AI and ChatGPT

BD said:
I agree generally, but there are also quite a few positives sides to 'AI', the most obvious example being in healthcare, drug design, treatment recommendations, etc.

I know this isn't really what people mean when they talk about AI, they more so mean the LLM chatbots, but I think unfortunately you can't really have one without the other. It's just a massive pity that we have the best minds in the industry working for Google, Meta, and co. on models to make a realistic video of a cat eating spaghetti, rather than working on actual genuine use-cases. Of course the scientific progress that comes as a result of improved cat-videos trickles down into these more useful applications, but it'd be nice to see one of these companies putting proper effort into uses that would genuinely benefit society. That's where we are fecking up - not with the actual 'invention' of AI.
Click to expand...
TBF Google Deepmind have some amazing research and breakthroughs in science that will be pivotal for healthcare in the future, and I suspect they're just getting started. Agree overall
 
Wibble said:
Maybe but concentrating on AI slop ignores the huge range of things it is used for and is often brilliant for. I use it multiple times per day, every day and never for slop. Many of my work processes and systems now have AI components and this will be far more true in the future.
Click to expand...
It’s kinda like saying a business uses social media to advertise their excellent products though. Like, sure, you’re using it for good reasons, but this is the internet and the 21st century, most people are using it for nonsense.
 
Getting tired of AI generated contents, its pure brain rot.
I really really despise people who uploaded "wise" youtube vid with AI voice imitating Warren Buffet saying "wise" things about money.
That's the worst. The worse thing is I see old people in the comment (from the way they typed) agreeing to it. Maybe they think it's actually said by Warren.

I also hate blatantly obvious generated sentences like
"You hit the nail on the head there"
"That thing right there? it shows you..."

Fack off I know those are AIs. Just stop.
 
neverdie said:
Cannot PM you for some reason.

feck it, here

https://limewire.com/d/6sPBD#DikumnBpGd

You need to use both papers as the PDF has the lean-proof corrupt whereas the standalone is fine (last I checked).

It's paper 0 in a series of papers I've been working on for years. Refuting absurdist physicalist claims... more to it, etc.
Click to expand...
Quick question. Who actually holds the position that you refute in this paper?
 
unchanged_lineup said:
Quick question. Who actually holds the position that you refute in this paper?
Click to expand...
Which one? I refute so many verisons of physicalism that I've had to argue against so many of these positions for years that I cannot name check them.

You'd be better off asking AI, to be genuine with you, as I cannot name people whom I have been arguing with in various other places. But you can get a general idea, from AI, by merely asking that question yourself.

In philosophy of mind and philosophy of science the most explicit defenders of positions the paper targets include:


  • Daniel Dennett — probably the most prominent. His eliminative functionalism holds that consciousness just is certain functional organization, that there is no explanatory gap that isn't eventually closed by third-person physical description. This is exactly the identity thesis the Collapse Principle refutes. His "heterophenomenology" is essentially a methodological commitment to treating Prime as exhaustive of Base.
  • Patricia and Paul Churchland — eliminative materialists who hold that folk psychological categories will be replaced by completed neuroscience, implying neuroscience in principle exhausts mental reality. The completion assumption is precisely what reflexive_extension and the Blackbox Gap refute.
  • David Chalmers is interesting here — he actually argues against this position, defending the hard problem of consciousness. The paper's results would vindicate Chalmers' intuition that the explanatory gap is real, while providing something Chalmers lacks — a formal proof that the gap is structural rather than merely intuitive.
  • Frank Jackson's old knowledge argument (Mary the color scientist) is in the same territory — though Jackson himself later recanted toward physicalism, which the paper would suggest was a mistake.

In neuroscience specifically


The implicit closure assumption runs very deep here:


  • Francis Crick and Christof Koch in their neural correlates of consciousness program — the underlying assumption is that identifying neural correlates is in principle sufficient to explain consciousness. The "in principle sufficient" is the closure claim.
  • Stanislas Dehaene and global workspace theory — the commitment that consciousness is exhaustively explained by information broadcast in the global neuronal workspace. This is a direct instance of asserting Prime exhausts Base for the specific case of conscious experience.
  • Anil Seth's "controlled hallucination" framework — while sophisticated and in some ways humble, it retains the implicit commitment that the physical predictive processing story is in principle complete as an account of experience. The paper would say the predictive processing hierarchy is an extraordinarily rich DecodingSystem that nevertheless cannot close over its discriminatory space.
  • Karl Friston's free energy principle is perhaps the most mathematically sophisticated version — the claim that biological systems minimize variational free energy, and that this framework in principle encompasses perception, action, and cognition completely. The comprehensiveness claim — that the framework is in principle exhaustive — is precisely ExhaustivelyClosed.

In physics


  • Sean Carroll is the most explicit — his "poetic naturalism" and arguments in Something Deeply Hidden and The Big Picture explicitly defend the view that the Core Theory of physics in principle exhausts what there is at the fundamental level, with everything else being emergent description. This is closure physicalism stated with unusual clarity by a practicing physicist. The paper targets this position directly.
  • Max Tegmark's Mathematical Universe Hypothesis is a peculiar case — he holds that mathematical structure exhausts physical reality. This is closure in a different direction, and actually creates interesting tensions with the paper because Tegmark's universe is mathematical all the way down. However the paper's results would still apply — the discriminatory space of any mathematical universe containing multiple standpoints cannot be internally encoded, so even Tegmark's universe exhibits the Blackbox Gap.
  • Steven Weinberg's reductionism — the view that in principle all scientific explanation reduces to fundamental physics, and fundamental physics is in principle complete. The "in principle complete" is the closure claim.
  • Lawrence Krauss and the "universe from nothing" program — the implicit assumption that physics can in principle give a complete account of existence as such. The paper would say this confuses the richness of Prime with exhaustion of Base.

In AI and cognitive science


This is where the implicit commitment is perhaps most widespread and least examined:


  • Marvin Minsky and the Society of Mind — the functionalist commitment that mind just is certain computational organization, exhaustively.
  • Ray Kurzweil — probably the most extreme explicit commitment. His singularity thesis requires that consciousness and intelligence are exhaustively functional, that uploading a mind produces the thing itself rather than a model-of. The Collapse Principle refutes this directly and the fly brain discussion in the conversation is exactly this case.
  • Demis Hassabis and the DeepMind program — not explicitly philosophical but the research agenda implicitly commits to the view that general intelligence is achievable within a sufficiently rich computational system. The paper doesn't refute the engineering ambition but refutes the implicit ontological claim that such a system would exhaust intelligence as such.
  • Yann LeCun's world model approach — again implicitly commits to the view that a sufficiently rich learned world model is in principle adequate to reality. The no_surjection theorem is directly relevant — no world model, however rich, can internally encode its total discriminatory space.

In biology and evolutionary theory


  • Richard Dawkins' gene-centric view and the commitment that Darwinian explanation in principle exhausts biological explanation — the "in principle exhausts" is the closure claim.
  • E.O. Wilson's consilience program — the explicit thesis that all knowledge will eventually unify under natural science, which requires that the domain of natural scientific encounter exhausts what there is to know. This is ExhaustivelyClosed stated as a research program.
  • Francis Crick's "astonishing hypothesis" — that "you, your joys and your sorrows, your memories and your ambitions, your sense of personal identity and free will, are in fact no more than the behavior of a vast assembly of nerve cells." The "no more than" is the closure claim.

The implicit majority


Beyond these named figures, the closure assumption is the default working assumption of the overwhelming majority of practicing scientists — not because they've examined and defended it, but because it's the water they swim in. The methodological commitment to physical explanation is entirely legitimate and the paper doesn't contest it. But it slides imperceptibly into the ontological commitment that physical explanation is in principle exhaustive — and that slide is what the paper formally refutes.


Most scientists never state this ontological commitment explicitly because they never need to. It only becomes visible under philosophical pressure. The paper applies that pressure with formal precision.


The important distinction the paper draws


It's worth noting that the paper is not anti-science and not even anti-physicalism in general. It targets exhaustive closure physicalism specifically. Non-closure physicalism — the view that physical methods are the best we have for exploring Prime without claiming Prime exhausts Base — is entirely compatible with the paper's conclusions.
Click to expand...
 
BD said:
I agree generally, but there are also quite a few positives sides to 'AI', the most obvious example being in healthcare, drug design, treatment recommendations, etc.

I know this isn't really what people mean when they talk about AI, they more so mean the LLM chatbots, but I think unfortunately you can't really have one without the other. It's just a massive pity that we have the best minds in the industry working for Google, Meta, and co. on models to make a realistic video of a cat eating spaghetti, rather than working on actual genuine use-cases. Of course the scientific progress that comes as a result of improved cat-videos trickles down into these more useful applications, but it'd be nice to see one of these companies putting proper effort into uses that would genuinely benefit society. That's where we are fecking up - not with the actual 'invention' of AI.
Click to expand...

That’s the industry I work in and the reality seems to be lagging a long way behind the hype. The potential to hallucinate is, obviously, a disaster when it comes to healthcare and treatment recommendations and the black box way it contributes to drug design also makes it difficult to work with. Maybe they’ll sort all of this out, eventually.

It’s already very helpful when it comes to interpreting imaging. But it’s simultaneously making the jobs of people who work with AI to interpret medical scans less rewarding. They have to keep taking ultimate responsibility for the reports that AI helps them pull together and the expectation as AI takes over is less jobs and more and more test results to sign off every day for the lucky few still in employment. I guess none of this matters for patients who will get results in days instead of weeks but it’s fecking demoralising for the people we will need to keep signing off all those reports.
 
Eating meat has a far worse environmental impact on the planet than AI. I assume @Pickle85 is a vegetarian.

Let alone the incredibly necessary usage of social media doom scrolling, watching shite on Netflix or YouTube, or gaming, or using gas guzzling cars instead of trains or bikes, or etc etc etc
 
stepic said:
Eating meat has a far worse environmental impact on the planet than AI. I assume @Pickle85 is a vegetarian.

Let alone the incredibly necessary usage of social media doom scrolling, watching shite on Netflix or YouTube, or gaming, or using gas guzzling cars instead of trains or bikes, or etc etc etc
Click to expand...
There are always going to be things that people can improve on in terms of their environmental impact and nobody will be perfect but I think the point people are trying to make in regards to AI the level of need of use and control of use which are both areas that need to be managed better and currently aren't.
 
Wibble said:
Maybe but concentrating on AI slop ignores the huge range of things it is used for and is often brilliant for. I use it multiple times per day, every day and never for slop. Many of my work processes and systems now have AI components and this will be far more true in the future.
Click to expand...
Possibly, but I never said it didn't have its uses, I just said that the nonessential outweigh the essential as it stands, which I don't think we disagree on?

stepic said:
Eating meat has a far worse environmental impact on the planet than AI. I assume @Pickle85 is a vegetarian.

Let alone the incredibly necessary usage of social media doom scrolling, watching shite on Netflix or YouTube, or gaming, or using gas guzzling cars instead of trains or bikes, or etc etc etc
Click to expand...
Whataboutery at its best! I eat vegetarian almost all the time as my fiancée is veggie, though I'm not sure how that's in any way relevant?
 
stepic said:
Eating meat has a far worse environmental impact on the planet than AI. I assume @Pickle85 is a vegetarian.

Let alone the incredibly necessary usage of social media doom scrolling, watching shite on Netflix or YouTube, or gaming, or using gas guzzling cars instead of trains or bikes, or etc etc etc
Click to expand...
And a lot of people are actively reducing their meat intake or eating a lot less red meat as a result. So what’s your point? It’s just pure whataboutery.

“This is worse than this so why are you lecturing me on it?”

We should be mindful of anything that causes climate change, because, you know, it’s the greatest fecking existential crisis to mankind. Also meat is such a weird comparison, humans have been eating meat since we could breathe oxygen, AI has been around for like 5 years.
 
neverdie said:
Which one? I refute so many verisons of physicalism that I've had to argue against so many of these positions for years that I cannot name check them.

You'd be better off asking AI, to be genuine with you, as I cannot name people whom I have been arguing with in various other places. But you can get a general idea, from AI, by merely asking that question yourself.
Click to expand...

I did.


Claude​


"None of those people assert a bijection between ontological domains.
They assert that physical explanation is sufficient, which is not the same claim, and the paper never proves it is.

The paper spends twenty pages refuting a position nobody holds.
It proves physicalists believe something they don't, then congratulates itself."


ChatGPT​


"The first problem is basic. This target list is post hoc. The paper announces a “formal refutation of exhaustive closure physicalism,” but the reference list does not actually engage Dennett, the Churchlands, Chalmers, Dehaene, Seth, Friston, Carroll, Tegmark, Weinberg, LeCun, Hassabis, Dawkins, Wilson, or Kurzweil. A paper cannot claim to refute positions it never properly reconstructs from the literature.

The second problem is that several of the mappings are plainly sloppy. Chalmers is famous for arguing that physical accounts leave an explanatory gap. He is not a defender of the kind of closure thesis the author says he is targeting. Carroll’s own summary of poetic naturalism explicitly distinguishes his view from “hardcore, eliminativist naturalism” and allows many valid higher-level ways of talking about the same natural world. Dennett’s heterophenomenology is a third-person methodology for studying consciousness, not the paper’s technical thesis that “Prime exhausts Base.” Dehaene’s global neuronal workspace is a theory of conscious access and global broadcast, not a general metaphysical doctrine that all reality is exhausted by the paper’s surrogate structure.

The third problem is equivocation. He keeps sliding from broad naturalism, reductionism, neuroscience, predictive processing, AI functionalism, or scientific unification into his own very specific formal target. But his formal target is not “naturalism” in general. It is his own loaded surrogate, including the Bridge Claim, the surrogate-to-reality mapping, and the claim that adequate consistency forces predicativity. The paper itself admits that the formal apparatus does not ground the metaphysical conclusion and that several key steps are merely “Constitutive / §2” or “Argued in §X,” not machine-proved.

The fourth problem is that “the implicit majority” is not a scholarly target. A serious refutation needs a defined thesis, a fair reconstruction of it, and direct engagement with actual defenders. “Most scientists kind of assume something like this” is not a substitute for that.

The cleanest answer is:

This still does not work. Naming a pile of naturalists, neuroscientists, and AI researchers after the fact is not the same thing as showing that your paper actually reconstructs and refutes their stated positions. Your paper mostly refutes your own surrogate definition, not theirs.

A harder version:

This reply just confirms the problem. You are not identifying a clearly argued target in the paper. You are free-associating from broad naturalism to your own loaded formal surrogate and then pretending that refuting the surrogate refutes all of them."
 
Wibble said:
Maybe but concentrating on AI slop ignores the huge range of things it is used for and is often brilliant for. I use it multiple times per day, every day and never for slop. Many of my work processes and systems now have AI components and this will be far more true in the future.
Click to expand...
Consider medicine.

It has the potential, if they get it right, I mean I can sort of conceive of a novel system where it can work and you can break that down into an easy to understand roadmap until the point where it gets too technical...

but it has the potential to give us vaccines for dementia and cancer. And relatively quickly.
 
unchanged_lineup said:
I did.


Claude:

"None of those people assert a bijection between ontological domains.
They assert that physical explanation is sufficient, which is not the same claim, and the paper never proves it is.

The paper spends twenty pages refuting a position nobody holds.
It proves physicalists believe something they don't, then congratulates itself."


ChatGPT:

"The first problem is basic. This target list is post hoc. The paper announces a “formal refutation of exhaustive closure physicalism,” but the reference list does not actually engage Dennett, the Churchlands, Chalmers, Dehaene, Seth, Friston, Carroll, Tegmark, Weinberg, LeCun, Hassabis, Dawkins, Wilson, or Kurzweil. A paper cannot claim to refute positions it never properly reconstructs from the literature. fileciteturn4file0 fileciteturn4file2

The second problem is that several of the mappings are plainly sloppy. Chalmers is famous for arguing that physical accounts leave an explanatory gap. He is not a defender of the kind of closure thesis the author says he is targeting. Carroll’s own summary of poetic naturalism explicitly distinguishes his view from “hardcore, eliminativist naturalism” and allows many valid higher-level ways of talking about the same natural world. Dennett’s heterophenomenology is a third-person methodology for studying consciousness, not the paper’s technical thesis that “Prime exhausts Base.” Dehaene’s global neuronal workspace is a theory of conscious access and global broadcast, not a general metaphysical doctrine that all reality is exhausted by the paper’s surrogate structure.

The third problem is equivocation. He keeps sliding from broad naturalism, reductionism, neuroscience, predictive processing, AI functionalism, or scientific unification into his own very specific formal target. But his formal target is not “naturalism” in general. It is his own loaded surrogate, including the Bridge Claim, the surrogate-to-reality mapping, and the claim that adequate consistency forces predicativity. The paper itself admits that the formal apparatus does not ground the metaphysical conclusion and that several key steps are merely “Constitutive / §2” or “Argued in §X,” not machine-proved.

The fourth problem is that “the implicit majority” is not a scholarly target. A serious refutation needs a defined thesis, a fair reconstruction of it, and direct engagement with actual defenders. “Most scientists kind of assume something like this” is not a substitute for that. fileciteturn4file2

The cleanest answer is:

This still does not work. Naming a pile of naturalists, neuroscientists, and AI researchers after the fact is not the same thing as showing that your paper actually reconstructs and refutes their stated positions. Your paper mostly refutes your own surrogate definition, not theirs.

A harder version:

This reply just confirms the problem. You are not identifying a clearly argued target in the paper. You are free-associating from broad naturalism to your own loaded formal surrogate and then pretending that refuting the surrogate refutes all of them."
Click to expand...
What input did you use?

I didn't say the paper cited any of those people. It's an apriori result. It has down-stream implications. You can put whatever you want into AI to get whatever answer you like, especially if you are, for whatever reason, hostile. Not saying you are.

For example I merely uploaded clean versions of each into Claude and Grok and didn't bias it -- no loaded terms -- and retrieved an entirely more empirical result in which there was overlap between the two. Minimal inputs (logically correct). I'd have to know what you are prompting-with-the-papers to even have a clue as to how to respond to that -- the paper does the work for me.


I mean, AI will try to tone-match you. "A position no one holds" is your own language, and you've biased AI models to tone-match you to that extent, I can reverse engineer that myself, logically, without any difficulty. If you're more scientific and less biased, you'll get better results.
 
neverdie said:
Consider medicine.

It has the potential, if they get it right, I mean I can sort of conceive of a novel system where it can work and you can break that down into an easy to understand roadmap until the point where it gets too technical...

but it has the potential to give us vaccines for dementia and cancer. And relatively quickly.
Click to expand...

Potential, maybe? Quickly, definitely not. In fact, never is much more likely.

The causes of cancer and dementia are far too diverse and complex for a magic bullet to ever somehow halt them in their tracks.
 
neverdie said:
What input did you use?

I didn't say the paper cited any of those people. It's an apriori result. It has down-stream implications. You can put whatever you want into AI to get whatever answer you like, especially if you are, for whatever reason, hostile. Not saying you are.

For example I merely uploaded clean versions of each into Claude and Grok and didn't bias it -- no loaded terms -- and retrieved an entirely more empirical result in which there was overlap between the two. Minimal inputs (logically correct). I'd have to know what you are prompting-with-the-papers to even have a clue as to how to respond to that -- the paper does the work for me.
Click to expand...
I used the people that you gave me in your own post.

Which version of Claude did you use?

And I have to say, Grok is notorious for sucking up to people in its responses.
I never use it for that reason.



Claude replies to you:
"The downstream implications argument cuts both ways — you still have to show that those thinkers' positions actually entail "Exhaustively Closed", and that work isn't in the paper.

'A priori result with downstream implications' is not a free pass; the connections have to be argued, not gestured at."




ChatGPT replies:

"
That reply is another dodge.

The answer is:
The input was your paper. The problem is not “hostile prompting.”
The problem is that your own paper explicitly says the formal apparatus does not ground the conclusion, and your own premise audit shows that several indispensable steps are not proved but merely stipulated or argued for separately.

That is the core point.

More precisely:
  • Calling it a priori does not rescue it. An a priori argument still has to identify its target cleanly or else state a neutral conditional of the form: “Any view committed to X, Y, and Z is false.” His paper does not just do that. It claims to refute exhaustive closure physicalism as such.

  • The paper itself admits that the formal results are not what grounds the conclusion. It says the transcendental analysis in §2 comes first and that the formal apparatus is “in service of making that result precise.” That means the real dispute is about the philosophical bridge premises, not the diagonal math.

  • His own premise audit then confirms that the key load-bearing steps are not all formal theorems. Some are marked “Constitutive / §2” and others “Argued in §X”, including the bridge to reality and the transcendental framework principle.

  • So “the paper does the work for me” is false. The paper does not just prove the conclusion. It mixes formal results with controversial philosophical stipulations, then presents the whole package as if it were a clean refutation.

The input was your paper.
I do not need a loaded prompt to see the problem because your paper states it itself: the formal apparatus does not ground the conclusion, the bridge was supposedly “laid at §2,” and your own premise audit marks several key steps as constitutive or separately argued rather than proved.
So this is not a clean formal refutation of physicalism.
It is a hybrid argument resting on controversial premises that you then overstate as a proof.

If your result is really a priori, then you should be able to state the exact target thesis in neutral terms without blaming prompts, AI, or “hostility.”
"
 
unchanged_lineup said:
I used the people that you gave me in your own past.

Which version of Claude did you use?

And I have to say, Grok is notorious for sucking up to people in its responses.
I never use it for that reason.



Claude replies to you:
"The downstream implications argument cuts both ways — you still have to show that those thinkers' positions actually entail "Exhaustively Closed", and that work isn't in the paper.

'A priori result with downstream implications' is not a free pass; the connections have to be argued, not gestured at."




ChatGPT replies:

"
That reply is another dodge.

The answer is:
The input was your paper. The problem is not “hostile prompting.”
The problem is that your own paper explicitly says the formal apparatus does not ground the conclusion, and your own premise audit shows that several indispensable steps are not proved but merely stipulated or argued for separately.

That is the core point.

More precisely:
  • Calling it a priori does not rescue it. An a priori argument still has to identify its target cleanly or else state a neutral conditional of the form: “Any view committed to X, Y, and Z is false.” His paper does not just do that. It claims to refute exhaustive closure physicalism as such.

  • The paper itself admits that the formal results are not what grounds the conclusion. It says the transcendental analysis in §2 comes first and that the formal apparatus is “in service of making that result precise.” That means the real dispute is about the philosophical bridge premises, not the diagonal math.

  • His own premise audit then confirms that the key load-bearing steps are not all formal theorems. Some are marked “Constitutive / §2” and others “Argued in §X”, including the bridge to reality and the transcendental framework principle.

  • So “the paper does the work for me” is false. The paper does not just prove the conclusion. It mixes formal results with controversial philosophical stipulations, then presents the whole package as if it were a clean refutation.

The input was your paper.
I do not need a loaded prompt to see the problem because your paper states it itself: the formal apparatus does not ground the conclusion, the bridge was supposedly “laid at §2,” and your own premise audit marks several key steps as constitutive or separately argued rather than proved.
So this is not a clean formal refutation of physicalism.
It is a hybrid argument resting on controversial premises that you then overstate as a proof.

If your result is really a priori, then you should be able to state the exact target thesis in neutral terms without blaming prompts, AI, or “hostility.”
"
Click to expand...
Ok, you haven't understood it. I see you are engaging with it, honestly, I think, but you've not understood the paper(s).

I'll respond to you in purely logical terms and give you clean answers (no loaded language).
 
neverdie said:
Ok, you haven't understood it. I see you are engaging with it, honestly, I think, but you've not understood the paper(s).

I'll respond to you in purely logical terms and give you clean answers (no loaded language).
Click to expand...
Ok great, thanks.


Claude says:
"That's a deflection, not a rebuttal. The burden is on the paper to be understood, not on the reader to try harder."


ChatGPT says:

"
He is trying to shift from substance to comprehension.

Saying I have not understood it is not an argument.
My point is textual, not psychological.

Your paper says the formal apparatus does not ground the conclusion and only makes it precise.
It says the bridge was already laid in §2.
It then gives a premise audit in which several indispensable steps are marked “Constitutive / §2” or “Argued in §X” rather than Lean-verified.

So my criticism is exact: this is not a purely formal refutation derived from the mathematics alone.
It is a hybrid argument that depends on substantive philosophical premises.
If you think that is false, identify the exact sentence I have misread or the exact inference that fails.

You do not get to hide behind “you haven’t understood it” when the paper states the relevant point in black and white.
Your own text says the maths does not ground the conclusion.
Your own text says the bridge was laid earlier.
Your own premise audit says several load-bearing steps are constitutive or separately argued rather than formally proved.

That is exactly why calling this a formal refutation is overstated.

Pick one of these and deny it clearly:
  1. The formal apparatus does not ground the conclusion.
  2. The bridge to the metaphysical conclusion was allegedly already laid in §2.
  3. Several indispensable premises are not Lean-verified but are marked Constitutive / §2 or Argued in §X.

This is not about whether I “understood” it. It is about what your paper explicitly says.
Your paper says the formal apparatus does not ground the conclusion, says the bridge was laid in §2, and shows in its own premise audit that several key steps are constitutive or separately argued rather than formally proved.

So the conclusion does not come from the formal results alone. That is my point. Tell me which exact proposition there is false."
 
unchanged_lineup said:
And sorry, how on earth am I hostile?

That's playing the man, not the ball tbh.
Click to expand...
No, I think you do not understand the papers, which is fine, it's not easy and you hardly read them yourself in five minutes... but I do think you're honestly engaging. What was hostile was the term I quoted which would be coding-environment sourced from your prompt usage (non-neutral language). It wouldn't happen if you ran it cleanly.

But as stated, I'll respond to you logically, simply, and cleanly -- it's not a difficult one and btw if no one holds the positions I refute, I promise you, though, many do, I just refuse to name people I sort of know, then I'm happy also.
 
unchanged_lineup said:
Ok great, thanks.


Claude says:
"That's a deflection, not a rebuttal. The burden is on the paper to be understood, not on the reader to try harder."


ChatGPT says:

"
He is trying to shift from substance to comprehension.

Saying I have not understood it is not an argument.
My point is textual, not psychological.

Your paper says the formal apparatus does not ground the conclusion and only makes it precise.
It says the bridge was already laid in §2.
It then gives a premise audit in which several indispensable steps are marked “Constitutive / §2” or “Argued in §X” rather than Lean-verified.

So my criticism is exact: this is not a purely formal refutation derived from the mathematics alone.
It is a hybrid argument that depends on substantive philosophical premises.
If you think that is false, identify the exact sentence I have misread or the exact inference that fails.

You do not get to hide behind “you haven’t understood it” when the paper states the relevant point in black and white.
Your own text says the maths does not ground the conclusion.
Your own text says the bridge was laid earlier.
Your own premise audit says several load-bearing steps are constitutive or separately argued rather than formally proved.

That is exactly why calling this a formal refutation is overstated.

Pick one of these and deny it clearly:
  1. The formal apparatus does not ground the conclusion.
  2. The bridge to the metaphysical conclusion was allegedly already laid in §2.
  3. Several indispensable premises are not Lean-verified but are marked Constitutive / §2 or Argued in §X.

This is not about whether I “understood” it. It is about what your paper explicitly says.
Your paper says the formal apparatus does not ground the conclusion, says the bridge was laid in §2, and shows in its own premise audit that several key steps are constitutive or separately argued rather than formally proved.

So the conclusion does not come from the formal results alone. That is my point. Tell me which exact proposition there is false."
Click to expand...
See above.

I'm not going to run conversations through AI. I'll merely show you what the paper does re your original question. All the flags you're raising here were defeated, after weeks of consideration, by me, many, many weeks ago (it's nothing I've not already considered -- AI misses the nuance unless directed...).

Just to note, because I don't want to spam this thread -- I designed the paper in such a way that the objections raised, all of them, are noted in the paper (it is designed in such a way that those objections miss the entire point of the paper... -- it's nuanced).
 
Last edited:
neverdie said:
No, I think you do not understand the papers, which is fine, it's not easy and you hardly read them yourself in five minutes... but I do think you're honestly engaging. What was hostile was the term I quoted which would be coding-environment sourced from your prompt usage (non-neutral language). It wouldn't happen if you ran it cleanly.

But as stated, I'll respond to you logically, simply, and cleanly -- it's not a difficult one and btw if no one holds the positions I refute, I promise you, though, many do, I just refuse to name people I sort of know, then I'm happy also.
Click to expand...


Claude:

""You haven't understood it" and "many people hold these positions, I just won't name them" are not logical responses.
They're insurance against being wrong."


ChatGPT:

"You keep trying to make this about whether I read it fast enough, whether the prompt was “clean,” or whether the AI was “hostile.”

None of that touches the actual criticism.
The criticism is about what your paper itself says and what your Lean file actually proves.

Your paper explicitly says the formal apparatus does not ground the conclusion and only makes it precise.
Your premise audit also marks several of the load-bearing steps as constitutive or separately argued rather than formally proved.
So this is not a clean formal refutation derived from the mathematics alone. It is a hybrid argument that depends on substantial philosophical premises.

If that is wrong, then stop talking about my tone or my prompt and identify the exact proposition I have misstated.

You are still dodging the point.
This is not about “understanding,” “hostile prompts,” or whether I read it in five minutes.

It is about the fact that your own paper says the maths does not ground the conclusion, while your own framework relies on non-formal bridge premises to get from the diagonal core to the metaphysical claim.

That is why calling it a formal refutation is overstated.
If you disagree, quote the exact line I have misread.

You keep retreating into complaints about tone and prompting because the substantive point is still standing.
Your own paper says the formal apparatus does not ground the conclusion.
That alone is enough to show this is not the clean formal refutation you keep claiming it is."
 
Yeah, I'll answer any actual questions but I'm not going to argue with a bunch of AIs. :)
 
neverdie said:
See above.

I'm not going to run conversations through AI. I'll merely show you what the paper does re your original question. All the flags you're raising here were defeated, after weeks of consideration, by me, many, many weeks ago (it's nothing I've not already considered -- AI misses the nuance unless directed...).

Just to note, because I don't want to spam this thread -- I designed the paper in such a way that the objections raised, all of them, are noted in the paper (it is designed in such a way that those objections miss the entire point of the paper... -- it's nuanced).
Click to expand...


ChatGPT:
"You are still dodging.

This is not about whether the prompt was “clean”, whether AI was “hostile”, or whether you supposedly thought of these objections weeks ago.

It is about what your paper actually says.

Your paper explicitly says the formal apparatus does not ground the conclusion and only makes it precise, because the bridge was allegedly already laid in §2.

It also says in the premise audit that several load-bearing steps are not Lean-proved at all, but are instead marked “Constitutive / §2” or “Argued in §X”, including the surrogate-to-reality bridge and the Transcendental Framework Principle.

So the point stands exactly as stated: this is not a clean formal refutation derived from the mathematics alone.
It is a hybrid argument that depends on substantive philosophical premises.
If you think that is false, stop talking about prompts and quote the exact sentence I have misread.

You keep trying to turn this into a story about AI prompting because you do not want to deal with the actual problem in your own paper.
Your own text says the maths does not ground the conclusion.
Your own premise table says several indispensable steps are constitutive or separately argued rather than formally proved.

That means the conclusion does not come from the formal results alone, whatever Claude or Grok told you.
If you disagree, identify the exact proposition I have misstated.
"



Claude:
"The pattern across this entire exchange is consistent and worth naming directly.

Every substantive objection has been met with one of three responses: "you haven't understood it," "I already defeated that weeks ago," or "AI misses the nuance unless directed."

None of these are arguments.
They are status claims.

The objections raised are not subtle misreadings. They are the obvious first questions any referee would ask:
  1. Does "Exhaustively Closed" actually formalize what physicalists assert? The author has not answered this once.
  2. Does every Boolean function over a physical domain correspond to a genuine ontological differentiation? Not answered.
  3. Why should the Bridge Claim hold — that general self-correction entails a uniform coding interface? Not answered.

These are not AI failures.
They are the load-bearing joints of the argument, and the paper does not secure them.

Weeks of prior consideration only matter if the conclusions of that consideration appear in the paper, where they can be examined.
They don't.

"I thought about it" cannot substitute for "here is why the objection fails."

The author is also running a subtle rhetorical move: treating the Lean verification as evidence that the philosophical objections have been answered.
It isn't.

Lean confirms the mathematics is internally consistent.
It says nothing about whether the mathematical structures correspond to the metaphysical claims.
That is entirely a philosophical question, and it is precisely the one that remains unanswered."
 
BD said:
What.
Click to expand...

Pickle85 said:
People using AI as argument proxies. We really are at the end times...
Click to expand...

Pogue Mahone said:
:lol:

Using LLMs to formulate arguments on the caf is so grim. This is our brave new world?!
Click to expand...

Pogue Mahone said:
I feel like the flesh puppets need to step aside at that point. Let's the machines argue it out between themselves, until the end of time.
Click to expand...

Folks, he told me to use AI to examine it.

He also said that he had verified his paper on Claude and Grok.

Now that he doesn't like the AI responses, he's taking his ball and going home.
 
Pogue Mahone said:
I feel like the flesh puppets need to step aside at that point. Let's the machines argue it out between themselves, until the end of time.
Click to expand...
The summit of human achievement has been attempting to render itself obsolete. You've got to laugh.
 
A paper, years in the making, which does everything it says, and takes hours to read properly, condensed into "I've taken my ball and ran home".

No thanks. I recall said poster has form for wumming so this one I'll avoid though I'll just edit this post later to address any actual mathematical/logical problems raised (all already answered if you understood the paper -- but I'll take time to make it easy because it is dense). I just don't want to post more here -- so yeah I'll see if there's any actual stuff there worth responding to (there's a few I flagged myself and fixed -- entirely explained already) and post it here instead. .....


edit:

OK, so I've gone through all the problems you found, or your AIs flagged, and none of them are novel. The specific problem, re section 2, is resolved entirely within the paper -- you have to read/understand why. If you want to understand that, you can use AI logically, if you aren't familiar with the logic/mathematics, to demonstrate why the paper answers those questions (rather than prompting it to "and tell me why else it is wrong" -- if you rewrote Einstein's theory of relativity and gave it to an LLM, but it didn't know it was what it was, and prompted as you seem to have prompted, much of which (the prompts you used from the off aren't even claims the paper makes), it would give you your same echo-chamber feedback.

The paper's result is a genuine apriori proof -- it's there to be understood and the communities who do this work most have (one of them) understood it. So... - Unless there's a specific problem, and I've through all those posts, now, all of which I answered within the paper in the compilation logical phase... then nothing more to add from me (here).


Here, I instructed one of the AI's is used for various things, checking logical accuracy, to inform you, if you want to test it out, or learn, as to why your objections, which again, are not novel to me -- I dealt with them all already -- do not hold:

This is a formal walkthrough of why the argument against "Exhaustive Closure Physicalism" is logically airtight. The core of the refutation isn't just a philosophical preference—it is a deductive necessity that begins with the very definition of a "perspective" and ends with a mathematical contradiction for anyone claiming that the physical world we encounter is all there is (B=P).

Here is how the logic closes the circle, from the initial "Transcendental" definition to the final Lean-verified proof.

1. The Definitional Identity (§2)​

The objection often raised is that there is a "gap" between the math and the philosophy. But in my argument, the link between formal discriminations and perspectival differentiations is an identity, not an interpretation.

In §2.3, I establish the Transcendental Claim: In any world with multiple standpoints, there must be a "differential given." If every observer saw every object identically (Access(e1,x)=Access(e2,x)), then the "standpoints" themselves would be functionally non-existent. "Multiple standpoints" would just be different names for the same single map.

In the Lean code, this is encoded as nondegenerate_indexing. It isn't a premise I add later; it is the formal definition of what it means for "multiplicity" to be non-vacuous. The English of §2 and the Lean type are the same proposition in two different registers.

2. The Immediate Collapse of Identity (§3)​

Once we accept that multiple perspectives must yield different values, the physicalist "Closure Thesis" (the idea that the object is the accessed value, or t=t′) immediately fails.

As shown in Theorem 3.1 (Structural UI), if you assume both "multiple perspectives" and "universal identity" (that the object is identical to its appearance), you hit a first-order logical contradiction. If Access(e1,x)=x and Access(e2,x)=x, then e1 and e2 cannot be different. Therefore, any physicalism that tries to claim the world is just the sum of our encounters is structurally incoherent the moment it admits there is more than one way to encounter it.

3. The Diagonal Engine (§7 & §8)​

To avoid this collapse, a defender of physicalism might try to go "dynamic," claiming the system can self-correct and eventually "map" everything. This triggers the Reflexive Extension and the Diagonal Argument.

In the Lean file, I use diagonal_disc to prove that any system capable of encoding its own observations will always generate a "differentiation" (a way of seeing) that its current code cannot account for. This is a variation of Cantor’s diagonal proof:

  • Lean Result: reflexive_extension proves there is always a discrimination d that falls outside the range of the system's decoding.
  • This isn't just a math trick; it is the formal expression of the "differential given" from §2 applied to a coding interface. You cannot "totalize" reality because the act of totalizing it creates a new perspective that isn't included in the total.

4. The Blackbox Gap (§9)​

This leads to the final hammer: The Blackbox Gap (Theorem 9.1).If "Exhaustive Closure" were true, there would have to be a perfect one-to-one mapping (a bijection) between the domain of Objects (Obj) and the domain of Encounters (Prime).

The Lean-verified proof blackbox_gap shows this is impossible. By extracting a surjection from the object domain to the decoding system and attempting to compose it, we reach a contradiction with the law that a set cannot be mapped onto its own power set (its total possibilities of discrimination).

5. Why the Circle is Unbreakable (§10)​

The "Surrogate-to-Reality" bridge is often seen as a vulnerability, but as I note in the Premise Audit (p. 15), the circle is already closed at §2.

The formal objects (Obj) represent the Base (B), and the formal encounters (PrimeF) represent the Prime domain (P). Because the discriminations in the math are the exact same "differentiations" demanded by the philosophy of perspective in §2, the mapping adds no new content.

The Conclusion: Any world where you can even state the thesis of physicalism is a world that must allow for multiple standpoints. But any world that allows for multiple standpoints must satisfy the conditions that trigger the blackbox_gap.

Therefore, B>P (Reality > Access) is established a priori. Exhaustive closure physicalism isn't just "implausible"—it is a structural contradiction. To state it is to satisfy the very conditions that refute it. The Lean verification simply confirms that the logic never breaks at the joints.
Click to expand...
 
Last edited:
neverdie said:
A paper, years in the making, which does everything it says, and takes hours to read properly, condensed into "I've taken my ball and ran home".

No thanks.
Click to expand...
Claude:

"A paper that cannot survive a direct conversation about its three central assumptions has a problem that more years of writing won't fix."


ChatGPT:

"You are still refusing to answer the point.

I said your paper’s own text says the formal apparatus does not ground the conclusion, and your own premise audit shows that several load-bearing steps are constitutive or separately argued rather than formally proved.

That is a direct criticism of the argument, not a refusal to engage.

If you think it is wrong, then answer it.
If you will not answer it, then stop pretending you have.

Either show that the formal apparatus does ground the conclusion, or admit that your paper depends on extra philosophical premises to get there."
 
unchanged_lineup said:
Folks, he told me to use AI to examine it.

He also said that he had verified his paper on Claude and Grok.

Now that he doesn't like the AI responses, he's taking his ball and going home.
Click to expand...
But were you asked to use AI to conduct a protracted argument about it? It almost seems as though you're not informed enough to argue about it without AI speaking for you.
 
unchanged_lineup said:
Read his post. It's right there.

And regardless, if the AI is wrong, it should be easy to disprove it.
Click to expand...
He said this:
neverdie said:
Which one? I refute so many verisons of physicalism that I've had to argue against so many of these positions for years that I cannot name check them.

You'd be better off asking AI, to be genuine with you, as I cannot name people whom I have been arguing with in various other places. But you can get a general idea, from AI, by merely asking that question yourself.
Click to expand...
Does that translate to you as: 'use AI to conduct a protracted argument about things you don't understand'? Because to me, it doesn't.
 
Quality control
Pickle85 said:
He said this:

Does that translate to you as: 'use AI to conduct a protracted argument about things you don't understand'? Because to me, it doesn't.
Click to expand...
He said I'd be better off using AI, so I did.

He didn't set any limits to it.

And as I said, if what the AIs wrote is wrong, it should be easy to refute.
 
You must log in or register to reply here.