{ "export": { "source": "https://urth.darkrealm.vip/thread/urth/4348", "note": "Exported from an unofficial mirror of a public mailing list. Copyright in each message remains with the person who wrote it. Email addresses are obscured as `user at host`, the form the original archive published. This file was rebuilt from parsed fields rather than retained headers, so it is faithful in content but not byte-exact.", "addresses": "obscured as 'user at host'" }, "archive": "urth", "thread": { "id": 4348, "subject": "5HC : Chinese boxes or tea chests?", "messages": 23, "voices": 8, "first": "2005-02-01T07:47:50+00:00", "last": "2005-02-03T11:21:26+00:00" }, "messages": [ { "message_id": "<20050201154712.13264.qmail@web25703.mail.ukl.yahoo.com>", "from": { "name": "Jim Raylor", "address": "rjraylor at yahoo.co.uk" }, "date": "2005-02-01T07:47:50+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": null, "reply_link": null, "blocks": [ { "kind": "text", "depth": 0, "text": "Searle's argument was that the guy in the room who\nmanipulated the symbols had no real knowledge of the\nChinese language, just the rules for manipulating the\nsymbols.\nIt could be argued that all language users 'merely'\nknow how to manipulate the symbols without a full\nawareness of their meanings.\nI think this when I use words like 'Notwithstanding'\nand 'Nevertheless'.\nI sometimes wonder if interpreters at the UN (for\nexample) pay any heed to the meaning of the messages\nthey translate, beyond what is required to translate\nthe words.\nFor example an eccentric translator could perversely\ntranslate a book by cutting it into individual\nsentences, translating each one in random order and\nthen reassembling them in the original order. \n\nI seem to recall a program which regularly fooled\npeople into thinking it was a client-centred\npsychotherapist.\n\nHow do you feel about a program which regularly fooled\npeople into thinking it was a client-centred psychotherapist?\n\n=====\nhttp://prisoner143.50megs.com/\n\nI am not a number I am a free person!\n\n\n\t\n\t\n\t\t" } ] }, { "message_id": "<002901c50877$e5604830$a51b0101@gensym.com>", "from": { "name": "David Lebling", "address": "dlebling at hyraxes.com" }, "date": "2005-02-01T08:06:09+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<20050201154712.13264.qmail@web25703.mail.ukl.yahoo.com>", "reply_link": "header", "blocks": [ { "kind": "quote", "depth": 1, "text": "From: \"Jim Raylor\" \nHow do you feel about a program which regularly fooled\npeople into thinking it was a client-centred psychotherapist?" }, { "kind": "text", "depth": 0, "text": "Joe Weizenbaum wept." } ] }, { "message_id": "<1f7617370502011143170eb5c8@mail.gmail.com>", "from": { "name": "Dan'l Danehy-Oakes", "address": "danldo at gmail.com" }, "date": "2005-02-01T11:43:27+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<20050201154712.13264.qmail@web25703.mail.ukl.yahoo.com>", "reply_link": "header", "blocks": [ { "kind": "quote", "depth": 1, "text": "On Tue, 1 Feb 2005 15:47:12 +0000 (GMT), Jim Raylor\n wrote:\nSearle's argument was that the guy in the room who\nmanipulated the symbols had no real knowledge of the\nChinese language, just the rules for manipulating the\nsymbols." }, { "kind": "text", "depth": 0, "text": "This is a significant point, because the neurons in my brain \nhave no real knowledge of the English language, just the \nrules for manipulating the symbols. Thus, the neurons do \nnot \"know English.\" But _I_ appear to.\n\nSimilarly, the man in the Chinese Room does not Chinese, \nbut the system-as-a-whole appears to. To say \"The Room \ndoes not know Chinese,\" Searle must, _at a minimum_,\n\n1. Define what it means to \"know Chinese,\"\n2. Propose an empirical test for the definition proposed\n in #1, and\n3. Demonstrate (from hypothesis, since no such Room\n exists) that the Room does not fit the definition \n proposed in #1.\n\nSearle has not, to the best of my knowledge, proposed \neither a definition of \"knowing Chinese\" or an empirical \ntest for determining whether the Room actually knows \nChinese. He has instead relied on his readers' paralogical \nemotional response to assure us that it does not. But that \ndoes not mean that it does not." }, { "kind": "quote", "depth": 1, "text": "I seem to recall a program which regularly fooled\npeople into thinking it was a client-centred\npsychotherapist." }, { "kind": "text", "depth": 0, "text": "That would be \"Eliza.\" It didn't really fool very many\npeople or for very long.\n\n--Dan'l" }, { "kind": "sig", "depth": 0, "text": "\"We're going to sit on Scorsese's head\"\n -- The Goodfeathers" } ] }, { "message_id": "<021d01c5089e$9f351bc0$4c1e4b0a@vcorp.ad.vrsn.com>", "from": { "name": "James Wynn", "address": "thewynns at earthlink.net" }, "date": "2005-02-01T12:43:04+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<1f7617370502011143170eb5c8@mail.gmail.com>", "reply_link": "header", "blocks": [ { "kind": "quote", "depth": 1, "text": "Searle has not, to the best of my knowledge, proposed\neither a definition of \"knowing Chinese\" or an empirical\ntest for determining whether the Room actually knows\nChinese. He has instead relied on his readers' paralogical\nemotional response to assure us that it does not. But that\ndoes not mean that it does not." }, { "kind": "text", "depth": 0, "text": "Yes, but to be fair Turing's criteria is just as subjective. Roughly, \"can a\ncomputer succeed in impersonating a human person well enough to fool\nhumans.\" Where is the \"control\" here? Mr. Raylor has brought up an excellent\npoint: What if a computer can succeed in fooling someone that it is\nsomething we KNOW it is not? That would be a useful control, I think. Is\nsuccessful mimicry really the final criteria for personhood for a machine?" }, { "kind": "quote", "depth": 1, "text": "From bird's point of view, many bugs succeed in fooling them that they are\nthorns, leaves and sticks. To the mother bird, the coo-coo is her child." }, { "kind": "text", "depth": 0, "text": "Another possible control that someone has pointed out is that a computer\nconceivably might be conscious and might think and NOT seem at all human in\nits responses. Such a computer would not test positive for the Turing test.\nI used to subscribe to AI Magazine but I'm not much of an authority on this.\nHow does the Turing criteria eliminate for an especially good mimic (to\nbring this discussion back to V.R.T)?\n\nWolfe, cleverly, has upped the ante on the Turing test by making the\n\"simulation\" much more elaborate:\n\n1) Mr. Million (referred to a \"simulation\" as a deliberate foil to V.R.T's\n\"simulation\") and Rose and the Whorl gods who are much the same thing in\nanother form\n\n2) The Lemur clan and Potto (are the chem bodies whose bodies they have\ncommandeered walkie-talkies as they apparently intended or are they only\nsimulating the psyches of the Ayuntamiento?)\n\n3) What does it mean that the souls of the inhumi look very similar to the\nphysical bodies of their hosts?\n\n~ Crush" } ] }, { "message_id": "", "from": { "name": "Maru Dubshinki", "address": "marudubshinki at gmail.com" }, "date": "2005-02-01T12:53:25+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<1f7617370502011143170eb5c8@mail.gmail.com>", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "The two of you are probably thinking of a) Parry; an improved Eliza\nwhich was a lot more effective and b) some experiments where\nunsuspecting students talked to a program similar to Eliza via a\nterminal and never suspected anything (that is, they were not told to\napply all their critical spirit to it, like they would in a real\nTuring test.).\n\n~Maru\nMicrosoft delenda est." }, { "kind": "quote", "depth": 2, "text": "On Tue, 1 Feb 2005 11:43:20 -0800, Dan'l Danehy-Oakes wrote:\nI seem to recall a program which regularly fooled\npeople into thinking it was a client-centred\npsychotherapist." }, { "kind": "quote", "depth": 1, "text": "That would be \"Eliza.\" It didn't really fool very many\npeople or for very long.\n\n--Dan'l\n\n--\n\"We're going to sit on Scorsese's head\"\n -- The Goodfeathers\n_______________________________________________" } ] }, { "message_id": "", "from": { "name": "Maru Dubshinki", "address": "marudubshinki at gmail.com" }, "date": "2005-02-01T12:55:39+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<20050201154712.13264.qmail@web25703.mail.ukl.yahoo.com>", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "Of course they do: context matters. When translating it's not a\nmatter of just swapping syntactic structures, grammar, and voabulary-\nif it was, machine translation woulda been licked a long time ago. \nThose are low level rules- there are rules that govern how the story\nacts, and how the meaning units interact. Contextual rules, not just\nsyntactic rules, if that makes sense.\n\n~Maru\nMicrosoft delenda est." }, { "kind": "quote", "depth": 1, "text": "On Tue, 1 Feb 2005 15:47:12 +0000 (GMT), Jim Raylor\n wrote:\nI sometimes wonder if interpreters at the UN (for\nexample) pay any heed to the meaning of the messages\nthey translate, beyond what is required to translate\nthe words.\nFor example an eccentric translator could perversely\ntranslate a book by cutting it into individual\nsentences, translating each one in random order and\nthen reassembling them in the original order.\n\nI seem to recall a program which regularly fooled\npeople into thinking it was a client-centred\npsychotherapist.\n\nHow do you feel about a program which regularly fooled\npeople into thinking it was a client-centred psychotherapist?\n\n=====\nhttp://prisoner143.50megs.com/\n\nI am not a number I am a free person!\n\n\n___________________________________________________________\nALL-NEW Yahoo! Messenger - all new features - even more fun! http://uk.messenger.yahoo.com\n_______________________________________________" } ] }, { "message_id": "<006201c508aa$3167dcc0$a51b0101@gensym.com>", "from": { "name": "David Lebling", "address": "dlebling at hyraxes.com" }, "date": "2005-02-01T14:05:43+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<1f7617370502011143170eb5c8@mail.gmail.com>", "reply_link": "header", "blocks": [ { "kind": "quote", "depth": 1, "text": "From: \"Dan'l Danehy-Oakes\" \nThat would be \"Eliza.\" It didn't really fool very many\npeople or for very long." }, { "kind": "text", "depth": 0, "text": "Indeed it was Eliza, written by Joseph Weizenbaum. For a long discussion of\nthe reaction to the program, see his book \"Computer Power and Human Reason,\"\na passionate critique of \"strong AI.\"\n\nEliza was widely distributed as \"Doctor,\" which imitated a Rogerian\npsychotherapist. Parry imitated (in my opinion in an even cruder way than\nEliza's effort) a paranoid schizophrenic. There is a classic dialog between\nthe two which is easy to Google for (try \"parry\" and \"doctor\").\n\nIt does not take long to see that either is not human, or able to understand\nEnglish in even a very crude way, but many people did not, for various\nreasons, most generally from the social surround of how the program was\npresented. If the source of the answers is described as a person, many\npeople will believe that it is in fact a person, until the evidence to the\ncontrary is overwhelming.\n\nWhat lessons about 5HC we may draw from this experience I hesitate to guess.\n\n-- Dave Lebling, aka vizcacha" } ] }, { "message_id": "<024101c508af$9e5b26c0$4c1e4b0a@vcorp.ad.vrsn.com>", "from": { "name": "James Wynn", "address": "thewynns at earthlink.net" }, "date": "2005-02-01T14:44:44+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<021d01c5089e$9f351bc0$4c1e4b0a@vcorp.ad.vrsn.com>", "reply_link": "header", "blocks": [ { "kind": "quote", "depth": 2, "text": "Yes, but to be fair Turing's criteria is just as subjective." }, { "kind": "quote", "depth": 1, "text": "Ummm, no. It _is_ subjective - or, rather, it is based on subjective\njudgments - but it provides a definition (admittedly, an ostensive\nrather than a descriptive definition), and a way of testing for\nwhether that definition has been met. Searle (to the best of\nmy memory) does neither." }, { "kind": "text", "depth": 0, "text": "But since a camouflage device can always get better and better (particularly\nsince our A.I.'s creators can experiment to discover the chinks in its\ndisguise) it is a standard that can likely be met without ever addressing\nwhether the definition is valid (just as the mother bird's criteria for\nwhether the coo-coo is her child is that it was hatched in her nest.\nSearle's analogy begs all sorts of questions, but it is a set up so that the\ndefinitions are implicit: the person in the box represents the expert system\nA.I. and it \"acts\" without \"thinking\". Remember, the original question that\nTuring was addressing was \"can a computer 'think'?\"." }, { "kind": "quote", "depth": 2, "text": "Roughly, \"can a computer succeed in impersonating a human\nperson well enough to fool humans.\" Where is the \"control\" here?" }, { "kind": "quote", "depth": 1, "text": "Read Turing's article. The \"imitation game\" was to be played by\ntwo players, a human and a computer, with a second human\njudging between them - \"which is the human and which is the\ncomputer?\" The test is, of course, blind - the judge doesn't know\nwhich player is which. (In a really rigorous test, I would suggest that\nruns be done with all possible combinations (not only human vs\ncomputer but human vs human and computer vs computer.) Both\nplayers are trying to convince the judge that they are the human.\nThus, the responses of the human player provides the control to\nwhich the responses of the computer player are compared." }, { "kind": "text", "depth": 0, "text": "That comes out to little more than an opinion poll if one assume that humans\nare generally equal in their ability to be duped. Not much of a \"control\".\nOn the other hand, even if a computer were conscious, a computer\nprogrammer who designed such systems would probably always be able\nto identify certain algorithms to recognize it.\n\n~ Crush" } ] }, { "message_id": "", "from": { "name": "Chris", "address": "rasputin_ at hotmail.com" }, "date": "2005-02-01T16:42:09+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<20050201154712.13264.qmail@web25703.mail.ukl.yahoo.com>", "reply_link": "header", "blocks": [ { "kind": "quote", "depth": 1, "text": "It could be argued that all language users 'merely'\nknow how to manipulate the symbols without a full\nawareness of their meanings." }, { "kind": "text", "depth": 0, "text": "Well, kinda. If you did make an argument that all language users 'merely' \nknow how to manipulate symbols, then you would need a matching conception of \n\"meaning\". (Such theories exist). In such an argument, there's simply isn't \nany \"deeper meaning\" for language users to lack awareness of." }, { "kind": "quote", "depth": 1, "text": "I sometimes wonder if interpreters at the UN (for\nexample) pay any heed to the meaning of the messages\nthey translate, beyond what is required to translate\nthe words." }, { "kind": "text", "depth": 0, "text": "If they didn't, they wouldn't be useful translators. Ambiguities don't map \none-to-one in different language; in order to know which French word to use \nin a translation of an ambiguous English word, you have to work out the \nprobable meaning of the expression it occurs in first.\n\nAlso the vast majority of typical communication is also not quite literal. \nMost of the time you intend to express something more than, or something \ndifferent from, the literal meanings of the words you use. Preserving this \nin a word-for-word translation, even without ambiguities, would be unlikely.\n\nTranslation is hard." }, { "kind": "quote", "depth": 1, "text": "I seem to recall a program which regularly fooled\npeople into thinking it was a client-centred\npsychotherapist.\n\nHow do you feel about a program which regularly fooled\npeople into thinking it was a client-centred psychotherapist?" }, { "kind": "text", "depth": 0, "text": "I think this was Dr. Sbaitso, although Dr. Sbaitso might well just be a \nvariant of Eliza (Dan'l's suggestion).\n\nCivet" } ] }, { "message_id": "", "from": { "name": "Chris", "address": "rasputin_ at hotmail.com" }, "date": "2005-02-01T17:33:10+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<021d01c5089e$9f351bc0$4c1e4b0a@vcorp.ad.vrsn.com>", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "Dan'l said:" }, { "kind": "quote", "depth": 2, "text": "Searle has not, to the best of my knowledge, proposed\neither a definition of \"knowing Chinese\" or an empirical\ntest for determining whether the Room actually knows\nChinese. He has instead relied on his readers' paralogical\nemotional response to assure us that it does not. But that\ndoes not mean that it does not." }, { "kind": "text", "depth": 0, "text": "Well, hold up for a second. Empirical tests aren't necessary, because this \nis a thought problem, and in thought problems you can actually *stipulate* \nsuch things; of course, this means that what you end up with is a conclusion \nthat a hypothesis is either logically consistent or inconsistent, it doesn't \nmean that an actual instance of what you describe in the thought problem can \nor will happen.\n\nI think it was Searle's intention to stipulate the layout of the Room such \nthat not only does it not \"know Chinese\", but that it *does appear* to \"know \nChinese\". Unfortunately the layout he describes doesn't really fit the bill, \nfor one thing. But really there are a multitude of problems here. Granting \nthat he does have a definition of what it is to understand a language \n(presumably laid out somewhere in the large body of work he's done in the \nphilosophy of language), for example, doesn't necessarily compel you to \nagree with that definition. And so on and so on and so on.\n\nCrush said:" }, { "kind": "quote", "depth": 1, "text": "Yes, but to be fair Turing's criteria is just as subjective. Roughly, \"can \na\ncomputer succeed in impersonating a human person well enough to fool\nhumans.\" Where is the \"control\" here? Mr. Raylor has brought up an \nexcellent\npoint: What if a computer can succeed in fooling someone that it is\nsomething we KNOW it is not? That would be a useful control, I think." }, { "kind": "text", "depth": 0, "text": "And I think this was Searle's intention in trying to lay out the Chinese \nRoom the way he did: to raise this as a logical possibility. Personally I \ndon't think he succeeded with his example, but even if he did succeed, what \nare the implications? Well, to start with it would mean that obviously you \ncan't *know* that something is intelligent just because it behaves \n\"intelligently\". And if this is the case then a strict behaviorist would \nhave a problem. (It is unclear to me but I get the general impression that \nSearle thinks that Turing's position is necessarily behaviorist, and I don't \nthink it really is.)" }, { "kind": "quote", "depth": 1, "text": "Is successful mimicry really the final criteria for personhood for a \nmachine?" }, { "kind": "text", "depth": 0, "text": "This is something I have been trying to get at here. There are two questions \nto be answered, an \"is\" question and an \"ought\" question. If Searle is \ncorrect the \"is\" question (is the machine intelligent?) can't be answered by \nexamination of behavior alone (or possibly can't be answered at all). \nHowever, \"personhood\" is really an \"ought\" question. How ought a machine \nthat appears to behave intelligently be treated, what rights should it have, \netc?\n\nI would say that Turing is really answering this \"ought\" question: a machine \nthat appears to be conscious should be granted personhood even if you're \nunsure of the truth of whether it \"is\" intelligent. And I think it is even \nmore clear that this is the position that Wolfe holds with the example of \nRose." } ] }, { "message_id": "", "from": { "name": "Maru Dubshinki", "address": "marudubshinki at gmail.com" }, "date": "2005-02-01T17:40:51+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "I think what he said is even more fundamental than that-(stop me if\nI've said this before). I think what Turing is saying, is that if you\ncan't tell the difference in output, can't make statistically\nmeaningful differentiation, than you have no choice but to grant it\npersonhood, because you will *never* find a better method, a better\ntest.\n\n~Maru\nNow here's an interesting thought problem- postulate a vast group of\nrobots who can pass the Turing Test, and mimic a bright, inventive\nyoung engineer so well nobody can tell the difference through a\nscreen, and who are also reasonably physically handy. Suppose every\nreal 'human' died off and all that is left are robots w/0 \"real\ncounsciousness\". What would happen if they discovered an asteroid\ncoming, and the only possible way of survival was to blow it up with a\nnuke? Suppose further that they made like a human, solving all\ntechnical problems, and an advanced alien civilization came by\nimpressed by what they had done. Would the aliens accord the robots\n'personhood' and would they be right or wrong?\n\nOn Wed, 02 Feb 2005 01:32:45 +0000, Chris wrote:\n...." }, { "kind": "quote", "depth": 1, "text": "I would say that Turing is really answering this \"ought\" question: a machine\nthat appears to be conscious should be granted personhood even if you're\nunsure of the truth of whether it \"is\" intelligent. And I think it is even\nmore clear that this is the position that Wolfe holds with the example of\nRose." } ] }, { "message_id": "", "from": { "name": "Chris", "address": "rasputin_ at hotmail.com" }, "date": "2005-02-01T23:09:02+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "Maru said:" }, { "kind": "quote", "depth": 1, "text": "Suppose further that they made like a human, solving all\ntechnical problems, and an advanced alien civilization came by\nimpressed by what they had done. Would the aliens accord the robots\n'personhood' and would they be right or wrong?" }, { "kind": "text", "depth": 0, "text": "You've stipulated they don't have \"real consciousness\", but the important \nunknown is: do the aliens have the ability to tell the difference, or if \nthey are uncertain, do they have an estimate of probability for the robots \nhaving \"real consciousness\"?\n\nFor fun, a few possibilities:\n\nKantian deontological Martians - treat the robots as persons if there is \neven the tiniest finite chance of them being conscious, because any attempt \nto universally apply a maxim which treats a being with a finite chance of \nbeing a person as a non-person will eventually use at least one person as a \nmeans and thus violates the Categorical Imperative.\n\nUtilitarian Neptunians - attempt to calculate the relative utilities of \ntreating the robots as persons vs. taking advantage of them, weighted for \nrelative probability that each is the case. If unable to accurately estimate \nthe utilities, they spin around in a circle three times chanting \"Do the \ndone thing\" and then ask the Martians what to do.\n\nHobbesian Plutonians - don't care whether the robots are persons or not. \nThey treat the robots as persons because cooperation is to their advantage.\n\nVenusian virtue-ethicists - treat the robots as persons even if they know \nthey're not conscious, because it reinforces their virtue of humane \ncompassion.\n\nMercurial relativists - Well, do the robots consider *themselves* persons? \nYou have to respect the robots' position on the matter. Unless, of course, \nthe current culture on Mercury has developed anti-robot values, in which \ncase who could blame them for destroying or enslaving the robots? Or, \nalternately, the more class-conscious Marxists among them elevate them as \nthe ultimate expression of the proletariat and elect the robots to be their \nleaders. But they might change their minds later. Or do all of the above, \nbecause after all there are no wrong answers." } ] }, { "message_id": "<20050202144646.96438.qmail@web50302.mail.yahoo.com>", "from": { "name": "Nathan Spears", "address": "spearofsolomon at yahoo.com" }, "date": "2005-02-02T06:47:21+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "I'm not sure I get the \"have to nuke the asteriod\" bit. If the computers ran off\nsolar energy than it would be a legimate problem, I suppose, although not for the\nreasons they were aware of. But don't you think that an entity which has\nconsciousness needs to be \"self-aware\" enough to make real decisions concerning its\nown future? Or were you saying that the computers destroyed the asteriod out of a\nlegitimate sense of self-preservation?\n\nChris, I would have gone with Platonic Plutonians: do the computers aspire to the\nForm of Personhood?\n\n\n\t\t" } ] }, { "message_id": "<1f7617370502020939286e70a5@mail.gmail.com>", "from": { "name": "Dan'l Danehy-Oakes", "address": "danldo at gmail.com" }, "date": "2005-02-02T09:39:19+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<20050202144646.96438.qmail@web50302.mail.yahoo.com>", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "Just as a side note:\n\nDifferentiate between behaviorism[1] and behaviorism[2]. (Sorry, my \nemail doesn't do subscripts.)\n\nBehaviorism[1] refers to the epistemological position that, since we\ndo not have access to the content of the consciousness of a subject\nof study (even if the subject describes the content of her consciousness,\nwe have to allow for inaccuracy, lies, etc.), the only aspect of the\nsubject we can actually study is her behavior (including statements\nabout the content of her consciousness, etc.).\n\nBehviorism[2] refers to the ontological position that no such thing \nas \"consciousness\" exists.\n\nBehaviorism[1] seems to-me a legitimate and scientific position.\n\nBehaviorism[2] seems to-me to have been concocted as a parody\nof behaviorism[1].\n\nHowever, some extreme behaviorists[1] often act and speak as if\nthey actually believed in behaviorism[2].\n\nThe point of all this? Turing's test works in the scientific realm\nof behaviorism[1], not the essentially religous (because \nincapable of empirical test) realm of behaviorism[2].\n\nSearle's Chinese Room gedankenexperiment works in the\nspace between behaviorism[1] and behaviorism[2].\n\nWhat I mean by this:\n\nTuring proposes an open test and says, \"If we obtained\na certain type of results, the most resaonable conclusion\nwould be that the machine in question was conscious.\" \nHe leaves the question open of whether the conclusion \nwould actually be true, or whether such a result might \nactually occur.\n\nSearle proposes a much more limited test and says, \"No \nmatter what the results, we must conclude that the Room \ndoes not 'know Chinese,' because we know _a priori_ that \nit does not.\"\n\n--Dan'l" }, { "kind": "sig", "depth": 0, "text": "\"We're going to sit on Scorsese's head\"\n -- The Goodfeathers" } ] }, { "message_id": "", "from": { "name": "Maru Dubshinki", "address": "marudubshinki at gmail.com" }, "date": "2005-02-02T10:29:53+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<20050202144646.96438.qmail@web50302.mail.yahoo.com>", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "What I meant was something along these lines: ' Suppose further that\nthe robots ran into some difficult, advanced, hard-to-solve\nlife-threatening problem. Let's call it X. Then the robots solve X\njust like the humans would have, if they were still around (because\nthe robots are so very good at imitating what humans would do). Then\nsome advanced aliens come...' etc.\n\n~Maru\nMicrosoft delenda est." }, { "kind": "quote", "depth": 1, "text": "Nathan Spears wrote:\nI'm not sure I get the \"have to nuke the asteriod\" bit. If the computers ran off\nsolar energy than it would be a legimate problem, I suppose, although not for the\nreasons they were aware of. But don't you think that an entity which has\nconsciousness needs to be \"self-aware\" enough to make real decisions concerning its\nown future? Or were you saying that the computers destroyed the asteriod out of a\nlegitimate sense of self-preservation?\n\nChris, I would have gone with Platonic Plutonians: do the computers aspire to the\nForm of Personhood?" } ] }, { "message_id": "", "from": { "name": "Maru Dubshinki", "address": "marudubshinki at gmail.com" }, "date": "2005-02-02T10:35:54+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "Ahh, now that depends on your philosphical position. If you go with\nPenrose and suchlike, then all you have to do is see if they pass the\nTuring Test and also run on the right substrate. But if you take the\nStrong AI position, and argue that intelligence and counsciousness are\nindependent of the physical instantiation, that what is relevant is\nthe pattern, then they have to fall back on observations, to what are\nessentially variations on the Turing Test.\nI'll leave out how every thought about reality is but a probability,\nsince it isn't really relevant.\nBut you left out the Bayesian Beta Regulisians: Examine carefully\ntheir actions, material culture, and the results of various tests you\ncarry out. Weight them all into the formula, obtain your best\nprobability, then evaluate expected utility and go from there.\n\n~Maru\nMicrosoft delenda est." }, { "kind": "quote", "depth": 1, "text": "Chris wrote:\nMaru said:" }, { "kind": "quote", "depth": 2, "text": "Suppose further that they made like a human, solving all\ntechnical problems, and an advanced alien civilization came by\nimpressed by what they had done. Would the aliens accord the robots\n'personhood' and would they be right or wrong?" }, { "kind": "quote", "depth": 1, "text": "You've stipulated they don't have \"real consciousness\", but the important\nunknown is: do the aliens have the ability to tell the difference, or if\nthey are uncertain, do they have an estimate of probability for the robots\nhaving \"real consciousness\"?" } ] }, { "message_id": "<20050202183844.60183.qmail@web50303.mail.yahoo.com>", "from": { "name": "Nathan Spears", "address": "spearofsolomon at yahoo.com" }, "date": "2005-02-02T10:38:48+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "But are the similarities superficial? That is, are they solving problems for\nthemselves using human methods, or for the same reasons humans would have? Do they\nbelieve that they are organic creatures dependent on oxygen and sunlight for\nsurvival, or are they \"aware\" (even if that only means, \"they factor into their\nproblem-solving the fact that they are machines\") that they are machines mimicking\nhuman thought-processes?" }, { "kind": "quote", "depth": 1, "text": "--- Maru Dubshinki wrote:\nWhat I meant was something along these lines: ' Suppose further that\nthe robots ran into some difficult, advanced, hard-to-solve\nlife-threatening problem. Let's call it X. Then the robots solve X\njust like the humans would have, if they were still around (because\nthe robots are so very good at imitating what humans would do). Then\nsome advanced aliens come...' etc.\n\n~Maru\nMicrosoft delenda est.\n\nNathan Spears wrote:" }, { "kind": "quote", "depth": 2, "text": "I'm not sure I get the \"have to nuke the asteriod\" bit. If the computers ran\noff\nsolar energy than it would be a legimate problem, I suppose, although not for\nthe\nreasons they were aware of. But don't you think that an entity which has\nconsciousness needs to be \"self-aware\" enough to make real decisions concerning\nits\nown future? Or were you saying that the computers destroyed the asteriod out of\na\nlegitimate sense of self-preservation?\n\nChris, I would have gone with Platonic Plutonians: do the computers aspire to\nthe\nForm of Personhood?" }, { "kind": "quote", "depth": 1, "text": "_______________________________________________" } ] }, { "message_id": "<029201c5095b$6a0d5620$4c1e4b0a@vcorp.ad.vrsn.com>", "from": { "name": "James Wynn", "address": "thewynns at earthlink.net" }, "date": "2005-02-02T11:14:29+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<1f7617370502020939286e70a5@mail.gmail.com>", "reply_link": "header", "blocks": [ { "kind": "quote", "depth": 1, "text": "Behviorism[2] refers to the ontological position that no such thing\nas \"consciousness\" exists.\nBehaviorism[1] seems to-me a legitimate and scientific position.\nBehaviorism[2] seems to-me to have been concocted as a parody\nof behaviorism[1]." }, { "kind": "text", "depth": 0, "text": "I don't know the whole history of Behaviorism[2], but wasn't it most\npopularized in B.F. Skinner's \"Beyond Freedom & Dignity\"? He is hardly a\nfringe personality in the science of Behavorialism. Skinner argued that\nproblems of human overpopulation could be handled by the emerging\n'technologies of behavior\" which in order to be fully developed had to be\nsundered from the notion of human \"free will\", that an \"automaton\" resides\nwithin us (as opposed to the understanding that we are the sum of Nature and\nNurture). There are lots of critics of Skinner (count me with them), but\nnone that I know of do so with empirical data." }, { "kind": "quote", "depth": 1, "text": "Searle proposes a much more limited test and says, \"No\nmatter what the results, we must conclude that the Room\ndoes not 'know Chinese,' because we know _a priori_ that\nit does not.\"" }, { "kind": "text", "depth": 0, "text": "I agree with this. The Searle's conclusion (and definitions) are implicit in\nhis thought experiment. But the argument against him always comes down to:\n\"Well, how do you know that *humans* \"think\" in the superlative manner\nSearle's analogy seems to presume they do?\" But that sounds like a tilt\ntoward B.F. Skinner to me.\n\nI think Wolfe has posed a much more complex scenario. Rather than building\nhis simulated personalities from scratch, he has them imprinted directly\nfrom a human pattern and says \"Are they people?\" Lemur says \"no\". He calls\nthem 'ghosts'. Sort of moving, monuments to people long dead. (And then he\nfound out that HE was such a monument as well). Mr. Million seems to agree\nwith Lemur.\n\nRose say \"yes\", because she already WAS a person. She's not a fraud,\nmimicking personality.\n\nV.R.T. is trying to *become* Marsch but does not seem to have succeeded\nbecause Number Five and his jailers seem to see right through him....but if\nhe got away with it, would he BE Marsch? Does this question not strike to\nthe heart of the Turing \"imitation game\"?\n\n~ Crush" } ] }, { "message_id": "", "from": { "name": "Maru Dubshinki", "address": "marudubshinki at gmail.com" }, "date": "2005-02-02T11:23:40+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<029201c5095b$6a0d5620$4c1e4b0a@vcorp.ad.vrsn.com>", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "There's a relevant idea in computer science which goes 'a perfect emulation \n*is* the thing being emulated.'\n\n~Maru " }, { "kind": "quote", "depth": 1, "text": "On Wed, 2 Feb 2005 14:14:25 -0500, James Wynn wrote:\n\nV.R.T. is trying to *become* Marsch but does not seem to have succeeded\nbecause Number Five and his jailers seem to see right through him....but if\nhe got away with it, would he BE Marsch? Does this question not strike to\nthe heart of the Turing \"imitation game\"?\n\n~ Crush" } ] }, { "message_id": "<1f76173705020211294435d5d5@mail.gmail.com>", "from": { "name": "Dan'l Danehy-Oakes", "address": "danldo at gmail.com" }, "date": "2005-02-02T11:29:55+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<029201c5095b$6a0d5620$4c1e4b0a@vcorp.ad.vrsn.com>", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "First, I do believe that we are reaching the (or a) heart of 5HC in \na way I've never seen before. I wonder if anyone here can tie\n\"A Story\" into this analysis?\n\nSecond...\n\nYes, Skinner is _the_ paradigmatic behaviorist. Behaviorism[2] is\nvery much pointed at the extreme behaviorist[1] position he takes\nin books like _Beyond Freedom and Dignity_ and, to a lesser \nextent, _Walden Two_. \n\nThe key point is that Skinner, as near as I can make out, does not \nclaim that \"consciousness\" does not happen - though he _does_\nclaim, if I'm not mistaken (it's been a looong time since I read S)\nthat there is no such \"thing\" - no nounish extant - as \"consciousness.\"\nRather, he is claiming that \"consciousness\" is irrelevant to \npragmatic matters (such as controlling the \"population explosion\"), \nwhich _can_ (he says) be handled by \"technologies of behavior\" \nthat don't take \"consciousness\" into account. \n\nIn this, he seems to me to fall into the same overall category as\nHitler, Stalin, Mao, Pol Pot, etc. - an anti-humanist. I also suggest\nthat this is a possible theme of 5HC, and particularly of \"VRT\".\n\n--Dan'l" }, { "kind": "sig", "depth": 0, "text": "\"We're going to sit on Scorsese's head\"\n -- The Goodfeathers" } ] }, { "message_id": "", "from": { "name": "Iorwerth Thomas", "address": "iorweththomas at hotmail.com" }, "date": "2005-02-03T04:29:05+00:00", "subject": "(urth) Behaviorism (footnote to a side note)", "in_reply_to": "<1f7617370502020939286e70a5@mail.gmail.com>", "reply_link": "header", "blocks": [ { "kind": "quote", "depth": 1, "text": "From: Dan'l Danehy-Oakes \n\nBehaviorism[1] seems to-me a legitimate and scientific position.\n\nBehaviorism[2] seems to-me to have been concocted as a parody\nof behaviorism[1].\n\nHowever, some extreme behaviorists[1] often act and speak as if\nthey actually believed in behaviorism[2]." }, { "kind": "text", "depth": 0, "text": "Hmm, I think (going by conversations with a psychologist flatmate) that [1] \nis historically prior to [2]... That may be because one gets so used to a \nstance taken for the sake of scientific method that one starts to take it as \nreality (add in logical postivism, and [1] pretty much becomes equivalent to \n[2], since the only things we can talk about are scientifically provable \nthings, if you're a logical positivist). Which do you class Skinner as?\n\nI keep coming across articles in the New Scientist that seem to be \nattempting to argue from [1] to [2]. That depresses me.\n\nIorwerth" } ] }, { "message_id": "", "from": { "name": "Iorwerth Thomas", "address": "iorweththomas at hotmail.com" }, "date": "2005-02-03T04:39:06+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<029201c5095b$6a0d5620$4c1e4b0a@vcorp.ad.vrsn.com>", "reply_link": "header", "blocks": [ { "kind": "quote", "depth": 1, "text": "From: \"James Wynn\" \n\nI don't know the whole history of Behaviorism[2], but wasn't it most\npopularized in B.F. Skinner's \"Beyond Freedom & Dignity\"? He is hardly a\nfringe personality in the science of Behavorialism. Skinner argued that\nproblems of human overpopulation could be handled by the emerging\n'technologies of behavior\" which in order to be fully developed had to be\nsundered from the notion of human \"free will\", that an \"automaton\" resides\nwithin us (as opposed to the understanding that we are the sum of Nature \nand\nNurture). There are lots of critics of Skinner (count me with them), but\nnone that I know of do so with empirical data." }, { "kind": "quote", "depth": 1, "text": "From conversations with a psychologist friend, I get the impression that he \nvaccilated between the two (he was an epiphenomenonologist (sp?), so \nconsciousness - when he believed in it - didn't matter anyway). He \napparently subjected his daughter to operant conditioning, but that may be \nexageration - my friend doesn't like him overmuch." }, { "kind": "text", "depth": 0, "text": "Iorwerth" } ] }, { "message_id": "", "from": { "name": "Chris", "address": "rasputin_ at hotmail.com" }, "date": "2005-02-03T11:21:26+00:00", "subject": "(urth) 5HC : Chinese boxes or tea chests?", "in_reply_to": "<1f76173705020211294435d5d5@mail.gmail.com>", "reply_link": "header", "blocks": [ { "kind": "text", "depth": 0, "text": "At the risk of terrible boredom, this is a reasonable reference:\n\nhttp://plato.stanford.edu/entries/behaviorism/\n\nAnalytical behaviorism seems to go just as far back as the other kind. But \nwhat I would point out here is that the mere epistemic position - that \nbehavior provides the only data we have to go by - is not enough to \nconstitute any of these types of behaviorism, and I don't think that what is \nbeing referred to here as \"behaviorism[1]\" is really behaviorism.\n\nTo draw a parallel with physics: when studying the motion of a ball in \nspace, the only data we have to go by is its position in time. Two \nscientists could entirely agree on this much. But let's say that one of \nthese scientists concludes that the proper way to study the motion of the \nball is by mathematically analyzing its motion, whereas the other insists \nthat the object of inquiry should not be the mere motion of the ball itself \nbut rather the action of the forces which act on the ball, and the laws of \nmotion in general.\n\nThis is not meant to be a perfect analogy. But my point is that being a \nbehaviorist of *any* stripe requires more than just a particular view of \nbehavior as data. It requires a certain outlook as to the limits, method, \nand object of study. And to accept Turing's position I don't think you have \nto accept any variety of behaviorism.\n\nAs a side note I think that the term has tended to refer to the more radical \nversions as time has gone on at least in part because people took the \ngeneral principles of the most extreme version and imported it into other \narenas under the name \"behaviorism\". For example \"linguistic behaviorism\", \nsomething Searle was concerned with, is the notion that there is no deeper \n\"meaning\" than simply the set of external stimuli that would lead one to \naffirm/assert or deny a sentence. [I can only say that his arguments against \nthis are much sharper, in my opinion, than the whole Chinese Room thing.] In \nany event anyone familiar with these other arenas rather than psychology is \ngoing to get a skewed impression of what \"behaviorism\" means." } ] } ] }