Using AI for reading

[Graham44]

Originally Posted by leebase

The IP issues will be worked out. So I’ll skip that part.

And the ideas and talks have merit on their own independent of my ability to source them.

For example - I don’t know you. You are a stranger on the internet. Everybody here is a stranger to me, even after all these years.

You share an opinion, I consider it. Some people I’m drawn to their way of thinking, some I don’t really care for. It my cup of tea.

AI is no different. I argue with it quite a lot. I find such discussions valuable and entertaining. It never occurs to me to wonder about the authority because it has NO authority in my life.

Some topics might be more important than others about having a specifically true response. No one should take outputs from AI as truthful. They hallucinate. They tell you not to trust them.

That’s not the same thing as them not being useful. I have 42 years of reading the Bible without AI. I’m not worried about it giving me an erroneous response.

Well we are all different and I can respect you find it helpful. For me AI is just really a no go, but as I say we are all different

[leebase]

Originally Posted by Graham44

Well we are all different and I can respect you find it helpful. For me AI is just really a no go, but as I say we are all different

Yep. Just like some of us have no problem reading on an iPad and others gotta have eink.

I will remind people of the context. I’m nit here saying everyone should be book talking with an AI. The OP asked if anybody was doing so.

I do. I get a lot of joy out of it. Not every book. Certainly the Bible. It’s the best “read the Bible in a year” journey I have ever done. I do have Bible commentaries and such. My Bible app has them integrated with the Bible’s making it easy to find commentary on specific passages I’m reading. Being able to talk to an AI is an elevated experience. IMHO.

If I’m reading a sci-fi book or tv show, sometimes I get interested in the science and talk to the AI. I see it as a form of self talk.

[DNSB]

Originally Posted by leebase

Or, I could do what I am doing. I am reading the “actual Bible”. And the AI is trained in all those commentaries. And I don’t treat the AI as authoritative, nor would I treat any of those commentaries as authoritative. Nor are my Christian friends. Nor is my pastor or elders.

Are you reading the original languages (Aramaic, Hebrew & Greek) or one or more of the many translations?

[Quoth]

Originally Posted by leebase

Or, I could do what I am doing. I am reading the “actual Bible”. And the AI is trained in all those commentaries. And I don’t treat the AI as authoritative, nor would I treat any of those commentaries as authoritative. Nor are my Christian friends. Nor is my pastor or elders.

But the LLM is not trained in human sense. It doesn't reliably repeat what was scraped. It doesn't reliably tell sources. Look at free eSword / Bible time study resources. The LLM solution is total garbage (full of nonsense) compared to that.

Also the LLM makers refuse to make clear what their models are fed, how it's fed, degree of curation (likely none, so garbage has the same weight as experts) and how reliable it might be.

[Quoth]

Originally Posted by DNSB

Are you reading the original languages (Aramaic, Hebrew & Greek) or one or more of the many translations?

@leebase
And even if you do, you need real historical, religious and cultural context to understand it, that's written by human experts with decades of study.

See the various Talmuds (Gemara and Mishna). https://en.wikipedia.org/wiki/Jewish...s_on_the_Bible

See Commentaries (Christian equivalent of Talmud) such as Mathew Henry etc.
Jewish & Christian: https://en.wikipedia.org/wiki/List_o...l_commentaries

A lot of the best PD material is on Bible Time. Real search. All local once downloaded.

[rantanplan]

Originally Posted by leebase

The IP issues will be worked out. So I’ll skip that part.

What do you mean by that? LLMs by definition only know what they are trained on. Which currently is what you can find on the internet. LLMs cannot, and will never be able to generate truly new things, it can only reproduce from what you trained it. It's a remix machine, that you have to actively prevent from plagiarizing.

[salamanderjuice]

Originally Posted by Quoth

But the LLM is not trained in human sense. It doesn't reliably repeat what was scraped. It doesn't reliably tell sources. Look at free eSword / Bible time study resources. The LLM solution is total garbage (full of nonsense) compared to that.

Also the LLM makers refuse to make clear what their models are fed, how it's fed, degree of curation (likely none, so garbage has the same weight as experts) and how reliable it might be.

I mean, they do actually say what they are fed and how it's curated:

https://arxiv.org/pdf/2302.13971
https://arxiv.org/pdf/2505.09388

[Quoth]

Originally Posted by salamanderjuice

I mean, they do actually say what they are fed and how it's curated:

https://arxiv.org/pdf/2302.13971
https://arxiv.org/pdf/2505.09388

In general they don't and can you believe them? MS and Google have not been great for truthiness and transparency.

However it make no difference. It's only any use if you can separately verify it. It's mostly plausible only if you are not already an expert.

[arjaybe]

I read the bible cover-to-cover once. It was hard. The writing doesn't really sweep you along. I found it had a lot of self-justification, and a lot of similar stories repeated. Like the one where the god says, kill them and take their stuff because they're heretics. I guess it would be all right for a scholar who could burrow down into one one book or one verse and squeeze all kinds of meaning out of it, but as a cohesive story end-to-end, it doesn't measure up. BTW, when I did this reading there was no AI to consult, and no internet to look up guides.

[nabsltd]

Originally Posted by DiapDealer

AI (and all of its adopted--albeit semantically incorrect--terminology) is here to stay.

If AI on its own doesn't become profitable soon, then it might go away, at least as far as access the general population currently has.

For example, Google makes the vast majority of their money through advertising. Companies (and some people) who pay for their software products directly are a tiny percentage of the advertising income. Unless Google can sell their AI directly to people/companies, they won't make money with it, so they will likely remove the ability for users to interact with it.

The extra data Google gets from AI prompts that allows better ad targeting can't possibly make up for the vast amounts of money spent on AI infrastructure. Without direct payments for the service, AI as we currently use it will go away.

[j.p.s]

Originally Posted by Quoth

MS and Google have not been great for truthiness...

Umm, truthiness is generally not a positive attribute except maybe to those that engage in it.

[DiapDealer]

Originally Posted by nabsltd

Without direct payments for the service, AI as we currently use it will go away.

Dream on. The tech, the terminology, and the tempest it causes is going nowhere. Learn to deal with it or lose your (that's a rhetorical you, by the way) mind over it. Of course how we use it may (more likely will) change, but that's how tech always evolves.

[Quoth]

Originally Posted by arjaybe

I read the bible cover-to-cover once. It was hard. The writing doesn't really sweep you along. I found it had a lot of self-justification, and a lot of similar stories repeated. Like the one where the god says, kill them and take their stuff because they're heretics. I guess it would be all right for a scholar who could burrow down into one one book or one verse and squeeze all kinds of meaning out of it, but as a cohesive story end-to-end, it doesn't measure up. BTW, when I did this reading there was no AI to consult, and no internet to look up guides.

But it's many books by many people and almost none of it written like a story. No way can it be cohesive.
I suppose Gospel of Luke and Acts can be read together. Esther is somewhat like a story as are parts of some books. Some books that are like stories might be only Canonical for Roman Catholics.

Song of Songs is Erotic Love Poetry.

The "Guides" (commentaries) were cheaply available and in libraries before the Internet. But "Bible Time" and other eSword resources are now on the Internet to download for free and are copies of accepted paper material, rather than the shuffled AI content that can be scrapped from entertainment and satirical sites.

[Quoth]

AI in many forms is mostly reliable, making money and useful. The question is where are the LLM & Generative AI systems going? Building and running them is costing billions. The operators have said they can't afford to pay for copyright content. None of them (unlike other AI systems) makes a profit. Microsoft might be in the best position to make money at it unless people figure out how to cancel the subscription extra features.

Max Leibman
@maxleibman@beige.party
Microsoft announces Xbox Copilot. This new AI gaming companion will play and enjoy video games for you, freeing you up to focus on work and chores.