Yoav Ravid | v1.3.0Nov 16th 2021 | (+50/-100) Small Fixes | ||
Ruben Bloom | v1.2.0Oct 1st 2020 | (+2864/-90) | ||
Multicore | v1.1.0Aug 1st 2020 | (+90/-2865) | ||
Joao Fabiano | v0.0.5Nov 24th 2014 | (+9/-10) /* Metaethics on Less Wrong */ | ||
Joao Fabiano | v0.0.4Nov 24th 2014 | (+142/-10) | ||
Eliezer Yudkowsky | v0.0.3Jan 15th 2013 | (+1104/-2087) for this page to be useful as a reference, it needs to plunge directly into the topic | ||
pedrochaves | v0.0.2Sep 29th 2012 | (+93/-18) | ||
pedrochaves | v0.0.1Sep 28th 2012 | (+2740/-390) | ||
TerminalAwareness | v0.0.0Jul 25th 2012 | (+1336) First part |
Metaethicsis inquiryone of the three branches of ethics usually recognized by philosophers, the others being normative ethics and applied ethics. It’s a field of study that tries to understand the metaphysical, epistemological and semantic characteristics as well as the foundations and scope of moral values. It worries about questions and problems such as "Are moral judgments objective or subjective, relative or absolute?", "Do moral facts exist?" or “How do we learn moral values?”. (As distinct from object-level moral questions like, "Ought I to steal from banks in order to give the money to the deserving poor?")
Eliezer Yudkowsky wrote a Sequence about metaethics, the Metaethics sequence, which Yudkowsky worried failed to convey his central point; he approached the same problem again from a different angle in Highly Advanced Epistemology 101 for Beginners. From a standard philosophical standpoint, Yudkowsky's philosophy is closest to Frank Jackson's moral functionalism / analytic descriptivism; Yudkowsky could be loosely characterized as moral cognitivist - someone who believes moral sentences are either true or false - but not a moral realist - thus denning that moral sentences refer to facts about the philosophical foundations underlyingworld. Yudkowsky believes that moral systems. Whycognition in any single human is anything right or wrongat least potentially about a subject matter that is 'logical' in the first place?sense that its semantics can be pinned down by axioms, and hence that moral cognition can bear truth-values; also that human beings both using similar words like "morality" can be talking about highly overlapping subject matter; but not that all possible minds would find the truths about this subject matter to be psychologically compelling.
Luke Muehlhauser has written a sequence, No-Nonsense Metaethics, where he claims that many of the questions of metaethics can be answered today using modern neuroscience and rationality. He explains how conventional metaethics or "Austere Metaethics" is capable of, after assuming a definition of 'right', choosing the right action given a situation - but useless without assuming some criteria for 'right'. He proposes instead "Empathic Metaethics" which utilizes your underlying cognitive algorithms to understand what you think 'right' means, helps clarify any emotional and cognitive contradictions in it, and then tells you what the right thing to do is, according to your definition of right. This approach is highly relevant for the Friendly AI problem as a way of defining human-like goals and motivations when designing AIs.
Metaethicsis one of the three branches of ethics usually recognized by philosophers, the others being normative ethics and applied ethics. It’s a field of study that tries to understand the metaphysical, epistemological and semantic characteristics as well as the foundations and scope of moral values. It worries about questions and problems such as "Are moral judgments objective or subjective, relative or absolute?", "Do moral facts exist?" or “How do we learn moral values?”. (As distinct from object-level moral questions like, "Ought I to steal from banks in order to give the money to the deserving poor?")
Eliezer Yudkowsky wrote a Sequence about metaethics, the Metaethics sequence, which Yudkowsky worried failed to convey his central point; he approached the same problem again from a different angle in Highly Advanced Epistemology 101 for Beginners. From a standard philosophical standpoint, Yudkowsky's philosophy is closest to Frank Jackson's moral functionalism / analytic descriptivism; Yudkowsky could be loosely characterized as moral cognitivist - someone who believes moral sentences are either true or false - but not a moral realist - thus denning that moral sentences refer to factsinquiry about the world. Yudkowsky believes thatphilosophical foundations underlying moral cognition in any single humansystems. Why is at least potentially about a subject matter that is 'logical'anything right or wrong in the sense that its semantics can be pinned down by axioms, and hence that moral cognition can bear truth-values; also that human beings both using similar words like "morality" can be talking about highly overlapping subject matter; but not that all possible minds would find the truths about this subject matter to be psychologically compelling.first place?
Luke Muehlhauser has written a sequence, No-Nonsense Metaethics, where he claims that many of the questions of metaethics can be answered today using modern neuroscience and rationality. He explains how conventional metaethics or "Austere Metaethics" is capable of, after assuming a definition of 'right', choosing the right action given a situation - but useless without assuming some criteria for 'right'. He proposes instead "Empathic Metaethics" which utilizes your underlying cognitive algorithms to understand what you think 'right' means, helps clarify any emotional and cognitive contradictions in it, and then tells you what the right thing to do is, according to your definition of right. This approach is highly relevant for the Friendly AI problem as a way of defining human-like goals and motivations when designing AIs.
Eliezer Yudkowsky wrote a Sequence about metaethics, the Metaethics sequence, which Yudkowsky worried failed to convey his central point; he approached the same problem again from a different angle in Highly Advanced Epistemology 101 for Beginners. From a standard philosophical standpoint, Yudkowsky's philosophy is closest to Frank Jackson's moral functionalism / analytic descriptivism; Yudkowsky could be loosely characterized as moral cognitivist - someone who believes moral sentences are either true or false - but not a moral realist - thus denning that moral statementssentences refer to facts about the world. Yudkowsky believes that moral cognition in any single human is at least potentially about a subject matter that is 'logical' in the sense that its semantics can be pinned down by axioms, and hence that moral cognition can bear truth-values; also that human beings both using similar words like "morality" can be talking about highly overlapping subject matter; but not that all possible minds would find the truths about this subject matter to be psychologically compelling.
Eliezer Yudkowsky wrote a Sequence about metaethics, the Metaethics sequence, which Yudkowsky worried failed to convey his central point; he approached the same problem again from a different angle in Highly Advanced Epistemology 101 for Beginners. From a standard philosophical standpoint, Yudkowsky's philosophy is closest to Frank Jackson's moral functionalism / analytic descriptivism; Yudkowsky could be loosely characterized as a moral cognitivist - someone who believes moral sentences are either true or false - but not a moral realist.realist - thus denning that moral statements refer to facts about the world. Yudkowsky believes that moral cognition in any single human is at least potentially about a subject matter that is 'logical' in the sense that its semantics can be pinned down by axioms, and hence that moral cognition can bear truth-values; also that human beings both using similar words like "morality" can be talking about highly overlapping subject matter; but not that all possible minds would find the truths about this subject matter to be psychologically compelling.
Metaethics is one of the three branches of ethics usually recognized by philosophers, the others being normative ethics and applied ethics. It’s a field of study that tries to understand the metaphysical, epistemological and semantic characteristics as well as the foundations and scope of moral values. It worries about questions and problems such as "Are moral judgments objective or subjective, relative or absolute?", "Do moral facts exist?" or “How do we learn moral values?”. (As distinct from object-level moral questions like, "Ought I to steal from banks in order to give the money to the deserving poor?")
The metaethical positions are defined byEliezer Yudkowsky wrote a Sequence about metaethics, the way authors tryMetaethics sequence, which Yudkowsky worried failed to answerconvey his central point; he approached the mentioned questions. Garner & Rosen (1967) have proposedsame problem again from a division of such theories based on three major concepts or types of questions: Semantic, Substantial and Justificationdifferent angle in Highly Advanced Epistemology 101 for Beginners.
The first group of theories deals with questions pertaining From a standard philosophical standpoint, Yudkowsky's philosophy is closest to Frank Jackson's moral functionalism / analytic descriptivism; Yudkowsky could be loosely characterized as a moral cognitivist but not a moral realist. Yudkowsky believes that moral cognition in any single human is at least potentially about a subject matter that is 'logical' in the meaning of moral terms. That is, it tries to solve questions such as "What does right or wrong actually mean?" or "What do people refer to when they say something is right?". Itsense that its semantics can be further divided into Cognitivepinned down by axioms, and Non-cognitive theories, regarding whether we considerhence that moral values to have a logical structure, expressing truth beliefs, or simply represent emotional responses.
The substantial theories, on the other hand, try tocognition can bear truth-values; also that human beings both using similar words like "morality" can be talking about highly overlapping subject matter; but not that all possible minds would find the answertruths about this subject matter to the nature of moral judgements. That is, they try to answer questions such as "Are moral values universal, or do they vary from context to context?". This theories can be grouped in three major clusters. Moral universalism postulates that a universal system of ethics is the foundation of all moral judgements and, as such, moral values are absolut. On the opposite side, Moral relativism introduces the idea that these values vary from culture to culture and are sensitive to context - there is no absolute standard. Parallel to both these kinds of theories emerges Moral nihilism, asserting that there are no actions or decisions better than others, right and wrong being...
Metaethics is one of the three branches of ethics usually recognized by philosophers, the otherothers being normative ethics and applied ethics. It’s a field of study that tries to understand the metaphysical, epistemological and semantic characteristics as well as the foundations and scope of moral values. It worries about questions and problems such as "Are moral judgments objective or subjective, relative or absolute?", "Do moral facts exist?" or “How do we learn moral values?”.
Luke Muehlhauser has begun a sequence, No-Nonsense Metaethics, where he claisclaims that many of the questions of metaethics can be answered today using modern neuroscience and rationality. He explains how conventional metaethics or "Austere Metaethics" is capable of, after assuming a definition of 'right', choosing the right action given a situation - but useless without assuming some criteria for 'right'. He proposes instead "Empathic Metaethics" which utilizes your underlying cognitive algorithms to understand what you think 'right' means, helps clarify any emotional and cognitive contradictions in it, and then tells you what the right thing to do is, according to your definition of right. This approach is highly relevant for the Friendly AI problem.problem as a way of defining human-like goals and motivations when designing AIs.
Metaethics is one of the branchthree branches of ethics devotedusually recognized by philosophers, the other being normative ethics and applied ethics. It’s a field of study that tries to understandingunderstand the naturemetaphysical, epistemological and semantic characteristics as well as the foundations and scope of the components of ethics.moral values. It seeks answers forworries about questions likeand problems such as "Are moral judgments objective or subjective, relative or absolute?" and, "Do moral facts exist?", termed second-order or “How do we learn moral questions. It is a field in rapid development, permittedvalues?”.
The metaethical positions are defined by the developing fieldway authors try to answer the mentioned questions. Garner & Rosen (1967) have proposed a division of Neurosciencesuch theories based on three major concepts or types of questions: Semantic, Substantial and Justification.
The first group of theories deals with questions pertaining the meaning of moral terms. That is, it tries to solve questions such as "What does right or wrong actually mean?" or "What do people refer to when they say something is right?". Rationality is well suitedIt can be further divided into Cognitive and Non-cognitive theories, regarding whether we consider moral values to analyzinghave a logical structure, expressing truth beliefs, or simply represent emotional responses.
The substantial theories, on the field of metaethics. Metaethics is also criticalother hand, try to find the answer to the developmentnature of moral judgements. That is, they try to answer questions such as "Are moral values universal, or do they vary from context to context?". This theories can be grouped in three major clusters. Moral universalism postulates that a Friendly Artificial Intelligenceuniversal system of ethics is the foundation of all moral judgements and, as such, moral values are absolut. On the opposite side, Moral relativism introduces the idea that these values vary from culture to culture and are sensitive to context - there is no absolute standard. Parallel to both these kinds of theories emerges Moral nihilism, asserting that there are no actions or decisions better than others, right and wrong being non-existent.
Finally, justification theories are worried with how we can justify moral positions or the the need to act morally. It's mainly an epistemological field that tries to solve the problem of deciding whether we can have actual knowledge about moral truths. If we deny such fact, we are faced with Moral skepticism, which asserts no one can have access to knowledge about moral decisions - thus encompassing most non-cognitive theories. On the other hand, if we assume knowledge is possible, we are faced with theories claiming we can gain it through inference - Empiricism or Ethical rationalism - or without it, with an Ethical intuitionism.
Luke Muehlhauser has begun a sequence, No-Nonsense Metaethics, and has claimedwhere he clais that many of the questions of metaethics can be answered today using modern neuroscience and rationality. He explains how conventional metaethics or "Austere Metaethics" is capable of, after assuming a definition of 'right', choosing the right action given a situation - but useless without assuming some criteria for 'right'. He proposes instead "Empathic Metaethics" which utilizes your underlying cognitive algorithms to understand what you think 'right' means, helps clarify any emotional and cognitive contradictions in it, and then tells you what the right thing to do is, according to your definition of right. This approach is highly relevant for the Friendly AI problem.
Metaethics is the branch of ethics devoted to understanding the nature of the components of ethics. It seeks answers for questions like "Are moral judgments objective or subjective, relative or absolute?" and "Do moral facts exist?", termed second-order moral questions. It is a field in rapid development, permitted by the developing field of Neuroscience. Rationality is well suited to analyzing the field of metaethics. Metaethics is also critical to the development of a Friendly Artificial Intelligence.
Luke Muehlhauser has begun a sequence, No-Nonsense Metaethics, and has claimed many of the questions of metaethics can be answered today using modern neuroscience and rationality. He explains how conventional metaethics or "Austere Metaethics" is capable of, after assuming a definition of 'right', choosing the right action given a situation - but useless without assuming some criteria for 'right'. He proposes instead "Empathic Metaethics" which utilizes your underlying cognitive algorithms to understand what you think 'right' means, helps clarify any emotional and cognitive contradictions in it, and then tells you what the right thing to do is, according to your definition of right.
Eliezer Yudkowsky wrote a Sequence about metaethics, the Metaethics sequence, which Yudkowsky worried failed to convey his central
point;point (this post by Luke tried to clarify); he approached the same problem again from a different angle in Highly Advanced Epistemology 101 for Beginners. From a standard philosophical standpoint, Yudkowsky's philosophy is closest to Frank Jackson's moral functionalism / analytic descriptivism; Yudkowsky could be loosely characterized as moral cognitivist - someone who believes moral sentences are either true or false - but not a moral realist - thusdenningdenying that moral sentences refer to facts about the world. Yudkowsky believes that moral cognition in any single human is at least potentially about a subject matter that is 'logical' in the sense that its semantics can be pinned down by axioms, and hence that moral cognition can bear truth-values; also that human beings both using similar words like "morality" can be talking about highly overlapping subject matter; but not that all possible minds would find the truths about this subject matter to be psychologically compelling.Metaethics sequenceNo-Nonsense MetaethicsHighly Advanced Epistemology 101 for Beginners