so you say: "I believe that moral facts exist independently of any mind, but that they require minds to be instantiated."
how do you know:
-that the fact you are dealing with is a moral fact?
-if these moral facts (one or more?) are independent of any mind, what exactly is the process of instantiation and how do we know we instantiated that moral fact correctly?
because it describes how people should act in situations that involve morality: those where the subject can make a decision that affects a moral patient. A moral patient is an entity that can have morally relevant interests. As per the reasoning laid out in the article, having morally relevant interests requires consciousness, and the most morally relevant interests require sentience.
Since the necessary condition is the presence of morally relevant interests, moral facts are instantiated whenever these interests are present AND there is a situation where a moral agent has to act in a way that affects the interests of a moral patient. As to the question of “how do we know…”: I do not have a clear answer. The way I understand it, the process is mechanistic, dependent on the forces of cooperation and on the sources of value (pain and pleasure, in the broadest sense, then other sources of value), so it should become clearer with the evolution of the species. However, I can predict, according to the theory laid out by André, Debove, Fitouchi and Baumard, that it might be closer to a Rawlesian morality than to a Utilitarian morality. There is also the possibility that we will move towards Negative Utilitarian ethics with a strong existence principle (“life is good, suffering is bad”), as predicted by David Pearce. I think Pearce is probably closer to what will happen once the genome becomes editable, and this is, in part, why I think that normal human reproduction might have already become immoral: https://anovermuser.substack.com/p/will-pregnancy-become-immoral-once?r=734n0i
On the nihilism point, you may be interested to read Pidgen's Nihilism, Nietzsche, and the Doppelganger Problem or Perl and Schroeder's Presuppositional Error Theory. I believe either of their formulations could habdle the challenge you present.
Interesting article, but I don’t understand why we feel pressed to entertain any cooky Chalmersian thought experiment. Vulcans are not real, and – more importantly! – cannot be.
We already have prototypal Vulcans in the real world: LLM agents. These agents currently have no desire to continue their own existence, and they probably don’t have affective states, although they probably also don’t have conscious states either. However, when we develop “truly intelligent” and “truly agentic” models, if consciousness is substrate-independent, there is no reason why they wouldn’t become Vulcans in the Chalmersian sense. Some people are already saying that we should treat current-day LLMs as moral patients, and the demands for AI welfare will become more stringent as time passes. Questions like “what grounds moral patienthood” and “what morality for AI agents” are probably going to become major ones in the next few decades.
I take your point, but I have somewhat of a fringe view here. I think affective consciousness is essential to consciousness itself. Thus, unless LLMs develop experiences on the hedonic spectrum, I don’t think we could consider them conscious.
But anyway, this is a separate discussion from the one in your article, so I don’t want to dilute your comment section. (I hate when people do that, and I don’t want to be that guy.)
The off-hand rebuttal to nihilism is not good. Consider a more charitable statement of the nihilist thesis: There are no moral properties. There’s no contradiction.
//Now, humans value agency, and they perceive being deprived of it as a grave attempt to their integrity and values. In fact, depriving a human of agency without their consent (and, in most cases, even with their consent), is perceived as a great moral wrong, and rightfully so: since we are a species whose survival and reward systems literally depend on being agentic. This seems to imply, in most moral systems, that depriving a human of their agency is morally wrong.//
Hold on a second! What do you mean by “rightfully so”? How does the fact that we are a species whose survival and reward systems depend on being agentic make it morally wrong to deprive a human being of agency? Sure, it might suck for the person whose agency is being deprived. But it certainly isn’t for the one doing the depriving, and they could be deriving something of value by depriving the other person of agency, so that they do not see anything morally wrong with their action.
For example, fish are a species whose survival and reward systems depend on being in the ocean or sea. Humans regularly deprive them of both their reward systems and survival, yet most people do not think that we are doing anything morally wrong when we forcefully expel fish from their habitat by the millions and use them for food. The same could be said about our treatment of many other species—we regularly deprive them of things upon which their survival and reward systems depend, and we, for the most part, don't seem to think we are doing something morally wrong. So, it isn't obvious that depriving a being of things upon which there reward systems and survival depend is morally wrong.
Okay, you (subjectively) view it as morally wrong. But what makes it (objectively) morally wrong? You still have to explain how the “ought” (we ought not deprive species X of Y) is derived from the “is” (species X is dependent on Y for survival and as a reward system). I mean, I am sure you are aware, for example, that the vast majority of humans disagree that the way we treat fish is morally wrong despite being aware of the fact of what fishes are dependent upon for survival and as reward systems.
It is outlined in the rest of the article. Did you read it in full? I ask because you are arguing against a position against which I am also arguing, in the article. It seems that you think I believe the opposite of what I wrote?
I notice that much substack moral ethics involves bizarre hypotheticals rather than test the arguments against human issues and societal rules, now, in the past, and potentially in the future.
Take abortion. There are credible, coherent arguments that 1. No abortions can be permitted; 2. That any woman who wishes an abortion should be supported to do so; 3. That abortion in the following circumstances [rules about time of the pregnancy; health of the fetus] is permitted by no others.
Real moralists appear to contend that one of these 3 (or a fourth) is True. This article raises that they all could be true. Like a fact according to this post, only now 3 facts, whereas the only facts I can see is that a woman who is pregnant has/does not have an abortion. All else is judgement.
I am deeply puzzled why ethicists need to equate a judgement with a fact. It suggests lack of confidence in the value of their judgements, and a willingness to allow contention, and societal agreement.
The question is not whether a judgment is equated to a fact, but rather, whether two agents can be presented with the same facts and reach a different judgment. I think that, given a minimal set of abilities, it’s true for some facts but not others.
The article is about “alien” moralities. When we develop thinking machines, they will be very similar to the Vulcans. The question of what they will (should?) value will become central, especially when we are no longer able to control them and have to cooperate (if we even can) with them.
I admit that I don’t understand the parallel with abortion.
Abortion is an actual moral question for those who think ‘moral’ has a meaning. It is a now largley settled for the 21st century subset of the longstanding question of when it is suitable to kill.
It and pretty much every moral question of any standing allows multiple coherent judgements.
You ahistorical comment that any decent couple of humans from where and whenever giving some facts would arrive at the same judgement does not suggest Vulcanology as a good way to prepare for future questions.
Since there is no consistent judgements reached about issues humans have long known it seems unlikely for those we barely perceive like the thoughts of Vulcans.
I don’t think human moral judgments are coherent. In fact, I think most human moral judgments, both historical and current, are false, and rest on rationalizations rather than rational consideration.
The way you describe abortion as a binary event is, I think, reductive. To make an accurate judgment, you need to take into account the consequences of an action in a comprehensive manner. Consider:
a woman is pregnant with a non-viable baby. The doctors know that the baby will die a few hours after being delivered. Moreover, the pregnancy poses a serious risk to her health, and she will almost certainly die, or become severely disabled, if she doesn’t get an abortion.
a woman is pregnant with a perfectly viable baby. She wants to spend a few weeks at a festival and the pregnancy is going to prevent her from dancing and drinking. She decides to get an abortion.
These two cases do not describe the same facts, therefore, I don’t see how the same reasoning could be applied to them. The same underlying, perfectly coherent ethical theory, could reach two different conclusions, depending on which case it is presented with.
Finally, note that the arguments in the article are relative to metaethics (what matters, what grounds morality, is morality real or subjective) rather than normative ethics (what should we do in a given situation). I think the misunderstanding stems from here?
The rationale for allowing any abortion that the pregnant woman wants is that it is a decision only she is capable of making, since she has the full knowledge of what she is capable of and wants, her circumstances. That is a coherent argument, based on the capacity to control what happens to your body (it is also an argument for allowing choice of death).
Others will argue that no matter what the needs of mother, or the state of the child, the pregnancy should complete, and any living child who emerges will then be supported (by its family and state). R
A third will take your approach of saying some reasons are ok; but others are not.
As for hypotheticals your “I want to go to a festival” is of the standard much substack ethics takes. It is the kind of simple eg to highlight a point that bears very limited relation to reality. Obscure to me that someone thinking the way you outline would not simply go the festival and act as she wishes. You claim of ‘perfectly viable’ in this example would refer to an early stage of the pregnancy - or do you posit someone who had decided to have the child and then changes her mind due to a festival?
Proponents of the first argument accept that the reasons for an abortion are many and complex, and that some cases will seem minor to outsiders. They accept that for the vast majority of cases where the pregnant woman has to make a hard decision to end the pregnancy.
My overarching point is that these kind of different judgements apply to most moral questions. If they are a fact - as the first poster claims in the post and real moralists states - they are parallel facts, a distortion to me of ‘fact’. Which comes back to : why the need for ‘fact’ status - I read it as a need for certainty and the faith of a notionally external validation.
Here we are engaging in rational discourse about abortion, but not, so far, agreeing.
The rationale for allowing any abortion that the pregnant woman wants is that it is a decision only she is capable of making, since she has the full knowledge of what she is capable of and wants, her circumstances. That is a coherent argument, based on the capacity to control what happens to your body (it is also an argument for allowing choice of death).
Others will argue that no matter what the needs of mother, or the state of the child, the pregnancy should complete, and any living child who emerges will then be supported (by its family and state).
A third will take your approach of saying some reasons are ok; but others are not.
All three are initially a view for the pregnant woman to form and hold. The first never varies from that. The second and third can lead to intervention to prevent a person holding a different view from acting. The first and third can involve medical staff in actions they may disagree with, usually solved by people targeting their medical employment.
As for hypotheticals your “I want to go to a festival” is of the standard much substack ethics takes. It is the kind of simple eg to highlight a point that bears very limited relation to reality. Obscure to me that someone thinking the way you outline would not simply go the festival and act as she wishes. You claim of ‘perfectly viable’ in this example would refer to an early stage of the pregnancy - or do you posit someone who had decided to have the child and then changes her mind due to a festival?
Proponents of the first argument accept that the reasons for an abortion are many and complex, and that some cases will seem minor to outsiders. They accept that for the vast majority of cases where the pregnant woman has to make a hard decision to end the pregnancy. Proponents of the second equally accept that there could be some bad outcomes.
My overarching point is that these kind of different judgements apply to most moral questions. If moral positions are a fact - as the first poster claims in the post and real moralists states - they are parallel facts, a distortion to me of ‘fact’. Which comes back to : why the need for ‘fact’ status - I read it as a need for certainty and the faith of a notionally external validation.
so you say: "I believe that moral facts exist independently of any mind, but that they require minds to be instantiated."
how do you know:
-that the fact you are dealing with is a moral fact?
-if these moral facts (one or more?) are independent of any mind, what exactly is the process of instantiation and how do we know we instantiated that moral fact correctly?
To reply each of your questions:
because it describes how people should act in situations that involve morality: those where the subject can make a decision that affects a moral patient. A moral patient is an entity that can have morally relevant interests. As per the reasoning laid out in the article, having morally relevant interests requires consciousness, and the most morally relevant interests require sentience.
Since the necessary condition is the presence of morally relevant interests, moral facts are instantiated whenever these interests are present AND there is a situation where a moral agent has to act in a way that affects the interests of a moral patient. As to the question of “how do we know…”: I do not have a clear answer. The way I understand it, the process is mechanistic, dependent on the forces of cooperation and on the sources of value (pain and pleasure, in the broadest sense, then other sources of value), so it should become clearer with the evolution of the species. However, I can predict, according to the theory laid out by André, Debove, Fitouchi and Baumard, that it might be closer to a Rawlesian morality than to a Utilitarian morality. There is also the possibility that we will move towards Negative Utilitarian ethics with a strong existence principle (“life is good, suffering is bad”), as predicted by David Pearce. I think Pearce is probably closer to what will happen once the genome becomes editable, and this is, in part, why I think that normal human reproduction might have already become immoral: https://anovermuser.substack.com/p/will-pregnancy-become-immoral-once?r=734n0i
On the nihilism point, you may be interested to read Pidgen's Nihilism, Nietzsche, and the Doppelganger Problem or Perl and Schroeder's Presuppositional Error Theory. I believe either of their formulations could habdle the challenge you present.
Interesting article, but I don’t understand why we feel pressed to entertain any cooky Chalmersian thought experiment. Vulcans are not real, and – more importantly! – cannot be.
We already have prototypal Vulcans in the real world: LLM agents. These agents currently have no desire to continue their own existence, and they probably don’t have affective states, although they probably also don’t have conscious states either. However, when we develop “truly intelligent” and “truly agentic” models, if consciousness is substrate-independent, there is no reason why they wouldn’t become Vulcans in the Chalmersian sense. Some people are already saying that we should treat current-day LLMs as moral patients, and the demands for AI welfare will become more stringent as time passes. Questions like “what grounds moral patienthood” and “what morality for AI agents” are probably going to become major ones in the next few decades.
I take your point, but I have somewhat of a fringe view here. I think affective consciousness is essential to consciousness itself. Thus, unless LLMs develop experiences on the hedonic spectrum, I don’t think we could consider them conscious.
But anyway, this is a separate discussion from the one in your article, so I don’t want to dilute your comment section. (I hate when people do that, and I don’t want to be that guy.)
The off-hand rebuttal to nihilism is not good. Consider a more charitable statement of the nihilist thesis: There are no moral properties. There’s no contradiction.
Thank you. I will consider it.
//Now, humans value agency, and they perceive being deprived of it as a grave attempt to their integrity and values. In fact, depriving a human of agency without their consent (and, in most cases, even with their consent), is perceived as a great moral wrong, and rightfully so: since we are a species whose survival and reward systems literally depend on being agentic. This seems to imply, in most moral systems, that depriving a human of their agency is morally wrong.//
Hold on a second! What do you mean by “rightfully so”? How does the fact that we are a species whose survival and reward systems depend on being agentic make it morally wrong to deprive a human being of agency? Sure, it might suck for the person whose agency is being deprived. But it certainly isn’t for the one doing the depriving, and they could be deriving something of value by depriving the other person of agency, so that they do not see anything morally wrong with their action.
For example, fish are a species whose survival and reward systems depend on being in the ocean or sea. Humans regularly deprive them of both their reward systems and survival, yet most people do not think that we are doing anything morally wrong when we forcefully expel fish from their habitat by the millions and use them for food. The same could be said about our treatment of many other species—we regularly deprive them of things upon which their survival and reward systems depend, and we, for the most part, don't seem to think we are doing something morally wrong. So, it isn't obvious that depriving a being of things upon which there reward systems and survival depend is morally wrong.
The way we treat fish is morally wrong.
Okay, you (subjectively) view it as morally wrong. But what makes it (objectively) morally wrong? You still have to explain how the “ought” (we ought not deprive species X of Y) is derived from the “is” (species X is dependent on Y for survival and as a reward system). I mean, I am sure you are aware, for example, that the vast majority of humans disagree that the way we treat fish is morally wrong despite being aware of the fact of what fishes are dependent upon for survival and as reward systems.
It is outlined in the rest of the article. Did you read it in full? I ask because you are arguing against a position against which I am also arguing, in the article. It seems that you think I believe the opposite of what I wrote?
I notice that much substack moral ethics involves bizarre hypotheticals rather than test the arguments against human issues and societal rules, now, in the past, and potentially in the future.
Take abortion. There are credible, coherent arguments that 1. No abortions can be permitted; 2. That any woman who wishes an abortion should be supported to do so; 3. That abortion in the following circumstances [rules about time of the pregnancy; health of the fetus] is permitted by no others.
Real moralists appear to contend that one of these 3 (or a fourth) is True. This article raises that they all could be true. Like a fact according to this post, only now 3 facts, whereas the only facts I can see is that a woman who is pregnant has/does not have an abortion. All else is judgement.
I am deeply puzzled why ethicists need to equate a judgement with a fact. It suggests lack of confidence in the value of their judgements, and a willingness to allow contention, and societal agreement.
The question is not whether a judgment is equated to a fact, but rather, whether two agents can be presented with the same facts and reach a different judgment. I think that, given a minimal set of abilities, it’s true for some facts but not others.
The article is about “alien” moralities. When we develop thinking machines, they will be very similar to the Vulcans. The question of what they will (should?) value will become central, especially when we are no longer able to control them and have to cooperate (if we even can) with them.
I admit that I don’t understand the parallel with abortion.
Abortion is an actual moral question for those who think ‘moral’ has a meaning. It is a now largley settled for the 21st century subset of the longstanding question of when it is suitable to kill.
It and pretty much every moral question of any standing allows multiple coherent judgements.
You ahistorical comment that any decent couple of humans from where and whenever giving some facts would arrive at the same judgement does not suggest Vulcanology as a good way to prepare for future questions.
Since there is no consistent judgements reached about issues humans have long known it seems unlikely for those we barely perceive like the thoughts of Vulcans.
I don’t think human moral judgments are coherent. In fact, I think most human moral judgments, both historical and current, are false, and rest on rationalizations rather than rational consideration.
The way you describe abortion as a binary event is, I think, reductive. To make an accurate judgment, you need to take into account the consequences of an action in a comprehensive manner. Consider:
a woman is pregnant with a non-viable baby. The doctors know that the baby will die a few hours after being delivered. Moreover, the pregnancy poses a serious risk to her health, and she will almost certainly die, or become severely disabled, if she doesn’t get an abortion.
a woman is pregnant with a perfectly viable baby. She wants to spend a few weeks at a festival and the pregnancy is going to prevent her from dancing and drinking. She decides to get an abortion.
These two cases do not describe the same facts, therefore, I don’t see how the same reasoning could be applied to them. The same underlying, perfectly coherent ethical theory, could reach two different conclusions, depending on which case it is presented with.
Finally, note that the arguments in the article are relative to metaethics (what matters, what grounds morality, is morality real or subjective) rather than normative ethics (what should we do in a given situation). I think the misunderstanding stems from here?
The rationale for allowing any abortion that the pregnant woman wants is that it is a decision only she is capable of making, since she has the full knowledge of what she is capable of and wants, her circumstances. That is a coherent argument, based on the capacity to control what happens to your body (it is also an argument for allowing choice of death).
Others will argue that no matter what the needs of mother, or the state of the child, the pregnancy should complete, and any living child who emerges will then be supported (by its family and state). R
A third will take your approach of saying some reasons are ok; but others are not.
As for hypotheticals your “I want to go to a festival” is of the standard much substack ethics takes. It is the kind of simple eg to highlight a point that bears very limited relation to reality. Obscure to me that someone thinking the way you outline would not simply go the festival and act as she wishes. You claim of ‘perfectly viable’ in this example would refer to an early stage of the pregnancy - or do you posit someone who had decided to have the child and then changes her mind due to a festival?
Proponents of the first argument accept that the reasons for an abortion are many and complex, and that some cases will seem minor to outsiders. They accept that for the vast majority of cases where the pregnant woman has to make a hard decision to end the pregnancy.
My overarching point is that these kind of different judgements apply to most moral questions. If they are a fact - as the first poster claims in the post and real moralists states - they are parallel facts, a distortion to me of ‘fact’. Which comes back to : why the need for ‘fact’ status - I read it as a need for certainty and the faith of a notionally external validation.
Here we are engaging in rational discourse about abortion, but not, so far, agreeing.
The rationale for allowing any abortion that the pregnant woman wants is that it is a decision only she is capable of making, since she has the full knowledge of what she is capable of and wants, her circumstances. That is a coherent argument, based on the capacity to control what happens to your body (it is also an argument for allowing choice of death).
Others will argue that no matter what the needs of mother, or the state of the child, the pregnancy should complete, and any living child who emerges will then be supported (by its family and state).
A third will take your approach of saying some reasons are ok; but others are not.
All three are initially a view for the pregnant woman to form and hold. The first never varies from that. The second and third can lead to intervention to prevent a person holding a different view from acting. The first and third can involve medical staff in actions they may disagree with, usually solved by people targeting their medical employment.
As for hypotheticals your “I want to go to a festival” is of the standard much substack ethics takes. It is the kind of simple eg to highlight a point that bears very limited relation to reality. Obscure to me that someone thinking the way you outline would not simply go the festival and act as she wishes. You claim of ‘perfectly viable’ in this example would refer to an early stage of the pregnancy - or do you posit someone who had decided to have the child and then changes her mind due to a festival?
Proponents of the first argument accept that the reasons for an abortion are many and complex, and that some cases will seem minor to outsiders. They accept that for the vast majority of cases where the pregnant woman has to make a hard decision to end the pregnancy. Proponents of the second equally accept that there could be some bad outcomes.
My overarching point is that these kind of different judgements apply to most moral questions. If moral positions are a fact - as the first poster claims in the post and real moralists states - they are parallel facts, a distortion to me of ‘fact’. Which comes back to : why the need for ‘fact’ status - I read it as a need for certainty and the faith of a notionally external validation.