you are talking about cases where a person exists to already need to share resources, space, and "achieve stuff". Hence why I really try to emphasize the unnecessary nature of creating these things for another person to encounter (the very imposition in question). — schopenhauer1

People do already exist with these needs. The current community. All of whom will suffer if there's no succeeding generation.

it is something that will (or at least could) affect another person so squarely fits in morality. It is not exempt because it doesn't fit with other cases. The rule itself must make room for this decision as well. — schopenhauer1

I agree. Procreation should not be exempt from any moral rule.

Your assumption seems to be that people NEED to be born TO build a better community.. Well, hold on, who says? Why do you get to make that decision? — schopenhauer1

See above. What is 'good' for a community cannot be a subjective decision because we all affect each other, so I have expectation of you and you of me on the grounds of our mutual need for each other. It is not only reasonable to expect others to adhere to the general consensus on what is good, it is completely necessary for a community to function.

But beyond this someone has to make a decision because inaction will definitely cause harm. we can't ask the unborn child, so we have to decide. Not deciding is just a cop out. It doesn't avoid harm, it just avoids direct responsibility. You create a situation where people will be harmed, but you personally get to wash your hands of it and say "wasn't me, not my problem".

As I said, if you don't care about whether a community functions, then that's fine, but it's nothing to do with morality then. You're just making up arbitrary rules.

Because it affects people in significant ways to have to do X. — schopenhauer1

So? Why care about affecting others?

I don't see any contradiction here, or any objective for that matter. — Tzeentch

Virtually all humans thrive on the company of others. The greatest harms we suffer are ostracisation and loneliness, far greater than any physical harm. Beyond that we cannot survive alone, we require the support of a community. If you do nothing to interact with other they will definitely come to this harm. You may not have directly caused it, but this is also true of procreation (I don't directly harm my children, I merely create a situation in which they may be harmed).

You cannot avoid interaction with others full stop. But even if you somehow became the world's first true hermit, you'd cause harm by depriving those who thrive on human interaction the pleasure (and utility) of your company.

First, individuals do not like being harmed. Their will is as good as mine, so I should take care not to harm them and thus violate their will.

Second, if I impose something on someone, I may violate their will. — Tzeentch

So your first clashes with your second. Your objective is unachievable. You might as well toss a coin.

You talk about meaning, I'm talking about good. Two different objectives. What is meaningful is fine being subjective. We can each make our own choices. What is 'good' is not fine being subjective because you sffdct me and I affect you, we share common resources, we share space, we collaborate to achieve stuff we couldn't achieve alone. We must come up with an agreement as to what constitutes the common good.

Once we have such an agreement, the maintenance of it is all morality is. Anything else is pointless. You could say "you mustn't do X", " you must do Y", but why? Who sets these bizarre rules and why ought we obey them?

If you don't want to build a better community, if you don't care for their welfare, then that's fine, you do you, but you've got no business with morality, the subject matter of which is the welfare of our community. Anything less is just a set of arbitrary rules for no purpose.

So your concern for the autonomy of the as yet unborn is noble, important, but completely pointless unless in the service of the larger goal of community welfare. Otherwise, why? Why bother with autonomy? Why bother with rights? Why bother with dignity? What's the purpose?

This is just garbled. — Tate

It's perfectly clear. I'm asking exactly the same of you as you just asked of me. The citations from which you've derived your view. If you don't understand my request then how did yours make sense to you?

if hidden states are only seen indirectly then what is it that is seen directly? — Michael

Nothing. 'Seeing' is a process of inference. Nothing is seen directly. Everything that is seen is seen indirectly. It's not a direct process, it has stages.

I see no such duty, except perhaps minimizing the harm I myself cause to others directly, that is to say by my action. — Tzeentch

This is the one I'm asking about. And...

the reason I would refrain from procreation is because I cannot see a justification for the imposition on another, — Tzeentch

...

Why?

Are these just spontaneous feeling you have, not derived from any deeper objective? They seem, no offense meant, really odd, and intriguing for that reason.

Then what is your chief concern? — Isaac


The search for truth and wisdom, I suppose. — Tzeentch

How does that concern affect the decision to procreate? Is non-procreation more truthful?

With inaction I mean non-interference. So the choice would be not to do anything about a given situation. — Tzeentch

And how does that assist the search for truth?

An inference regarding a hidden state is never called an "observation." — Tate

I've just given three examples in which the subject of observation is the hidden state. As I said, why don't we start with the papers from which you've arrived at the impression that what we observe are inferences, and we can see what the differences are.

My chief concern was never the minimizing of harm to others (besides that which might be caused by myself), or the welfare of others. — Tzeentch

Then what is your chief concern?

Inaction does not cause harm. It's a neutral state. — Tzeentch

So you don't breathe, eat or move then? You are never inactive, so you're always doing. The choice is over what to do.

If you recall a specific case of a hidden state being referred to as "observed", could you point that out? — Tate

inference on models and their parameters given data) that considers hidden states

First mention, first paper. It is the hidden states which are being inferred by the process in question (in this case 'seeing') ie what we 'see' is the hidden state.

They are often referred to as hidden states because they are seldom observed directly.

Second mention, still the first paper. They are seldom observed directly (ie they are observed, indirectly).

Empirical priors (priors that depend on variables) relate to transitions between hidden states, which are encoded in the B matrix

Third paper.

Short of going through every mention in every paper I'm not sure what it is you want.

Why don't you start with a reference informing your opinion that hidden states are not observed.

Because causing harm to others is bad.

This is turning into a silly game. Please get to a point. — Tzeentch

You've not demonstrated that being certain one's actions don't cause irreversible harm before acting minimises harm to others though. The inaction resulting from your uncertainty might cause harm to others.

The point was made way back. Why privilege inaction over action if your concern is the welfare of others? Your inaction is just as likely to result in harm to the welfare of others as your action.

We directly experience the idea of a tree — Art48

We don't. As I said in the other thread, if you're going to start saying that 'direct' requires no intervening data nodes, then we do not 'directly' experience the idea of a tree either. The closest we could get to 'the idea of a tree' might be some of specialised neural clusters in the frontal cortex. You don't 'directly' experience the output from there, you experience a reconstructed memory of neural events a few seconds ago.

Could you give an example of the two being used to refer to the same thing? — Tate

See the papers I cited earlier. Or any papers on inference systems.

To be clear, I stated earlier that there is no need to assume any "external world" at all, and extreme skepticism which doubts the reality of an "external" world. is well justified. To expand on this, I will say that it is highly possible that all of our relations with the so-called external world, including sensations, and communications with others, is done internally. — Metaphysician Undercover

This is simply not possible (where 'internal' applies to some self-organsing system). To recognise a system, a self organising one, there has to be an 'internal' and an 'external' otherwise you're just referring to 'everything', and a self organising system has to have a probability distribution function that is opposed to the Gaussian distribution, as this is just the definition of self-organising.

So simply by the definition of a discrete system we've got, of necessity, an internal state, an external state, a Markov boundary, and two different probability functions on either side of that boundary. We can then infer that for the probability function of the internal state to work against the probability function of the external state, which is required for self-organisation, as above, there needs to be a model of the probability function of the external state in the function of the internal state.

Because if one chooses to interfere in the affairs of others, one should be certain their actions don't cause irreversible harm. — Tzeentch

Why? You just keep repeating arbitrary rules without basing them on any potentially shared objectives. We don't just follow rules for no reason.

The Heisenberg article explicitly states '“This new ontological picture requires that we expand our concept of ‘what is real’.” — Wayfarer

"Expand", maybe. "Replace" is what I'm less convinced by.

Mathematical physics has necessitated the development of both new forms of mathematics, and the coining of new words ('spin', 'color', 'charm') to describe new discoveries. — Wayfarer

Indeed, but not told us that words we've been using for one purpose are 'wrong'. Correct use of language is determined by the community using the word, not by some subset.

It certainly has an impact on how we think about the nature of reality - so many of the books on physics have sub-titles that refer to that (‘Einstein, Bohr, and the Great Debate about the Nature of Reality’, for instance.) — Wayfarer

Yes, I think that each time we discover something new (say quarks) we have to expand what is real to accommodate. A scientist suffering from temporary hallucinations might still say (looking at his machine) "is that quark real, or am I seeing things again".

What I object to is the idea that 'that teacup isn't real'. It is real because thst teacup is the sort of thing we use the word 'real' to describe and that's the only measure there is of what a word means.

What we might have discovered is some property of real things we didn't know about before (like they don't exist until we observe them).

The single-minded emphasis on harms and implicit contempt for the positive does not seem right to me. — DA671

Yes, I agree. There's no point in reducing harm just for the sake of following some rule about reducing harm. We usually weigh predicted harms and benefits. I can't see any reason why procreation should be treated differently.

I was talking about hidden states in general. It makes more sense to me to say we have some degree of confidence in our inferences. That lets me distinguish hidden states from observations. — Tate

Ah, then we're using two different meanings of 'hidden states' which is causing the confusion. I'm using hidden states in its technical sense with regards to Markov bounded systems. An observation, in this sense, is an observation of a hidden state.

I know it's a bit confusing to call a state 'hidden' and then talk about us directly seeing it, but that's just the technical terminology, it will be impossible to read about active inference, nor understand any of the work in this field without understanding what is meant by a 'hidden state' in its technical sense.

When it comes to the well-being of others, to take only those actions the outcomes of which we can predict with great accuracy. — Tzeentch

Why?

You're saying for one's intention to match the outcome it requires omniscience? — Tzeentch

Yes, clearly. One does not otherwise know the outcome in advance and cannot match it to one's intention.

Morality is about what we ought to to. It is necessarily predictive about outcomes.

If one intends to do great good but does only great harm, they clearly cannot be said to be moral. — Tzeentch

Why not? If the harmful outcome was completely unforeseeable we'd commonly refer to such good intentions as virtuous.

But let's take this away from semantics. What ought we do? We cannot predict the future with great accuracy, our inaction could cause as much harm as our action, so what ought we do?

One cannot have both, so you've made moral action impossible. — Isaac


Why not? — Tzeentch

Because that would involve omniscience and none of us are.

If intentions were good but the outcome was bad, then there must have been ignorance at play. In my view that does not justify the action or make it moral as per virtue ethics. — Tzeentch

Well then you're not using the word 'moral' correctly. The degree of prior certainty you're describing is not the kind of action we use the word 'moral' to describe (in fact it describes no class of actions at all, since that level of prior certainty in unobtainable but...). You're describing a different type of action. Let's call it a 'y-moral' action. Now, why ought we only act in ways which are y-moral?

I think a just intention alone does not suffice, though it is a prerequisite for a moral action.

Similarly, a good outcome alone is not enough either.


One needs both. — Tzeentch

One cannot have both, so you've made moral action impossible. Thus the word 'moral' becomes pointless. I suggest, therefore we find it new use - perhaps to describe actions which are reasonably likely to bring about good outcomes?

ignorance is not virtuous — Tzeentch

Ignorance, of the sort you describe here, is neither virtuous non non-virtuous. It's as relevant to virtue as having a nose. We are all ignorant in the manner you describe and cannot be any other. As such the state is irrelevant to virtue. One cannot make into a virtue that which is unobtainable.

All we can reasonably say is a virtue is for one to have taken achievable steps to predict the outcome of one's actions.

would you say that there is a risk that we can start thinking about the community or the collective as this distinct being of sorts whilst disregarding the needs of the individuals, which could be deleterious for the community itself (eventually)? — DA671

Yes, I think one can reason badly about the courses of action which would most benefit the community and ignoring individual welfare would be one such bad reasoning practice.

procreation can be a source of unfathomable value for many (and I haven't even mentioned the indirect value it could have due to the fact that it could lead to the creation of people who would help their communities and would contribute towards the common good). — DA671

Yes. And also many projects which increase the welfare of a community take more than one generation to complete, so it's reasonable to have children to help progress such projects.

The harms to existing beings is also a good point worth thinking about. — DA671

It's the basis of my argument against anti-natalism. One only need a reasonable belief that one's child will not suffer greatly for any small suffering they might experience to be outweighed by the benefits it is reasonable to believe they will bring to one's community. Ethics should be about the welfare of the community as a whole, not of individuals, otherwise no sacrifice for the wider community could ever be ethically advocated.

That isn't known when the decision is made. One may very well be making a decision that greatly damages their community and their future child. — Tzeentch

True. It's a consideration one ought to take into account.

I guess one approach would be to say something like "but the chance of improvement is larger than the chance of damage", or is there some other way to continue in the face of these unknown consequences? — Tzeentch

My preferred solution to the unknown consequences problem is to consider ethics about virtue, not consequence. Virtue only requires that we do our best. Once one has positively answered the question "have I reasonable grounds to think this action will increase the welfare of my community" then one can act virtuously in carrying out that action because the ethic is about the virtue, not the consequence (one might be wrong). Consequentialist ethics is a nightmare because we have no cut-off point. Maybe it will be good in ten years, but bad in a hundred, but then good again in a thousand... where do we stop?

I think most people though, could answer that question positively. Without a future generation almost all communities would suffer great harms, and almost every human alive benefits more from the company of other humans than they do from solace (very few of us are hermits).

I agree with this line of thought. — Tzeentch

By it, the simplest justification for having a child is that it will do more to improve the welfare of one's community (including the future child) than not doing so would.

if one's goal is to do good onto others, one should be humble and seek to do so in ways that are within one's control. — Tzeentch

I don't follow. On what grounds is inaction morally superior to action?

To expand. It seems to me that a person standing by without acting whilst another is clearly in pain is callously immoral, whereas a person arranging a huge benefit concert for starving Africans, but in doing so causing a minor trip hazard which breaks someone's foot, is clearly a good person (despite the unforeseen harm).

Insisting that one only act in ways that are totally in one's control seems a bizarre restriction.

I see "good odds" as the only attempt at a justification here, but I'm trying to see if there are other possibilities. — Tzeentch

Good odds is the requirement, not the justification. One cannot reasonably justify having a child on the basis of good odds that they'll be happy alone. We should no more be wanting to maximise the number of happy people in existence (by procreation) than we should be wanting to minimise the number of unhappy ones (by culling). We ought be concerning ourselves with the welfare of our community. That's what ethics is for, otherwise it's just baseless.

Interestingly enough, negative utilitarians who only wish to reduce suffering might be forced to say that allowing a million people to suffer is good as long as it removes the greater suffering of a single being. Therefore, NU isn't as perfect as some of its proponents might think. — DA671

Yes. NU is as bizarre a ethic as any. Why would we eliminate harm with no-one around to enjoy their harm-free life? One might as well have an ethic around eliminating cheese.

I see no reason why future negative consequences should be in our minds but not the positive ones. — DA671

I agree to an extent, but one has to have careful limits if one is to do that and I think those limits create an asymmetry. Looking at just utility we end up with the utility monster problem (is a million moderately happy people better than a few extremely happy people). You'd have to look at is from a virtue perspective - ought I create some happiness, ought I create some harm (or risk)? I think the answer to both depends on the scale of either. One ought not create the risk of massive unwarranted harm. One ought to create happiness where one can, but one seems less obliged to do the latter than avoid the former.

reality comprises a real 'realm of possibility', from which particular ('determinate') outcomes are precipitated. — Wayfarer

I have a lot of sympathy for this view, it's basically the same as my own, but I'd quibble with the word 'reality'. I don't think we use the word 'reality' that way. We use the word 'reality' to describe our collective 'precipitated outcomes'. When I ask (say, suffering from temporary hallucinations) "Is that teapot real" I mean something more like "does everyone else see that?", not "does the foundational state of the universe contain a teapot"

I think science (which is at least partly what Wheeler is doing) should merely describe the nature of the things we already use words for. I don't think it should be in the business of re-defining how we ought to use words.

what you're suggesting is absurd - that people have no moral obligation to take into account the consequences of their actions. — Tzeentch

I agree. I cannot for the life of me think why arguments against anti-natalism always seem to descend into this particular lunacy. It is obvious that one ought consider the as-yet-unborn child's welfare prior to their birth. It's why prospective parents buy nappies, because they don't want the newborn to be without them (and so uncomfortable). It seems so obvious, I'm genuinely baffled by the popularity of the counter-argument.

There is no point questioning memory as such; if we have no faith at all in memory, then we can have no faith in any knowledge at all. Memory is the foundation of who we are, to question it in a general way would be absurd. — Janus

No. That was my point (reductio ad absurdum, I believe it's called?).

You say we know hidden states, via inference; that our actual experience just is inference, if I've understood you. This makes no sense to me. We infer that there are hidden states, and by definition, they being hidden, we don't know them in the common sense of "knowing'. We don't know what those inferred hidden states are; if we did they would not be hidden. — Janus

That's not how we use the word 'know'. We use the word 'know' to refer to successful models of hidden states. I say something like "I know where the pub is", by which I mean that if I go to the place I believe the pub is, I will find it there. We don't use the word 'know' to refer to thinks we have a direct node-to-node connection with - I don't think anyone outside of cognitive computational neuroscience would be able to use the word 'know' if it were thus restricted.

I am saying that naming is just a matter of fiat: we can call the hidden state a chair, or we can call it the unknown whatever that appears to us a chair. — Janus

Yes. all language is by fiat. There's no book of what things 'really' mean.

if we want to say it is a hidden state at all, then it seems contradictory to say that it is "really" a chair, since we have already acknowledged that we think it is "really" a hidden state, and a chair cannot be anything but a familiar object. — Janus

Uh huh. And why can we not be familiar with hidden states? If we have good models of them, we can be very familiar with them.


Active inference is a scientific theory, that means it's an attempt to describe the world and be able to better predict how aspects of it work. It's not a normative theory, it's not saying we ought use words this way or that, it's describing what is happening in the events we already use words for. We do not need to change our words to match the science. The science describes better what the words are already referring to.

Active inference describes, for example, what 'seeing' is. The intention is not that we say "Ah so we don't really 'see' things then", what 'see' means doesn't change, we're just describing what goes on in the process in more detail. Same with the word 'know'. The process of 'knowing' is not what we (perhaps) thought it was, we have provided some more detail describing what the process of 'knowing' something involves. We've not proven that we don't 'really' know anything. The human activity/state that the word 'know' describes still exists unaltered, we just now have a better model of how it works.

Well, if you consider that each and every internal organ has its own purpose, function, relative to the existence of the whole, which is a living being, then you would understand that each of these organs has a reason for its existence. — Metaphysician Undercover

Not how I use the word reason. If you're talking about purpose, then fine. The purpose of neurons is not to represent the outside world. We know this because there doesn't appear to be anywhere to store that representation, nor do any of the areas involved in seeing, hearing, feeling etc seem to have inputs from areas outside the external-facing sensory organs which are sufficient to explain their activity. Hence we do not 'see' a representation, we 'see' the external world. I can literally watch a message go from the external world to the retina, to the occipital cortex and be recognised by the V4 region as being a certain colour. If I damage, or temporarily cut off that route the subject can no longer see colour. I can't think of a more clear demonstration that the seeing of colour is data from the external world, not an internal representation.

If one expands “the network doing the inference” to include the sensorimotor systems, what happens to the hidden state? — NOS4A2

Then the hidden state would be whatever lay outside whatever nodes you had as the new Markov boundary. The existence of hidden states is just a mathematical outcome of the focus on a system.

If we're talking about a system (whether that's a person or a teacup), it has an internal and an external - that which is it, and that which is not it. Without those basics, we can't be talking about anything, we're simply talking about everything. The idea of hidden states is simply saying that with internal/external defined, you must also accept that there is a boundary, and with a boundary there must be (in information terms) a (set of) final node(s) at that boundary.

So that's a system.

Active inference occurs in self-organising systems (a subset of all the systems described above). The self organising bit meaning that they actively work against a probability gradient (the most probably distribution of our component parts is entropic - randomly scattered). The components of our body are a stage above in terms of less randomness than the components of the world we find ourselves in (which are somewhat more randomly dispersed). So we (our bodies) represent an active inference system, we are working against a probability distribution to maintain our arrangement against random dispersal, at a rate slightly higher than that of our surrounding system (the ecosystem).

As each system has final nodes (which cannot form part of the inference) we should look for final nodes in the human system (having identified it a s an active inference system above). Those nodes appear to be the sensory neurons (the passive element), and our motor actions (the active element - though many include secretion here too).

You can, of course, focus on any system you like with any boundary, but some systems would not count as active inference systems because they cross a threshold of the degree to which they are working against a particular probability distribution of the dispersal of their components.

can an activity that only organisms can be shown to perform—experiencing, thinking, inferring, believing, seeing—be isolated to a single part of it? — NOS4A2

I think so, yes. We can simply remove parts and see if the process continues without them (if we were evil experimenters - otherwise we can just wait until it happens anyway, then study the result). People without arms show absolutely no sign of reduced thinking, inferring, believing or seeing. So we can infer that arms are not necessary. People with damage to the occipital cortex cannot see properly, so we can infer that without an occipital cortex, you can't see.

You were talking about perception. If you're now talking about the future, then no, I don't think we can know the future (in general). I'm claiming we know that which we have a successful model of. The success obviously requires testing in the present.

There's an 80 percent chance of rain. I don't know it's going to rain. — Tate

The hidden state is not a future state, it's a current one.

So we're not trying to be serious here? — Tate

Nothing non-serious about it. If you want to say we don't actually 'know' a hidden state because all we 'really' have access to is our inference about it from experiment, then it is no less true to say that we don't 'really' know that either because all we 'really' have access to is our memory of what the inference was when we made it.

It's the problem with putting 'really's everywhere.

We know the hidden state. We sometimes make errors. I know the colour of the dress. Sometimes I'm wrong.

That's really interesting, thanks. Did you ever read the article here...

https://journals.sagepub.com/doi/full/10.1177/1059712319862774

...where Friston responds to some of the enactivist critique? I'd be interested to hear how well you think it answers the criticisms.

You know what happened when you tested the model. — Tate

Well, if we're not 'overstating', you only know what you currently remember about what happened when you tested the model.

All thought is post hoc by at least a few milliseconds.