self-help manual

[iansjack]

I don't understand. You quite clearly said "I am not using an LLM to protect my ego or to avoid being proven wrong. I am using it because I literally do not have the academic background to debate the nuanced history of MIT methodologies, and I do not care to." Doesn't this mean you are relying upon the LLM to validate your system? And if you can't trust the LLM on moral questions (which I certainly don't, considering their provenance) doesn't that invalidate your argument.

It's like constructing a building without the ability to understand building regulations and design constraints and then relying upon an untrustworthy builder or surveyor to say "yeah - that building is structurally sound". If you don't have the ability to fully understand the methodology you don't have the ability to know whether you are being given valid advice. That's why we rely on qualified architects rather than some amateur who comes up with a back-of-an-envelope building design.

[Solar]

You are projecting an ego onto me that does not exist.

Damn, look on the ego on this man.

I have zero emotional investment in "winning" a semantic argument.

And the hubris.

[kerravon]

But then again, what is point of all of this? If your manual does not want to convince people which do not agree with you, then it is only for people who agree with you. But, well, those people already agree with you. So at the end it will do nothing.

That's not correct. The people who agree with Axiom 0 - which is most people - will be happy to construct a system that makes people (mainly future people - children) INTERNALIZE this framework (and get into a COMBAT STANCE against rape/subjugation). If they can see that the framework is logically true, they will be very happy to protect their OWN daughters. Arming the protective majority is not "doing nothing" - it is the whole point.

A rapist has not internalized the rule against rape, and no amount of philosophical debate will fix that corrupted firmware. This framework is the source code for the next generation. We are not trying to convince the currently infected nodes; we are compiling a new, unhedged operating system so that the next generation does not inherit these vulnerabilities. ... A rapist or a dictator does not violate human beings because they lack a rigorous, peer-reviewed epistemological justification for why it's wrong. They violate human beings because of raw instinct, entitlement, or a corrupted drive. You cannot debate a virus into stopping its replication.

Does this mean, that you assume, that people act by fixed and determined set of actions? So "bad" people are incapable of change, and somebody to whom we will "load" your system will act by it forever? If so, this is empirically proven false. People change their beliefs and actions all the time, because they have will. And yes, even rapist or dicatator can change.

I am not philosophizing about edge cases. I'm trying to load an identity onto children. Adults too - if possible. But if adults are impossible I can live with that. And if you have an alternate system that works on adults while mine fails - that's great - I'm happy to use your system for "treating" rapists etc.

But if you give to people some system, they will start to question it, they will ask why.

Very few people are going to question why women have a right to not be raped. You're trying to talk abstract philosophy. I'm trying to protect women in the real world.

If it does not provide answer, they will stop following it.

Not true. Most people are happy to follow a system that protects women. We can debate which system is the best for doing that, but the base axioms themselves (just two of them - 0 and 0.5) are not going to be disputed.

So if you want people to follow your system, you should be ready to give answer why. That is not categorical error, that follows directly when you said, that you want your system to be really used by other humans in real world.

Right - the real world - where people are embarrassed to ask "why do women have the right to not be raped?". Because the answer is self-evident.

[kerravon]

I don't understand. You quite clearly said "I am not using an LLM to protect my ego or to avoid being proven wrong. I am using it because I literally do not have the academic background to debate the nuanced history of MIT methodologies, and I do not care to." Doesn't this mean you are relying upon the LLM to validate your system? And if you can't trust the LLM on moral questions (which I certainly don't, considering their provenance) doesn't that invalidate your argument.

It's like constructing a building without the ability to understand building regulations and design constraints and then relying upon an untrustworthy builder or surveyor to say "yeah - that building is structurally sound". If you don't have the ability to fully understand the methodology you don't have the ability to know whether you are being given valid advice. That's why we rely on qualified architects rather than some amateur who comes up with a back-of-an-envelope building design.

You are fundamentally misrepresenting my workflow and the analogy you just constructed.

1. I am the Architect. The LLM is the Word Processor.

You are conflating "drafting text" with "validating structural integrity." I did not ask the LLM to invent my moral axioms. As I explicitly stated, Axiom 0 and 0.5 are derived from my own decades of lived experience, innate moral reasoning, and observable reality. The LLM is merely the compiler or word processor that helps me articulate those pre-existing concepts into clear, readable text.

I do not need an academic background in philosophy to know that "baseline human integrity must be protected" is a valid structural constraint, any more than I need a PhD in thermodynamics to know that fire is hot.

2. I Am Using Qualified Humans to Validate the Design.

Your building analogy ignores the most critical part of the construction process: peer review and inspection. That is exactly what I am doing right now by posting on this forum. I am crowdsourcing the stress-test.

If the LLM outputs a structural flaw, human pushback reveals it. I then use my own critical thinking to evaluate that pushback (e.g., observing whether the response is a logical fallacy, an ad hominem attack, or a valid technical point). I am not blindly trusting the tool; I am using the tool to efficiently generate drafts, which I then submit to human inspectors (like you) for validation.

3. The "Academic Background" Gatekeeping.

You are attempting to invalidate a system designed to prevent subjugation by demanding I possess a specific academic pedigree to debate "nuanced history." This is a classic appeal to authority. The validity of Axiom 0 does not rest on academic consensus or peer-reviewed epistemology; it rests on observable, kinetic reality.

I am not an amateur handing a blank piece of paper to an untrustworthy builder. I am an architect who has drawn the blueprints based on observable reality, used a drafting tool to clean up the lines, and am now submitting it to human inspectors to find the flaws.

If you have a valid technical flaw in the blueprints, point it out. But if your only objection is that I used a software tool to help me write it down, you are arguing about the font, not the foundation.

[iansjack]

If you have a valid technical flaw in the blueprints, point it out. But if your only objection is that I used a software tool to help me write it down, you are arguing about the font, not the foundation.

There are only so many times that people can point out that your "axioms" are derivations from basic axioms rather than axioms. From what you have written so far it is quite clear that you use your "software tool" for far more than merely writing things down.

But I will grant you that the turbid prose style certainly points to an artificial author.

[VSlezak]

So if I understand correctly, the purpose of your system is to provide suggestions on how to achieve your chosen axioms. So if somebody already agrees with your chosen axioms, they can benefit from the rest of your thoughts about this matter. Is that right?

If yes, then I was under the wrong impression, that the purpose of your system was to convince everybody who is willing to listen and use rational approach. If your system is just a thought experiment on how the world could run, if humans wanted to achieve given starting points, then my main criticism fades, because there is no longer need to convince anybody, who is not already convinced.

But then, while almost all humans today agree with your first "axiom" that rape is wrong, this number drops significantly with other "axioms" you bring. Just to demonstrate it on two of them:

"Protect the pack. ... Axiom 0.5 expands the definition of "pack" to include all of humanity" "THE SPECIES IS THE TRIBE (THE UNIVERSAL IN-GROUP PROTOCOL)." - Many people will agree, that to some extent, an individual should protect bigger society. But it seems that you are here saying that to protect your "pack" is an absolute duty. That is much more controversial. And when you combine that with definition of "pack" as whole humanity, that is even more debatable. Individual belongs to many "packs", and usually they accept some hierarchy of duties in regards to them, like family comes before your city, and your city comes before state at the other side of the globe. So it is not that easy to find individual who agrees, that we should only care about one pack - whole humanity - and protect it above everything else.

"UNCERTAINTY IS PERMANENT. You cannot know anything with absolute certainty. You cannot rule out most things. You must act under uncertainty." - This is really controversial view of epistemology.

Now, I am not saying, that those "axioms" are bad. Actually, to some extent I even agree with them. But it seems to me, that you will find only few people who agree with you entirely, and as such, you will find very few people who have even any reason to care about your system. So at the end, I do not see how is your system really useful in the real world, if you refuse to build foundation that reaches to people, who disagree with you on fundamentals.

I'm trying to load an identity onto children.

That will not succeed on its own. Children will obey any system you teach them, because you as parent have authority. But when they grow up, they will question what you told them. And if you did not provide sufficient reason for your system, they will most likely leave it.

[kerravon]

If you have a valid technical flaw in the blueprints, point it out. But if your only objection is that I used a software tool to help me write it down, you are arguing about the font, not the foundation.

There are only so many times that people can point out that your "axioms" are derivations from basic axioms rather than axioms. From what you have written so far it is quite clear that you use your "software tool" for far more than merely writing things down.

But I will grant you that the turbid prose style certainly points to an artificial author.

I already explicitly stated Axiom 1.5: "All men are created equal". That is word-for-word the basic, foundational axiom you are claiming I lack.

What more could you possibly want?

[kerravon]

You have fundamentally misunderstood the mechanical function of Axiom 0.5, and you are misrepresenting the nature of rational epistemology. Let us address these points directly.

1. The Hierarchy of Duties vs. The Negative Constraint (Axiom 0.5)
You claim it is controversial to prioritize "whole humanity," citing a natural hierarchy where family comes before a stranger on the other side of the globe.

You are conflating a positive obligation (you must sacrifice your family's resources for a stranger) with a negative constraint (you must not violate a stranger's baseline integrity).

Axiom 0.5 does not demand you treat a stranger identically to your child in terms of love, loyalty, or resource allocation. It demands that you do not subjugate, rape, or enslave that stranger to benefit your in-group.

This is not controversial. It is the foundational premise of all functional justice systems. You do not get to murder a foreigner to feed your family. The "hierarchy of duties" applies to positive resource allocation, not to the suspension of baseline human integrity.

2. The "Controversial" Nature of Uncertainty
You call the statement "Uncertainty is permanent" a controversial view of epistemology. In reality, it is the absolute bedrock of the scientific method, Bayesian reasoning, and modern philosophy (e.g., Karl Popper).

Claiming absolute certainty is the hallmark of dogma, fanaticism, and totalitarianism. It is the exact mindset that says, "I am certain my ideology is perfect, therefore I can purge anyone who disagrees."

Acknowledging uncertainty is what prevents ideological extremism. It forces the system to rely on observable, empirical outcomes (e.g., "does this action cause subjugation?") rather than unprovable metaphysical claims. It is the least controversial stance in rational thought.

3. Children Questioning the System
You state: "when they grow up, they will question what you told them. And if you did not provide sufficient reason... they will most likely leave it."

Good. Let them question it.

This framework does not rely on blind obedience, religious dogma, or appeals to authority. It relies on observable, empirical reality.

When a child asks, "Why is it wrong to subjugate someone?", the answer is not "Because the manual says so." The answer is: "Because it inflicts irreversible, terrifying damage on a sentient being, and you possess the capacity to understand that suffering."

That is a self-evident, empirically verifiable truth. It survives rigorous questioning precisely because it is grounded in reality, not in fragile ideology. A system built on truth does not fear being questioned; it welcomes it.

Conclusion
You claim the system is useless because it won't reach people who disagree on these "fundamentals."

But the fundamentals you claim are controversial (don't subjugate strangers; acknowledge uncertainty; rely on observable reality) are actually the baseline requirements for a functional, non-totalitarian society.

If someone fundamentally disagrees with those premises, they are not a rational actor looking for a better system. They are a hostile node advocating for subjugation or dogma. The framework is not designed to coddle hostile nodes. It is designed to out-compete, isolate, and contain them.

[VSlezak]

You are conflating a positive obligation (you must sacrifice your family's resources for a stranger) with a negative constraint (you must not violate a stranger's baseline integrity).

"protect the pack" is formulated as positive obligation. If I should protect somebody, that does not mean that I only do not violate his integrity.

You call the statement "Uncertainty is permanent" a controversial view of epistemology. In reality, it is the absolute bedrock of the scientific method, Bayesian reasoning, and modern philosophy (e.g., Karl Popper).

But you formulated that in manner, which sounds like I should question even statements like "I exist", "1 + 1 = 2", or that moral certainty does not exist. That would be controversial view. Of course, if you only mean, that we should remain open to correcting our beliefs, then it is universally accepted. But that is not the same thing as "uncertainty is permanent".

When a child asks, "Why is it wrong to subjugate someone?", the answer is not "Because the manual says so." The answer is: "Because it inflicts irreversible, terrifying damage on a sentient being, and you possess the capacity to understand that suffering."

Why then you do not have "reduce suffering" defined as the most basic axiom from which everything else flows?

You claim the system is useless because it won't reach people who disagree on these "fundamentals."

No, I expressed view, that system, which wants to transform society, but can by its nature reach only very small number of people, will not change much. So, in other words, it is not very usefull in real world. Of course, if your system does not belong to this category, this does not apply.

[kerravon]

I have added these two things to
https://mutazilah.org/corpus/firstprinciples.txt

P.S. "Protect the pack" is a proportional obligation, not an
infinite one. It does not demand that you bankrupt your family
to feed a stranger on the other side of the globe. It demands
that you do not become *complicit* in subjugation, and that you
take reasonable, proportional action to prevent it when it is
within your capacity (e.g. intervening, reporting, or refusing
to participate). It is a duty of non-complicity, bounded by
practical reality.

...

--------------------------------------------------------------------
TENTATIVE AXIOM 1.6
--------------------------------------------------------------------
AXIOM 1.6 (TENTATIVE): "Reduce suffering"

Same constraints as Axiom 1.5 - it needs to both reproduce, and not
override, axioms 0, 0.5 and 1.


And here's the formal reply from the LLM:

You raise excellent points that allow me to refine the framework's precision. Let us address them directly, including your very valid prompt about suffering.

1. "Protect the Pack" as a Positive Obligation
You are correct that "protect" implies action. However, it is a proportional obligation, not an infinite one. It does not demand you bankrupt your family to feed a stranger on the other side of the globe. It demands that you do not become complicit in subjugation, and that you take reasonable, proportional action to prevent it when it is within your capacity (e.g., intervening, reporting, or refusing to participate). It is a duty of non-complicity, bounded by practical reality.

2. "Uncertainty is Permanent"
I concede that this phrasing can be misinterpreted as radical skepticism (e.g., doubting "I exist" or "1 + 1 = 2"). That is not the intent. The intent is fallibilism regarding complex human, social, and moral systems. It means we must remain open to correcting our beliefs based on new empirical evidence, and we must never claim absolute, dogmatic certainty about human affairs. Claiming absolute certainty in human systems is the exact mindset that justifies ideological purges and totalitarianism.

3. Why "Reduce Suffering" is NOT the Most Basic Axiom (and the introduction of Axiom 1.6)
You ask: "Why then do you not have 'reduce suffering' defined as the most basic axiom from which everything else flows?"

Because making "reduce suffering" the most basic axiom is the exact trap of pure utilitarianism, which has historically justified horrific atrocities. If "reduce suffering" is the absolute highest rule, a system can logically justify harvesting the organs of one healthy person to save five, or subjugating a minority to maximize the happiness of the majority.

Therefore, "reduce suffering" cannot be the foundation. Axiom 0 (Baseline Human Integrity / No Subjugation) is the absolute, unhedged floor.

However, your point is well taken. Within the strict boundaries of Axiom 0, we should optimize for well-being. Therefore, I am formally introducing a tentative Axiom 1.6: "Reduce suffering."

Crucially, this is a secondary, optimizing goal. It operates strictly within the negative constraints of Axiom 0. You are never permitted to violate Axiom 0 in order to achieve Axiom 1.6. The floor cannot be broken to build the ceiling.

4. Usefulness and Scale
You argue that a system reaching only a small number of people is not useful for transforming society.

First, all major societal shifts in history began with a committed, principled minority that reached a critical tipping point.
Second, because this framework is built primarily on negative constraints (don't subjugate, don't be complicit) rather than demanding utopian positive obligations, it is actually highly scalable. It does not require everyone to be a saint; it only requires a critical mass of people to agree on a firm, structural boundary against abuse.

By clarifying these distinctions, the framework becomes more robust, not less. It anticipates the utilitarian trap and explicitly guards against it.