{ "export": { "source": "https://urth.darkrealm.vip/thread/urth/3188", "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": 3188, "subject": "Mr Million v Oreb", "messages": 10, "voices": 7, "first": "2002-04-21T20:45:49+00:00", "last": "2002-04-24T17:35:29+00:00" }, "messages": [ { "message_id": "", "from": { "name": "Nigel Price", "address": "nigel.a.price at virgin.net" }, "date": "2002-04-21T20:45:49+00:00", "subject": "(urth) Mr Million v Oreb", "in_reply_to": null, "reply_link": null, "blocks": [ { "kind": "text", "depth": 0, "text": "Apologies to hartshorn (no capital H, no terminal e). My friends the\nHartshornes do spell their name with a capital \"H\" and a terminal \"e\", and\nthat's probably why I wrote it that way. Sorry. Should have been\nconcentrating..." }, { "kind": "quote", "depth": 2, "text": "A cubic millimeter of brain has 14 kilometers\nof biological wiring packed into it. There\nis no comparion between the brain of any living\ncreature and ANY contemporary computer." }, { "kind": "text", "depth": 0, "text": "A splendidly precise answer to my inane speculations.\n\nMaybe the difference between Mr Million's implementation of a stored human\npersonality and Oreb's, then, (apart from the obvious one that they appear\nin different books, written at different times, when the author may have had\nmore or less experience of computer and/or completely different artistic\nintentions!) arises from the fact that Oreb's bird brain is already stuffed\nfull to capacity with night chough subroutines (datatypes? function calls?),\nleaving relatively little room to spare for downloaded cyber-deities.\nPerhaps loading Scylla into Oreb is like installing two operating systems\nand a double boot option onto your PC. There was plenty of space when you\nwere running just one OS, but it gets a bit crowded when you've got two.\n\nOr then again, Oreb's characteristic speech pattern may be the result of\nsome sort of \"bottleneck\" in the capacity of Oreb's speech centres.\n\nOr maybe it just made him interesting and distinctive as a character." }, { "kind": "quote", "depth": 2, "text": "I won't comment on the Pharasees/Saducsees debate" }, { "kind": "text", "depth": 0, "text": "I mentioned it once, but I think that I got away with it...\n\nNigel (no lowercase \"W\", no terminal \"z\")" } ] }, { "message_id": "", "from": { "name": "Andy Robertson", "address": "andywrobertson at clara.co.uk" }, "date": "2002-04-21T21:40:01+00:00", "subject": "Re: (urth) Mr Million v Oreb", "in_reply_to": "", "reply_link": "subject", "blocks": [ { "kind": "text", "depth": 0, "text": "----- Original Message -----" }, { "kind": "quote", "depth": 1, "text": "From: \"Nigel Price\" \n\nintentions!) arises from the fact that Oreb's bird brain is already\nstuffed\nfull to capacity with night chough subroutines (datatypes? function\ncalls?),\nleaving relatively little room to spare for downloaded cyber-deities." }, { "kind": "text", "depth": 0, "text": "The trouble is, of course, that a human personality needs a human brain to\nhold it. A smaller brain will not do. Only a fraction of a human\nprsonality could be embedded in a chough .\n\nToday we guess that the fastest supercomputers carry out about the same\nnumber of switching events per second as a human brain does. But they are\nBIG and consume a LOT of power.\n\nOne can only assume that the cybernetics of Wolfe's future is far advanced\nover ours, because a handful of \"cards\" seem to be able to hold something\nlike a human personality" }, { "kind": "quote", "depth": 1, "text": "Perhaps loading Scylla into Oreb is like installing two operating systems\nand a double boot option onto your PC. There was plenty of space when you\nwere running just one OS, but it gets a bit crowded when you've got two." }, { "kind": "text", "depth": 0, "text": "Candidly, organic data processing is not digital and it is probably\nimpossible to \"download\" a human or animal mind into a computer, full stop.\nThe whole buisness of the Holy Hues being visual downloading of personality\nis very, very, dodgy. But it is a familiar SF trope and Wolfe uses it\nwell.\n\n\n hartshorn" } ] }, { "message_id": "", "from": { "name": "Jeff Wilson", "address": "jwilson at io.com" }, "date": "2002-04-22T04:59:44+00:00", "subject": "Re: (urth) Mr Million v Oreb", "in_reply_to": "", "reply_link": "subject", "blocks": [ { "kind": "quote", "depth": 1, "text": "From: \"Andy Robertson\" " }, { "kind": "quote", "depth": 1, "text": "----- Original Message -----\nFrom: \"Nigel Price\" " }, { "kind": "quote", "depth": 2, "text": "intentions!) arises from the fact that Oreb's bird brain is already\nstuffed\nfull to capacity with night chough subroutines (datatypes? function\ncalls?),\nleaving relatively little room to spare for downloaded cyber-deities." }, { "kind": "quote", "depth": 1, "text": "The trouble is, of course, that a human personality needs a human brain to\nhold it. A smaller brain will not do. Only a fraction of a human\nprsonality could be embedded in a chough ." }, { "kind": "text", "depth": 0, "text": "There are numerous documented cases of microcephalics with normal mental\nfunctionality; a lot of the human brain is redundant, something that is\nmentioned as a performance boosting factor in \"5HC\". This can be read\nabout in Carl Sagan's book _Broca's Brain_, somethign along with Ray\nKurzweil's book would have been available to Wolfe during the 70s." }, { "kind": "quote", "depth": 1, "text": "Today we guess that the fastest supercomputers carry out about the same\nnumber of switching events per second as a human brain does. But they are\nBIG and consume a LOT of power." }, { "kind": "text", "depth": 0, "text": "Measuring raw switching events per second is misleading. The 10^11\nneurons in the brain might change state 20-30 times a second each, for\n2.5 2.5 x 10^12 switching events a second. However, almost 10 years ago,\nPentium processors with 3.3 million transistors were running at 100\nmegahertz, for 3.3 x 10^14 switching events a second, and just consuming\na few watts.\n\nRay Kurzweil writes on this sort of comparison extensively; at current\nrate of development, computers will approach the complexity of the human\nbrain in terms of informational and interconnected capacity in about 30\nyears." }, { "kind": "sig", "depth": 0, "text": "Jeff Wilson\nHow Am I Posting? 1-800-555-6789\n\"If your SecOp can see you, so can the enemy.\" -Cpt Law\n\n\n-- " } ] }, { "message_id": "", "from": { "name": "Andy Robertson", "address": "andywrobertson at clara.co.uk" }, "date": "2002-04-22T07:49:07+00:00", "subject": "Re: (urth) Mr Million v Oreb", "in_reply_to": "", "reply_link": "subject", "blocks": [ { "kind": "text", "depth": 0, "text": "----- Original Message -----" }, { "kind": "quote", "depth": 1, "text": "From: \"Jeff Wilson\" " }, { "kind": "quote", "depth": 2, "text": "There are numerous documented cases of microcephalics with normal mental\nfunctionality; a lot of the human brain is redundant, something that is\nmentioned as a performance boosting factor in \"5HC\"." }, { "kind": "text", "depth": 0, "text": "I am begnning to have serious doubts about the \"numerous microcephalics\"\nbeing true, and I'd appreciate a link to a primary source (not a dodgy\npopulist like Sagan). I have chased these rumors up before and they seem\nto come to nothing, or to ignore the fact that the \"normal\" person in fact\nhas serious behavioral deficiencies, though they may score averaage on an IQ\ntest. But I bow to correction on this point.\n\nI agree that a human brain can work at half human size. After all, children\nmanage! But not at 1/10 or 1/40 human size." }, { "kind": "quote", "depth": 1, "text": ". However, almost 10 years ago,\nPentium processors with 3.3 million transistors were running at 100\nmegahertz, for 3.3 x 10^14 switching events a second, and just consuming\na few watts.\n\nRay Kurzweil writes on this sort of comparison extensively; at current\nrate of development, computers will approach the complexity of the human\nbrain in terms of informational and interconnected capacity in about 30\nyears." }, { "kind": "text", "depth": 0, "text": "Thanks for the correction. But the \"30 years to equality\" figure is one I\nhave heard before. Indeed I was hearing it 30 years ago, and I suspect we\nwill be hearing it 30 years from now.\n\n hartshorn" } ] }, { "message_id": "", "from": { "name": "David DiGiacomo", "address": "david at slack.com" }, "date": "2002-04-22T18:21:47+00:00", "subject": "(urth) Re: Mr Million v Oreb", "in_reply_to": "", "reply_link": "subject", "blocks": [ { "kind": "quote", "depth": 1, "text": "The trouble is, of course, that a human personality needs a human brain to\nhold it. A smaller brain will not do. Only a fraction of a human\nprsonality could be embedded in a chough ." }, { "kind": "text", "depth": 0, "text": "This is still unknown. Perhaps the personality is stored only in the\nhuman-specific parts of the brain, which are a fraction of the total\nsize. If the information were organized efficiently, and non-essentials\nremoved, it might be possible to find room for it in a smaller brain." }, { "kind": "quote", "depth": 1, "text": "Today we guess that the fastest supercomputers carry out about the same\nnumber of switching events per second as a human brain does. But they are\nBIG and consume a LOT of power." }, { "kind": "text", "depth": 0, "text": "I don't think today's computers have anywhere near the storage capacity of\nan animal brain." }, { "kind": "quote", "depth": 1, "text": "Candidly, organic data processing is not digital and it is probably\nimpossible to \"download\" a human or animal mind into a computer, full stop." }, { "kind": "text", "depth": 0, "text": "The conclusion may be valid, but the premise doesn't support it. Audio\nsignals are not digital, but we have no problem downloading music into our\ncomputers.\n\n(There is also no technical obstacle to incorporating analog circuits in\ncomputers.)" }, { "kind": "quote", "depth": 1, "text": "The whole buisness of the Holy Hues being visual downloading of personality\nis very, very, dodgy. But it is a familiar SF trope and Wolfe uses it\nwell." }, { "kind": "text", "depth": 0, "text": "The downloading is questionable enough, but what about the uploading?!?\nJust how do the bios of the Whorl transmit information through their eyes?\nAll I can think of is that the upload version of the Holy Hues triggers\nevoked potentials which can be detected in some way, but that's not very\nsatisfying." } ] }, { "message_id": "", "from": { "name": "Grant Peacock", "address": "grantpeacock at yahoo.com" }, "date": "2002-04-22T18:15:24+00:00", "subject": "Re: (urth) Mr Million v Oreb", "in_reply_to": "", "reply_link": "subject", "blocks": [ { "kind": "quote", "depth": 1, "text": "--- Jeff Wilson wrote:\nMeasuring raw switching events per second is misleading. The 10^11\nneurons in the brain might change state 20-30 times a second each, for\n2.5 2.5 x 10^12 switching events a second. However, almost 10 years ago,\nPentium processors with 3.3 million transistors were running at 100\nmegahertz, for 3.3 x 10^14 switching events a second, and just consuming\na few watts." }, { "kind": "text", "depth": 0, "text": "I'm not sure this is number makes any sense. A 100 meg processor executes a\nmaximum of 10^8 operations per second, no matter how much memory it has\navailable.\n\nMy source of information is _I am Right, You are Wrong_ by Edward de Bono. I\nrecommend this book. Unfortunately my copy is on loan so I might miss a lot of\ndetails here. The computing knowledge is also shaky. But here goes. A neuron\nhas 2 states, firing and not firing, but don't let this fool you into thinking\nof the brain as binary. A neuron also has dendrites pointing to dozens of\nother neurons, and it also has a variable stimulation threshold. That is, it\nneeds a certain amount of \"juice\" from neurons that point to it before it\nfires. This limit changes depending on the amount of time since it fired\npreviously, and on levels of whatever chemicals to which it is sensitive. \nFinally, the amount of influence one neuron has over another can be increased\nover time. This happens when two connected neurons fire at the same time.\n\nSo, without too much effort to be efficient about the coding, let's estimate\nhow much computing power one would need to try to simulate a brain, using a\nstandard style of computer. For each neuron, we need a list of other neurons\nto which this one has a dendrite, and the sensitivity of each. Here we're\ntalking probably 3 bytes * 20 or 30 neurons. This is the hard part,\nspace-wise. Keeping it on a disk would be way too slow -- I think it has to be\nin memory. So we need about 100 bytes per neuron, or 10^7 megs of RAM. This\nis 10,000 times the memory of today's PC but it is conceivable.\n\nNow, the real bottleneck is that we only have one processor. We can only look\nat one neuron at a time. The processor will have to make \"passes\" through the\nentire list of neurons. For each pass, on each neuron which is firing, it will\ngo down the list of dendrites and add the right amount of juice to all the\nneighbors. Let's suppose that a maximum of 10% of a brain's 10^11 neurons can\nbe firing at any given time. Still, that's 10^10 * 20 additions that must be\nperformed in the slow part of the pass. (I'm assuming an average of 20\ndendrites/neuron, this may be too low.) Adding probably takes more than 8 or\n10 clock cycles but let's say 10. So each pass takes on the order of 10^12\nclock cycles.\n\nNow, when we consider that a neuron might go on and off 20 times per second,\nI'd say the passes I just described need to be performed at least 1000 times\nper second to achieve realistic brain activity. So our processor needs to run\nat around 10^16 hertz, that is, do the work of 100 million pentiums.\n\nThe reason we're still behind by such a large factor is that each neuron is not\njust a memory cell, it also performs some of the functions assumed by the\nprocessor in a computer. \n\n-Peacock" } ] }, { "message_id": "", "from": { "name": "matthew.malthouse", "address": "matthew.malthouse at guardian.co.uk" }, "date": "2002-04-23T11:57:22+00:00", "subject": "Re: (urth) Re: Mr Million v Oreb", "in_reply_to": "", "reply_link": "subject", "blocks": [ { "kind": "quote", "depth": 1, "text": "On 22/04/2002 19:21:47 david wrote:\nThe downloading is questionable enough, but what about the uploading?!?\nJust how do the bios of the Whorl transmit information through their\neyes?\nAll I can think of is that the upload version of the Holy Hues triggers\nevoked potentials which can be detected in some way, but that's not very\nsatisfying." }, { "kind": "text", "depth": 0, "text": "This is entirely speculative as I don't think there's any supporting text\nbut it would make sense if as you suggest the Holy Hues were a trigger\nrather than a mechanism. Perhaps akin to the susceptibility some\nepileptics have to stroboscopic effects.\n\nAfter all Pas had the opportunity to modify the 'cargo' when loading the\nWhorl to insert any 'equipment' required and this is a universe where a\nperson's life experience can be assimilated by the ingestion of a finger's\nworth of (cooked!) flesh and a preperation of a xenoc animal's gland.\n\nActually I don't think that the 'how' has any pertinant significance to\nthe understanding of the text: what has is that the process works.\n\nMatthew" } ] }, { "message_id": "", "from": { "name": "Jeff Wilson", "address": "jwilson at io.com" }, "date": "2002-04-24T06:03:39+00:00", "subject": "Re: (urth) Mr Million v Oreb", "in_reply_to": "", "reply_link": "subject", "blocks": [ { "kind": "quote", "depth": 1, "text": " From: \n Grant Peacock " }, { "kind": "quote", "depth": 1, "text": "--- Jeff Wilson wrote:" }, { "kind": "quote", "depth": 2, "text": "Measuring raw switching events per second is misleading. The 10^11\nneurons in the brain might change state 20-30 times a second each, for\n2.5 2.5 x 10^12 switching events a second. However, almost 10 years ago,\nPentium processors with 3.3 million transistors were running at 100\nmegahertz, for 3.3 x 10^14 switching events a second, and just consuming\na few watts." }, { "kind": "quote", "depth": 1, "text": "I'm not sure this is number makes any sense. A 100 meg processor executes a\nmaximum of 10^8 operations per second, no matter how much memory it has\navailable." }, { "kind": "text", "depth": 0, "text": "Not true; many processors have multiple instruction piplines inside that\nlet them do multiple operations each clock cycle. A Pentium I averaged\none instruction per Hertz, but that averages includes instances where it\ncan do two operations at once, each taking a couple of cycles to\ncomplete. Later processors do even better, as they finish operations\nsooner and are more likely to be able to handle dual instructions in\ntandem. Regardless of the external speed, the processor is still made up\nof millions of interconnected transistors, each of which can change\nstate hundreds of millions, and lately, billions of times per second." }, { "kind": "quote", "depth": 1, "text": "My source of information is _I am Right, You are Wrong_ by Edward de Bono. I\nrecommend this book. Unfortunately my copy is on loan so I might miss a lot of\ndetails here. The computing knowledge is also shaky. But here goes. A neuron\nhas 2 states, firing and not firing, but don't let this fool you into thinking\nof the brain as binary. A neuron also has dendrites pointing to dozens of\nother neurons, and it also has a variable stimulation threshold. That is, it\nneeds a certain amount of \"juice\" from neurons that point to it before it\nfires. This limit changes depending on the amount of time since it fired\npreviously, and on levels of whatever chemicals to which it is sensitive. \nFinally, the amount of influence one neuron has over another can be increased\nover time. This happens when two connected neurons fire at the same time." }, { "kind": "text", "depth": 0, "text": "This sounds similar in scope to the interconnected transistor\narchitecture inside a microprocessor; their exact configuration and\nlevels of influence on one another is also adjustable nowadays, thanks\nto the embarrassment over bugs publicly discovered in previous\ngenerations of chips." }, { "kind": "quote", "depth": 1, "text": "So, without too much effort to be efficient about the coding, let's estimate\nhow much computing power one would need to try to simulate a brain, using a\nstandard style of computer. For each neuron, we need a list of other neurons\nto which this one has a dendrite, and the sensitivity of each. Here we're\ntalking probably 3 bytes * 20 or 30 neurons. This is the hard part,\nspace-wise. Keeping it on a disk would be way too slow -- I think it has to be\nin memory. So we need about 100 bytes per neuron, or 10^7 megs of RAM. This\nis 10,000 times the memory of today's PC but it is conceivable.\n\nNow, the real bottleneck is that we only have one processor. We can only look\nat one neuron at a time. " }, { "kind": "text", "depth": 0, "text": "\tThis is wrong; there's no need to limit the computers to one processor\napiece. Multiprocessor computing is old hat, in fact the ancient ENIAC\nhad separate processing in each of its 64 interconnectable modules.\nHardwire multiprocessor designs run into the tens of thousands, letting\nus simulate physical processes at subatomic levels; try googling for\nASCI White. You might also try reading about Beowulf clusters; freely\ndownloadable software exists to let you join arbitrary numbers of\nnetworked computers into a virtual multiprocessing machine. A Beowulf\ncluster of 100 rackmounted PCs with 10,000 megs of RAM each would take\nup about as much room as Mr. Million\n\tEven if you are supposing the brain is to be simulated on a single\nprocessor, general purpose machine instead of a special purpose\nmechanism like Mr. Million, procs today are being designed with SIMD\ninstructions; Single Instruction, Multiple Data. A SIMD instruction to\nadd could be defined; it would quickly sum the adjacent neuron's juice x\nsensitivity values and return true if they exceed the central neuron's\nthreshold. This would not be much different from the digital signal\nprocessing used to de-blur photographs and reduce audio noise." }, { "kind": "quote", "depth": 1, "text": "The processor will have to make \"passes\" through the\nentire list of neurons. For each pass, on each neuron which is firing, it will\ngo down the list of dendrites and add the right amount of juice to all the\nneighbors. Let's suppose that a maximum of 10% of a brain's 10^11 neurons can\nbe firing at any given time. Still, that's 10^10 * 20 additions that must be\nperformed in the slow part of the pass. (I'm assuming an average of 20\ndendrites/neuron, this may be too low.) Adding probably takes more than 8 or\n10 clock cycles but let's say 10. So each pass takes on the order of 10^12\nclock cycles.\n\nNow, when we consider that a neuron might go on and off 20 times per second,\nI'd say the passes I just described need to be performed at least 1000 times\nper second to achieve realistic brain activity. So our processor needs to run\nat around 10^16 hertz, that is, do the work of 100 million pentiums.\n\nThe reason we're still behind by such a large factor is that each neuron is not\njust a memory cell, it also performs some of the functions assumed by the\nprocessor in a computer. " }, { "kind": "text", "depth": 0, "text": "Good reasons for multiprocessing and other options I mentioned above. A\nneuron is much more complex than a transistor, but much less so than a\nmicroprocessor. It would not be inconceivably difficult to design a\ncomputing device with thousands or millions of subordinate computing\nelements to simulate the effects of individual neurons, and to\ncoordinate their action in ways suitable to model the way real neurons\nare connected. Hans Moravec and his ilk are keen on doing it." }, { "kind": "sig", "depth": 0, "text": "Jeff Wilson\nHow Am I Posting? 1-800-555-6789\n\"If your SecOp can see you, so can the enemy.\" -Cpt Law\n\n\n-- " } ] }, { "message_id": "", "from": { "name": "Michael Straight", "address": "straight at email.unc.edu" }, "date": "2002-04-24T17:35:29+00:00", "subject": "Re: (urth) Re: Mr Million v Oreb", "in_reply_to": "", "reply_link": "subject", "blocks": [ { "kind": "quote", "depth": 1, "text": "On Mon, 22 Apr 2002, David DiGiacomo wrote:\nThe downloading is questionable enough, but what about the uploading?!?\nJust how do the bios of the Whorl transmit information through their eyes?\nAll I can think of is that the upload version of the Holy Hues triggers\nevoked potentials which can be detected in some way, but that's not very\nsatisfying." }, { "kind": "text", "depth": 0, "text": "Maybe this is what you mean by \"evoked potentials,\" but I always assumed\nthat the eye thing was merely a trigger, that the Holy Hues put the brain\ninto some sort of receptive state so that some sort of electromagnetic\ndoohicky could read/write to the brain directly, not actually transferring\ndata through the optic nerve.\n\n-Rostrum" } ] }, { "message_id": "", "from": { "name": "Andy Robertson", "address": "andywrobertson at clara.co.uk" }, "date": "2002-04-24T16:34:52+00:00", "subject": "Re: (urth) Re: Mr Million v Oreb", "in_reply_to": "", "reply_link": "subject", "blocks": [ { "kind": "text", "depth": 0, "text": "----- Original Message -----" }, { "kind": "quote", "depth": 1, "text": "From: \"Michael Straight\" \nMaybe this is what you mean by \"evoked potentials,\" but I always assumed\nthat the eye thing was merely a trigger, that the Holy Hues put the brain\ninto some sort of receptive" }, { "kind": "text", "depth": 0, "text": "TBOTLS is among other things a retrospective of SF.\n\nHere, we see Cyberpunk being treated.\n\nThe main idea of Cyberpunk was to take \"cyberspace\" and fuse it with the\nSpirit world. Cyberspace becomes the place where the Gods live, where you\ndream, where you may go after death. It is this (utterly uncientific) idea\nthat gave Gibson's first novels their tremndous punch.\n\nWolfe is nodding to this, among other things.\n\n\n harthsorn" } ] } ] }