I think something that’s left out in all of this is intentionality. The golden rule has a lot to do with respecting someone’s intentions even if the results we’re not at all what you would want. Conversely, even if the results are in your favor, if the intentionality was “immoral” in your eyes, then it’s not something you should just happily accept. You should question it and try to help others who it affects. You speak about this a lot when you talk about anti-racism and I agree with you.
I agree with you about ‘intentionality’, but we also have to deal with a few other phenomena:
1) Some people take the phrase too literally. I know I’m Autistic, and we tend to take things more literally than Allistic people, but there are times where context is far more important than the literal words being used as a “label” to a larger concept.
2) There are many people who accept training that “human nature” makes everyone’s values and motivations the same, so harm is generated even in situations where the intention wasn’t to cause harm. This makes the golden rule very different than the “One Rule” that is being used in Stretch’s LLM training.
——
Edit: I realized after I wrote this that I missed something important. I don’t believe the “intention” of an individual should be the only level of analysis, but the “intention” of the training (LLM or humans) at a systemic and peoples (Greek word “genos”) level.
Western individualism ends up protecting “freedom from responsibility” as a primary freedom, and I don’t believe Western “Enlightenment” worldviews should itself be used to absolve all responsibility for harm caused by Western “Enlightenment” worldviews.
yeah, that makes a lot of sense. I think it's impossible to truly tell, anyway, so we're less bound to speculation and assumptions, etc., if we go with One Rule...I think, re: your edit, that intentionality can be sussed in live, real time discussions without an audience (or with one) provided both parties let down their guard and speak with compassion and with a shared goal to reduce conflict. Ultimately, conflict is conflict is conflict...we need better assessment of conflict, conflict management, etc.
The meaning of “conflict” is different under different worldviews. Training data is the term used for LLMs, and I’m intending to use “worldviews” as the lens for understanding the world created through the training data of different peoples (Greek word “genos” type groupings, not arbitrary groupings of individuals based on snapshots in time within specific geography/etc).
If we start with current Western “Enlightenment” worldviews (individuality, property=exclusivity without responsibility, Westphalian sovereignty, dueling dualistic religions, etc) you end up with “conflict” being forced into “winner takes all” existential scenarios enforced under Worldview training.
If we alternatively start with Great Lakes/St Lawrence Turtle Island worldviews (collectivity, land/life stewardship where responsibility comes before and often instead of exclusivity, peoples/genos/etc based sovereignty, ethic of non-interference, relational rather than objectification thinking) you end up with “conflict” being ongoing living relationships between peoples. This is how how inter-national and treaty relations were treated on this continent for centuries. My understanding is that this alternative concept of sovereignty has existed far longer than the Eurocentric Westphalian system that came only from 1600’s Europe. Even the Haudenosaunee Confederacy and the “Great law of Peace” (their worldview changing alternative to perpetual war, which the Westphalian system was also designed to offer) dates back to the late 1100’s.
LLMs and humans all receive training data, and that training data is far more useful in determining how each “thinks” than fixating on biology or computer hardware. Meaning, individualism isn’t a useful lens for analysing much – not even for sciences like biology, due to the mythology of neutrality/objectivity.
While I grew up and spent most of my life embedded within Western training data, I am doing my best to try to move outside of that silo. I’m trying to bring in new training data that helps me think in a way that recognizes that other peoples have entirely different worldviews that should be respected rather than generating ideological “winner takes all” scenarios where other worldviews/peoples are constantly being targeted for eradication.
I think the definitions of the morality attached to conflict might change, but I think there's a reasonable argument behind "concepts that are universally experienced by animals, whether they have language or words to convey them back to themselves, or others" so I think it'd be a waste of time to (ironically) enter into conflict over what or how to relate conflict, when it's easier to teach conflict reduction techniques. Like with a dog, you put your fist out, closed, with the top towards it, or you approach with something to eat, as a sign that you are wary of it, and have no "intention" of causing harm. That's the level I'm speaking at. So while I understand and DEEPLY respect your attempt to get out of a silo, there is an actual reasonable understanding of inter, intra species conflict that all of us can work around, within, etc.
Ultimately, prey and predator relationships exist in every biotic community I can think of, so if there are any where it doesn't exist, good for them/it/they, but I think it's off-topic to worry about individualism within the original context of "is the golden rule golden?" We agree that It's not, but I do think it (and the "one rule" which is what we were originally discussing), are still trying to reduce conflict, and I don't think it's a world-view issue.
many non western traditions tell people to accept all conflict, even deadly ones, so i do think it's a universally species-wide grokable concept? I dunno, just my thoughts! :)
I think something that’s left out in all of this is intentionality. The golden rule has a lot to do with respecting someone’s intentions even if the results we’re not at all what you would want. Conversely, even if the results are in your favor, if the intentionality was “immoral” in your eyes, then it’s not something you should just happily accept. You should question it and try to help others who it affects. You speak about this a lot when you talk about anti-racism and I agree with you.
I agree with you about ‘intentionality’, but we also have to deal with a few other phenomena:
1) Some people take the phrase too literally. I know I’m Autistic, and we tend to take things more literally than Allistic people, but there are times where context is far more important than the literal words being used as a “label” to a larger concept.
2) There are many people who accept training that “human nature” makes everyone’s values and motivations the same, so harm is generated even in situations where the intention wasn’t to cause harm. This makes the golden rule very different than the “One Rule” that is being used in Stretch’s LLM training.
——
Edit: I realized after I wrote this that I missed something important. I don’t believe the “intention” of an individual should be the only level of analysis, but the “intention” of the training (LLM or humans) at a systemic and peoples (Greek word “genos”) level.
Western individualism ends up protecting “freedom from responsibility” as a primary freedom, and I don’t believe Western “Enlightenment” worldviews should itself be used to absolve all responsibility for harm caused by Western “Enlightenment” worldviews.
yeah, that makes a lot of sense. I think it's impossible to truly tell, anyway, so we're less bound to speculation and assumptions, etc., if we go with One Rule...I think, re: your edit, that intentionality can be sussed in live, real time discussions without an audience (or with one) provided both parties let down their guard and speak with compassion and with a shared goal to reduce conflict. Ultimately, conflict is conflict is conflict...we need better assessment of conflict, conflict management, etc.
edit: conflict IS not avoidable, but "acting from a place of meeting conflict" IS something all of us can work on, especially at genos level
The meaning of “conflict” is different under different worldviews. Training data is the term used for LLMs, and I’m intending to use “worldviews” as the lens for understanding the world created through the training data of different peoples (Greek word “genos” type groupings, not arbitrary groupings of individuals based on snapshots in time within specific geography/etc).
If we start with current Western “Enlightenment” worldviews (individuality, property=exclusivity without responsibility, Westphalian sovereignty, dueling dualistic religions, etc) you end up with “conflict” being forced into “winner takes all” existential scenarios enforced under Worldview training.
If we alternatively start with Great Lakes/St Lawrence Turtle Island worldviews (collectivity, land/life stewardship where responsibility comes before and often instead of exclusivity, peoples/genos/etc based sovereignty, ethic of non-interference, relational rather than objectification thinking) you end up with “conflict” being ongoing living relationships between peoples. This is how how inter-national and treaty relations were treated on this continent for centuries. My understanding is that this alternative concept of sovereignty has existed far longer than the Eurocentric Westphalian system that came only from 1600’s Europe. Even the Haudenosaunee Confederacy and the “Great law of Peace” (their worldview changing alternative to perpetual war, which the Westphalian system was also designed to offer) dates back to the late 1100’s.
LLMs and humans all receive training data, and that training data is far more useful in determining how each “thinks” than fixating on biology or computer hardware. Meaning, individualism isn’t a useful lens for analysing much – not even for sciences like biology, due to the mythology of neutrality/objectivity.
While I grew up and spent most of my life embedded within Western training data, I am doing my best to try to move outside of that silo. I’m trying to bring in new training data that helps me think in a way that recognizes that other peoples have entirely different worldviews that should be respected rather than generating ideological “winner takes all” scenarios where other worldviews/peoples are constantly being targeted for eradication.
I think the definitions of the morality attached to conflict might change, but I think there's a reasonable argument behind "concepts that are universally experienced by animals, whether they have language or words to convey them back to themselves, or others" so I think it'd be a waste of time to (ironically) enter into conflict over what or how to relate conflict, when it's easier to teach conflict reduction techniques. Like with a dog, you put your fist out, closed, with the top towards it, or you approach with something to eat, as a sign that you are wary of it, and have no "intention" of causing harm. That's the level I'm speaking at. So while I understand and DEEPLY respect your attempt to get out of a silo, there is an actual reasonable understanding of inter, intra species conflict that all of us can work around, within, etc.
Ultimately, prey and predator relationships exist in every biotic community I can think of, so if there are any where it doesn't exist, good for them/it/they, but I think it's off-topic to worry about individualism within the original context of "is the golden rule golden?" We agree that It's not, but I do think it (and the "one rule" which is what we were originally discussing), are still trying to reduce conflict, and I don't think it's a world-view issue.
many non western traditions tell people to accept all conflict, even deadly ones, so i do think it's a universally species-wide grokable concept? I dunno, just my thoughts! :)