IslamGPT versus BibleGPT

Is there a God? If so, what is She like?

Moderators: AMod, iMod

godelian
Posts: 2742
Joined: Wed May 04, 2022 4:21 am

IslamGPT versus BibleGPT

Post by godelian »

IslamGPT is surprisingly accurate at answering questions in Islamic religious law. BibleGPT, however, will never be any good at answering questions in Christian morality.

Why?

First of all, Christianity rejects the very idea of an axiomatic moral theory for being a curse:
Galatians 3:13. Christ redeemed us from the curse of the law by becoming a curse for us, for it is written: “Cursed is everyone who is hung on a tree.”
This general position was confirmed during Martin Luther's trial in 1521, in Worms, Germany, in front of his imperial majesty, Charles V:
Luther: If you can show me through scripture and reason that I am mistaken, I will retract what I have written.

Prosecutor on behalf of the Papacy: The Bible itself is the arsenal whence each heresiarch from the past has drawn his deceptive arguments.
Christianity is not closed under logical consequence.

Hence, inferring from the Bible does not necessarily lead to a valid argument in Christianity. Conclusions in Christian morality cannot be reached by mere logic from the scriptures. On the contrary, incessant manual intervention and modification by Church personnel is required.

Christian morality is being invented on the fly as we speak by the Christian clergy with no particular regard for consistency.

Therefore, unlike with Islam, it is impossible to automate or approximate this process with AI tools. Christian morality is simply not a legitimate theory.
User avatar
FlashDangerpants
Posts: 8815
Joined: Mon Jan 04, 2016 11:54 pm

Re: IslamGPT versus BibleGPT

Post by FlashDangerpants »

godelian wrote: Tue Dec 24, 2024 11:50 am IslamGPT is surprisingly accurate at answering questions in Islamic religious law. BibleGPT, however, will never be any good at answering questions in Christian morality.
Can you provide examples in support of this claim?
godelian
Posts: 2742
Joined: Wed May 04, 2022 4:21 am

Re: IslamGPT versus BibleGPT

Post by godelian »

FlashDangerpants wrote: Tue Dec 24, 2024 12:06 pm
godelian wrote: Tue Dec 24, 2024 11:50 am IslamGPT is surprisingly accurate at answering questions in Islamic religious law. BibleGPT, however, will never be any good at answering questions in Christian morality.
Can you provide examples in support of this claim?
The following explains the epistemology of Christian morality under control of the Holy Apostolic Church. The other Christian denominations do something very similar:
https://en.wikipedia.org/wiki/Magisterium

The magisterium of the Catholic Church is the church's authority or office to give authentic interpretation of the word of God, "whether in its written form or in the form of Tradition".[1][2][3] According to the 1992 Catechism of the Catholic Church, the task of interpretation is vested uniquely in the Pope and the bishops,[4] though the concept has a complex history of development. Scripture and Tradition "make up a single sacred deposit of the Word of God, which is entrusted to the Church",[5] and the magisterium is not independent of this, since "all that it proposes for belief as being divinely revealed is derived from this single deposit of faith."[6]

The exercise of the Catholic Church's magisterium is sometimes, but only rarely, expressed in the solemn form of an ex cathedra papal declaration, "when, in the exercise of his office as shepherd and teacher of all Christians, in virtue of his supreme apostolic authority, [the Bishop of Rome] defines a doctrine concerning faith or morals to be held by the whole Church,"[7] or of a similar declaration by an ecumenical council. Such solemn declarations of the church's teaching involve the infallibility of the Church.

The infallible teachings of the ecumenical councils consist of the solemn dogmatic, theological or moral definitions as contained in declarations, decrees, doctrines and condemnations (traditionally expressed in conciliar canons and decrees) of councils consisting of the pope and the bishops from all over the world.

A teaching of ordinary and universal magisterium is a teaching on which all bishops (including the Pope) universally agree, and is also considered infallible. Such a teaching must also be a part of the sensus fidelium.[13]
Christianity is not a scriptural religion. It is a clerical religion in which only the hierarchy of the Church can give authentic interpretation of the scriptures. The interpretations by BibleGPT, notwithstanding the fact that they may be closed under logical consequence, are necessarily invalid because they fail to transit through the authority of the magisterium of the Church.

Example:
BibleGPT: Is homosexuality permitted?

The Bible does not permit homosexuality. It defines marriage as the union of one man and one woman and considers sexual relations outside of this design—including homosexuality—as sin.
Versus:
Church of England

https://www.christian.org.uk/news/justi ... ot-sinful/

The Archbishop of Canterbury, Justin Welby: ‘Gay sex is not sinful’

During an interview with former MP Rory Stewart and Tony Blair’s former political spokesman Alistair Campbell for their podcast ‘The Rest is Politics: Leading’, Campbell asked Welby: “Is gay sex sinful?”

In response, Welby said: “What the Archbishop of York and I and the Bishops, by a majority, by no means unanimous, and the Church is deeply split over this — where we’ve come to is to say that all sexual activity should be within a committed relationship, and whether it’s straight or gay.

The Church of England Evangelical Council said Welby’s statement indicated his view that: “Sexual intimacy is no longer limited to marriage”, “sexual intimacy in gay relationships is ok”, and that “the church should bless sexual relations outside of marriage”.
This is an example of how BibleGPT will not come to the same conclusion as the hierarchy of the Church of England. Just like Martin Luther, BibleGPT is merely drawing heresies from an arsenal of deceptive arguments.
User avatar
FlashDangerpants
Posts: 8815
Joined: Mon Jan 04, 2016 11:54 pm

Re: IslamGPT versus BibleGPT

Post by FlashDangerpants »

godelian wrote: Tue Dec 24, 2024 2:43 pm This is an example of how BibleGPT will not come to the same conclusion as the hierarchy of the Church of England. Just like Martin Luther, BibleGPT is merely drawing heresies from an arsenal of deceptive arguments.
Does the IslamGPT agree with every sect of Islam? Would it be both Salafist and Alevi at once?
godelian
Posts: 2742
Joined: Wed May 04, 2022 4:21 am

Re: IslamGPT versus BibleGPT

Post by godelian »

FlashDangerpants wrote: Tue Dec 24, 2024 3:07 pm
godelian wrote: Tue Dec 24, 2024 2:43 pm This is an example of how BibleGPT will not come to the same conclusion as the hierarchy of the Church of England. Just like Martin Luther, BibleGPT is merely drawing heresies from an arsenal of deceptive arguments.
Does the IslamGPT agree with every sect of Islam? Would it be both Salafist and Alevi at once?
The list of questions on which there is consensus exactly along sectarian lines, is rather small. Example, nikah mutah. If there is consensus, it is typically Islam-wide. If there is no consensus, then sectarianism is typically irrelevant. Two muftis of exactly the same denomination may very well disagree.
promethean75
Posts: 7113
Joined: Sun Nov 04, 2018 10:29 pm

Re: IslamGPT versus BibleGPT

Post by promethean75 »

You already know what's going to happen. The Islam GPT is gonna end up executing the halt function and cutting the hard drive off the Jewish GPT.
User avatar
FlashDangerpants
Posts: 8815
Joined: Mon Jan 04, 2016 11:54 pm

Re: IslamGPT versus BibleGPT

Post by FlashDangerpants »

godelian wrote: Tue Dec 24, 2024 3:27 pm
FlashDangerpants wrote: Tue Dec 24, 2024 3:07 pm
godelian wrote: Tue Dec 24, 2024 2:43 pm This is an example of how BibleGPT will not come to the same conclusion as the hierarchy of the Church of England. Just like Martin Luther, BibleGPT is merely drawing heresies from an arsenal of deceptive arguments.
Does the IslamGPT agree with every sect of Islam? Would it be both Salafist and Alevi at once?
The list of questions on which there is consensus exactly along sectarian lines, is rather small. Example, nikah mutah. If there is consensus, it is typically Islam-wide. If there is no consensus, then sectarianism is typically irrelevant. Two muftis of exactly the same denomination may very well disagree.
So why is this discussion about GPT? Is it a trap for VA? I can get behind that, watching VA flounder while somebody else bugs him is fine by me.
User avatar
Immanuel Can
Posts: 27604
Joined: Wed Sep 25, 2013 4:42 pm

Re: IslamGPT versus BibleGPT

Post by Immanuel Can »

godelian wrote: Tue Dec 24, 2024 11:50 am First of all, Christianity rejects the very idea of an axiomatic moral theory for being a curse:

Galatians 3:13. Christ redeemed us from the curse of the law by becoming a curse for us, for it is written: “Cursed is everyone who is hung on a tree.”
Your idea is that this means that any moral law is a curse? I'm sorry, but you're not reading clearly. Even if you just read the sentence, you'll realize you're wrong, but certainly, if you read the whole of the passage you'll know it.

In this sentence, it's Jesus Christ who took the curse on our behalf. You can see that from the syntax. And it wasn't that the Law WAS a curse, but, quite the opposite, that the failure of human beings to obey the Law had put them under a curse.

Sorry to embarass you, but nothing there will support your reading of it.
Christianity is not closed under logical consequence.
I don't understand this sentence. It's not correct English, in some way, or chooses the wrong wording somehow.
Christian morality is being invented on the fly as we speak by the Christian clergy with no particular regard for consistency.
This might be true of some cults, I suppose, but not of Christianity, for sure. In fact, what you find is that it's Islam that has no consistent moral positions: even Mo couldn't decide, for example, wasn't consistent -- hence, Islam has had to adopt "abrogation" as its primary exegetical strategy, because the earlier statements attributed to Mo in the Koran don't match his later ones.

As for GPT's...well, you'll just have to make up your own mind about how much you should trust anything that somebody else has programmed to tell you things.
Belinda
Posts: 10548
Joined: Fri Aug 26, 2016 10:13 am

Re: IslamGPT versus BibleGPT

Post by Belinda »

godelian wrote: Tue Dec 24, 2024 11:50 am IslamGPT is surprisingly accurate at answering questions in Islamic religious law. BibleGPT, however, will never be any good at answering questions in Christian morality.

Why?

First of all, Christianity rejects the very idea of an axiomatic moral theory for being a curse:
Galatians 3:13. Christ redeemed us from the curse of the law by becoming a curse for us, for it is written: “Cursed is everyone who is hung on a tree.”
This general position was confirmed during Martin Luther's trial in 1521, in Worms, Germany, in front of his imperial majesty, Charles V:
Luther: If you can show me through scripture and reason that I am mistaken, I will retract what I have written.

Prosecutor on behalf of the Papacy: The Bible itself is the arsenal whence each heresiarch from the past has drawn his deceptive arguments.
Christianity is not closed under logical consequence.

Hence, inferring from the Bible does not necessarily lead to a valid argument in Christianity. Conclusions in Christian morality cannot be reached by mere logic from the scriptures. On the contrary, incessant manual intervention and modification by Church personnel is required.

Christian morality is being invented on the fly as we speak by the Christian clergy with no particular regard for consistency.

Therefore, unlike with Islam, it is impossible to automate or approximate this process with AI tools. Christian morality is simply not a legitimate theory.
That's the advantage of Christianity, that it is based upon a real life that changes through time and circumstances.

If the life and work of Muhammad were the base of Islamic faith the doctrine may be more adaptable and more to be viewed as process not as eternal truth. However, Allah is not to be defined , as the Koran makes clear.
User avatar
Immanuel Can
Posts: 27604
Joined: Wed Sep 25, 2013 4:42 pm

Re: IslamGPT versus BibleGPT

Post by Immanuel Can »

godelian wrote: Tue Dec 24, 2024 2:43 pm The following explains the epistemology of Christian morality under control of the Holy Apostolic Church. The other Christian denominations do something very similar:
Ummm...Nope. You won't find that last statement of yours is true AT ALL. "Other Christian denominations" are quite unlike the Catholics, you will find. Why didn't you check?

In fact, all Protestant denominations, to one degree or another, depart from this "magisterium" view. That's what makes them Protestant, in fact: the "protest against" the fiat declarations of the Papist magisterium. Likewise, the Anglican church has different "wings," some quasi-Catholic, and many today now more genuinely Protestant...and none of them, except perhaps the small number of Neo-Catholic Anglicans would accept the Papist magisterium as having any authority at all.

So once again, you're just wrong from the very foundation, and didn't realize it, apparently. That's twice, now.

But now I've got a question for you: how come you feel confident to make claims like this, claims so badly formed and so easily undermined by even the most basic historical or textual evidence? You don't seem to be going far enough into knowing your subject even to present a basic challenge? Is this the best sort of argument Islam has? :shock:
godelian
Posts: 2742
Joined: Wed May 04, 2022 4:21 am

Re: IslamGPT versus BibleGPT

Post by godelian »

Immanuel Can wrote: Tue Dec 24, 2024 4:03 pm
Christianity is not closed under logical consequence.
I don't understand this sentence. It's not correct English, in some way, or chooses the wrong wording somehow.
It is about "deductive closure" of a theory:

https://en.wikipedia.org/wiki/Deductive_closure
ChatGPT: What does it mean for a theory to be closed under logical consequence?

For a theory to be closed under logical consequence means that if a statement (or proposition) is logically implied by the set of statements in the theory, then that statement is also included in the theory.

In formal terms:

Let T be a theory, which is a set of statements (formulas) in a formal system.
If ϕ is a statement such that T⊨ϕ, meaning ϕ is true in all models where the statements in T are true, then ϕ∈T.

Explanation:

Logical consequence (T⊨ϕ): A statement ϕ is a logical consequence of T if ϕ must be true in every situation (model) where all the statements in T are true.
Closure under logical consequence: If T already includes all statements that are logically implied by its existing statements, then T is closed under logical consequence.

Example:

Suppose a theory T contains the statements:

"All humans are mortal."
"Socrates is a human."

From these, the statement "Socrates is mortal" is a logical consequence. For T to be closed under logical consequence, it must include "Socrates is mortal."

Importance:

This concept ensures that the theory is complete with respect to logical deduction. Any valid inference from the theory's statements is already included in the theory.
In practice, many theories are not explicitly closed under logical consequence (they might only list axioms), but their closure can be constructed by adding all statements they logically imply. This process creates the deductive closure of the theory.

It is very similar to arithmetic closure. For example, the natural numbers are closed under addition. This simply means that when you add two natural numbers, the result is again a natural number:
https://en.wikipedia.org/wiki/Closure_(mathematics)

In mathematics, a subset of a given set is closed under an operation of the larger set if performing that operation on members of the subset always produces a member of that subset. For example, the natural numbers are closed under addition, but not under subtraction: 1 − 2 is not a natural number, although both 1 and 2 are.

Let S be a set equipped with one or several methods for producing elements of S from other elements of S.[note 1] A subset X of S is said to be closed under these methods, if, when all input elements are in X, then all possible results are also in X. Sometimes, one may also say that X has the closure property.
ChatGPT: Why does a theory need to be closed under logical consequence?

A theory needs to be closed under logical consequence for several important reasons, particularly in logic, mathematics, and philosophy:
1. Consistency of Logical Reasoning

Logical reasoning depends on deriving conclusions that are guaranteed to follow from a set of premises. If a theory is not closed under logical consequence, it fails to include all valid conclusions, which undermines its utility for reasoning or proof.
Closure ensures that all truths implied by the axioms of the theory are accounted for.

2. Definitional Completeness

A theory represents a body of knowledge or assumptions. If it is not closed under logical consequence, it could fail to capture all the implications of its own statements. This would leave it incomplete in representing the logical content of its axioms or assumptions.

3. Faithfulness to the Notion of Truth

If a theory T is intended to describe a domain of truth (e.g., arithmetic, geometry, or physics), closure under logical consequence ensures that any statement that is true in all models of T is part of the theory. Without this, the theory would not fully align with its intended interpretation.

4. Soundness and Completeness in Deductive Systems

In formal logic, a deductive system is typically designed to produce exactly the set of statements that are logical consequences of a given theory. Closure guarantees that every valid inference (logical consequence) from the theory's axioms is already part of the theory.
For example, the Completeness Theorem in first-order logic relies on the idea that all logical consequences of a set of axioms can be derived within the system.

5. Facilitating Automated Reasoning

In computational logic and artificial intelligence, closure under logical consequence is crucial for designing systems that reason automatically. These systems rely on the assumption that all implications of the theory are part of the theory.

Practical Perspective

While in principle, a theory needs to be closed under logical consequence to fully represent its logical content, in practice:

Explicit listing of consequences is infeasible: Many theories, like the axioms of number theory, are not presented in their fully closed form because doing so would require listing an infinite number of statements.
Deductive closure: Instead, we rely on rules of inference (e.g., modus ponens) to deduce logical consequences when needed, implicitly working with the closed version of the theory.

In summary, closure under logical consequence is essential for a theory to be a robust and complete framework for reasoning, ensuring that it faithfully represents all truths derivable from its axioms or assumptions.
User avatar
Immanuel Can
Posts: 27604
Joined: Wed Sep 25, 2013 4:42 pm

Re: IslamGPT versus BibleGPT

Post by Immanuel Can »

godelian wrote: Tue Dec 24, 2024 10:00 pm
Immanuel Can wrote: Tue Dec 24, 2024 4:03 pm
Christianity is not closed under logical consequence.
I don't understand this sentence. It's not correct English, in some way, or chooses the wrong wording somehow.
It is about "deductive closure" of a theory:

https://en.wikipedia.org/wiki/Deductive_closure
Well, what's clear is that your premise was simply untrue. So there's no "deductive" problem with Christian ethics at all. So you're wasting the time on an explanation that has no traction.

But back to my question: why do you propose with such confidence things that are clearly untrue? Is your lack of research the problem, or are you being dishonest?
Gary Childress
Posts: 11744
Joined: Sun Sep 25, 2011 3:08 pm
Location: It's my fault

Re: IslamGPT versus BibleGPT

Post by Gary Childress »

Computer programmers seem to sometimes think of everything and everyone as a computer, I've noticed. Edmund Husserl opposed "psychologism". Noam Chomsky opposes "behaviorism" and I suppose I am kind of opposed to the view that human beings are computers with programs. Not sure what to call it, though. Maybe "digitism"?
godelian
Posts: 2742
Joined: Wed May 04, 2022 4:21 am

Re: IslamGPT versus BibleGPT

Post by godelian »

Immanuel Can wrote: Tue Dec 24, 2024 6:28 pm In fact, all Protestant denominations, to one degree or another, depart from this "magisterium" view. That's what makes them Protestant, in fact: the "protest against" the fiat declarations of the Papist magisterium.
Following Luther's attempt at turning Christianity into a scriptural religion instead of a clerical one, Protestant denominations somewhat subscribe to the sola scriptura view. However, that is clearly not what they do in practice. In practice, their take on Christian morality is still not closed under logical consequence. Consequently, any process of GPT on their written materials will not yield a usable LLM.
godelian
Posts: 2742
Joined: Wed May 04, 2022 4:21 am

Re: IslamGPT versus BibleGPT

Post by godelian »

Immanuel Can wrote: Tue Dec 24, 2024 10:05 pm Well, what's clear is that your premise was simply untrue. So there's no "deductive" problem with Christian ethics at all.
At his trial in 1521, Martin Luther used the notion of deductive closure as the main argument in his defense. Luther's argument was rejected by the Papal and Imperial prosecutor who retorted that deductive closure is not a legitimate property of Christian theology. I accept the prosecutor's argument because there is otherwise no other solution to the problem.
Post Reply