn't know anything really well, and nonetheless is very keen to have this job of assisting you with their knowledge. They have a lot of chutzpah, and they can put up a good fight up to a point with glib phrases that sound like what someone who knew stuff would say. They're a good salesperson. They're the kind of person who gets the job, at least in plenty of places. Now they're with you.

That's a large language model A.I.. For some things, it's just great. It's not exactly intelligence.

[malkie]

No, three LLMs agreeing on a subject like this doesn't add much. A little bit, but not much.

Simple logic with words that are often used casually, but whose strict meanings are more precise than that, is the kind of thing that LLMs don't do well. They'll have absorbed thousands of Internet examples of statements using the words in loose senses. Most of the people that made those statements knew very well that the words technically had stricter meanings than that. The LLMs don't have that kind of knowledge, however. Their whole existence is about not even trying to have that kind of knowledge. They just repeat what they've heard, on the Internet.

So imagine that you have hired a personal assistant: a young B+ student who knows some things, but doesn't know anything really well, and nonetheless is very keen to have this job of assisting you with their knowledge. They have a lot of chutzpah, and they can put up a good fight up to a point with glib phrases that sound like what someone who knew stuff would say. They're a good salesperson. They're the kind of person who gets the job, at least in plenty of places. Now they're with you.

That's a large language model A.I.. For some things, it's just great. It's not exactly intelligence.

If (as sometimes seems the case) disinformation multiplies faster than accurate information, it may be that LLMs will become progressively poorer as sources that make sense on almost any topic.

[MG 2.0]

…multiplying the probability of three independent events DECREASES the likelihood…

Of what?

…and if there are three completely dependent events the probability would REMAIN THE SAME as any one of the events…

That doesn’t make sense in this context. One event. Three independent witnesses. They all stay true to their story independently in regards to the One Event.

…so all you are saying is that in the entire universe of possibilities, the likelihood is the same or less.

No. You’re not making sense again. The likelihood of all three witnesses telling the truth increases exponentially based on there being more than one witness.

There really isn’t a “universe of possibilities” as you would like to think

Your nonsensical statements actually work AGAINST your argument, by definition decreasing the likelihood of your assertion.

From what I’m seeing this is almost the perfect example of what I’ve been concerned with before in your responses. You twist things around in order to come up with conclusions that dovetail with a preset agenda.

To me, anyway, it seems rather transparent.

Regards,
MG

[MG 2.0]

No, three LLMs agreeing on a subject like this doesn't add much. A little bit, but not much.

Simple logic with words that are often used casually, but whose strict meanings are more precise than that, is the kind of thing that LLMs don't do well. They'll have absorbed thousands of Internet examples of statements using the words in loose senses. Most of the people that made those statements knew very well that the words technically had stricter meanings than that. The LLMs don't have that kind of knowledge, however. Their whole existence is about not even trying to have that kind of knowledge. They just repeat what they've heard, on the Internet.

So imagine that you have hired a personal assistant: a young B+ student who knows some things, but doesn't know anything really well, and nonetheless is very keen to have this job of assisting you with their knowledge. They have a lot of chutzpah, and they can put up a good fight up to a point with glib phrases that sound like what someone who knew stuff would say. They're a good salesperson. They're the kind of person who gets the job, at least in plenty of places. Now they're with you.

That's a large language model A.I.. For some things, it's just great. It's not exactly intelligence.

I still think that in the way I’ve been using A.I. on this board that it has mostly added rather than detracted to the relevant information which it has made available ‘on the fly’. I’m not opposed to then looking at actual source material that may act as a correction or buffer against false information and/or conclusions.

More often than not, however, I’ve seen simple rejections of A.I. rather than any kind of complex or relevant argument against the information presented.

Regards,
MG