API tools faq
paste
Advertisement
JaysonSunshine

Convo with commie

JaysonSunshine
May 7th, 2018
81
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 12.71 KB | None | 0 0
raw download clone embed print report
  1. 3:11:14 PM+commieIt's going to take a bit before I can argue the position, but I've been reading quite a bit about Assemblage Theory lately and it is blowing my mind constantly. It's a whole ontology that rises up to challenge everything I thought I knew and it connected to observations I've made in interesting ways.
  2. 3:11:49 PM→ |WP| has joined
  3. 3:11:56 PMⓘ Mabus set mode +v |WP|
  4. 3:12:44 PM+LowInformationVotercommie: Can haz preview?
  5. 3:13:18 PM+commieI'm not asking anyone to change their view, I'm just finding it interesting and challenging to my own views.
  6. 3:13:20 PM+LowInformationVoterThe last very influential construct I came across was probably some of the statistical foundations of machine learning.
  7. 3:13:44 PM+LowInformationVoterI've previously read a bit about Deleuze -- he has some verbose constructs.
  8. 3:13:49 PM+commieYes
  9. 3:13:56 PM+commieI'm presently reading Manuel DeLanda
  10. 3:14:10 PM+commieWho fleshes out Assemblage Theory beyond the mentions Guattari and Deleuze put forward
  11. 3:14:20 PM+LowInformationVoterFor example, in regards to statistical foundations of machine learning, you can always have a binary classifier for any system.
  12. 3:14:29 PM+LowInformationVoterSo, yeah, you can say all humans are of type X or non-X.
  13. 3:14:43 PM+LowInformationVoterOr you have five categories. Or one category per object.
  14. 3:15:56 PM+LowInformationVoterAlso, I now appreciate essentially all language constructs are partially false.
  15. 3:16:07 PM+commieSo far what I'm getting is that wholes are not defined by their parts i.e. they are not aggregations and the properties of the parts can not explain the whole. In Assemblage Theory, you focus on the actual synthesis, or interaction, of component parts.
  16. 3:16:17 PM+LowInformationVoterLanguage ends up being a kind of classifier in relatively low-dimensional space, as natural language is not precise.
  17. 3:16:45 PM+LowInformationVoterIf you're discussing a very high dimensional object like, 'What is moral?', you'll have many, many, many, many ways you can project that to a lower dimensional subspace.
  18. 3:17:25 PM@ProfFarnsworthhttps://scontent-ort2-1.xx.fbcdn.net/v/t1.0-9/31113767_791678974356004_4301163850804455448_n.png?_nc...
  19. 3:17:28 PM+LowInformationVoterThis is explains, for example, why some really smart people think consequentialist thinking is correct, why some other really smart people think categoricalism is correct.
  20. 3:17:55 PM+LowInformationVoterFinally, it's helpful to think of the precision of the analyzing process in regards to the precision with which you wish to discuss the object under analysis.
  21. 3:18:17 PM+commieLanguage is an interesting subject.
  22. 3:18:29 PM+LowInformationVoterIf the system -- for example a court proceeding -- can only handle five dimensions of analysis, with some maximum resolution, you will immediately understand this cannot discern some objects from one another.
  23. 3:18:36 PM+LowInformationVoterFor example, Are these two acts both examples of murder?
  24. 3:18:46 PM+commieFor awhile I've had a *sense* of what I called a "lattice." That is, a kind of emergent platform upon which higher actualizations can be achieved, but I didn't have any framework for it.
  25. 3:18:57 PM+commiei.e. genetics or language
  26. 3:19:28 PM+LowInformationVoterIf you feel you're lacking precision for this intuition, I recommend studying more mathematics.
  27. 3:19:37 PM+LowInformationVoterMathematics is the most precise language system in the noosphere.
  28. 3:20:21 PM+LowInformationVotercommie: I am thinking about your preview of Assembling Thinking.
  29. 3:20:31 PM+LowInformationVoterAre you familiar with Buddhist metaphysics?
  30. 3:20:36 PM+commieI am not.
  31. 3:20:50 PM+LowInformationVoterAt the core of Buddhist metaphysics is that everything is composed of non-self parts.
  32. 3:21:04 PM+LowInformationVoterAnd, also, that all boundaries/categories are in the perception and not reflect in the world.
  33. 3:21:11 PM+LowInformationVoterSo, people are composed entirely of non-people components.
  34. 3:21:19 PM+LowInformationVoterAnd you get interesting conclusions like "There are no people".
  35. 3:21:26 PM+commieheh
  36. 3:21:29 PM+LowInformationVoterWhere you can replace "People" with every object you think exists.
  37. 3:21:34 PM+commieI think the boundaries themselves are in the real world.
  38. 3:21:42 PM+LowInformationVoterCan you give an example?
  39. 3:22:20 PM+LowInformationVoterThe reason I bring this up is because Buddhist would agree with you that you cannot understand an object merely by what it is or the whoel.
  40. 3:22:42 PM+LowInformationVoterHmm. I have a misperception of your words, it seems.
  41. 3:23:02 PM+commiePhase changes in water. Hydrogen, oxygen, non-water parts make up the substance water, which has energy boundaries that produce state changes (at a certain threshold, they suddenly change morphogeneticly
  42. 3:23:05 PM+LowInformationVoterOkay, scrap my last post starting with "The reason".
  43. 3:23:27 PM+LowInformationVotercommie: I like that example.
  44. 3:23:42 PM+LowInformationVoterMy worldview is some combination of Buddhism, physics/science, and others.
  45. 3:24:22 PM+LowInformationVoterI think physics has been able to peer sufficiently deep and precisely so as to refute some notions in Buddhism.
  46. 3:24:30 PM+commieI was reading a book (I need to reread it though) called Dividuum which argued for a concept of "dividual."
  47. 3:24:47 PM+LowInformationVoterI think that there is some dynamical system of the matter/energy that behave according to some complex rules/patterns/functions.
  48. 3:25:12 PM+LowInformationVoterAnd, at some points in that phase space, there are abrupt changes or steep gradients.
  49. 3:25:29 PM+commieI think in Deleuzian ontology the space in which matter arises has structure and determines how matter can operate morphogenetically.
  50. 3:25:35 PM* LowInformationVoter pauses
  51. 3:25:38 PM+commieAnd these manifest in assemblages.
  52. 3:26:05 PM← Guest23816 has quit (Ping timeout)
  53. 3:26:25 PM+commieLowInformationVoter: Ah, attractors?
  54. 3:27:09 PM→ Banquo has joined
  55. 3:27:13 PM+LowInformationVoterWell, phase space changes in water are not attractors, to my understanding, but merely areas with a steep gradient (perhaps in the potential energy).
  56. 3:27:14 PMⓘ Mabus set mode +v Banquo
  57. 3:27:31 PM+commieLowInformationVoter: Oh, sorry, I was thinking you meant more generally, I would agree with that.
  58. 3:29:06 PM+commieSo given populations of assemblages and all the configurations and degrees of freedom they may operate in... I get the fleeting sense of simulation - of a system generating as many permutations as possible.
  59. 3:30:12 PM+LowInformationVoterI'm reading the Assemblage Theory article.
  60. 3:30:15 PMⓘ Debates set mode -b *!*@anonysaurus.com
  61. 3:30:22 PM+LowInformationVoterI guess the construct he is leveraging is a dynamic graph.
  62. 3:30:28 PM+commieMhm
  63. 3:30:40 PM+commieFor non-linear causality
  64. 3:31:05 PM+commieAmong other things, I imagine, but the example I remember was for that purpose.
  65. 3:31:09 PM+LowInformationVoterSo, what we might call 'Islam' is actually some semi-stable metagraph that has some current instantiation which is composed of the people, artifacts, who 'belong' to Islam.
  66. 3:31:21 PM+LowInformationVoterWhat do you mean @ non-linear causality
  67. 3:31:56 PM+LowInformationVoterThat actual object: the graph of people and artifacts belonging to Islam is constantly changing, and sometimes may undergo substantial reformulations.
  68. 3:32:04 PM+LowInformationVoterThe Islam that we discussed yesterday isn't precisely the Islam we are discussing today.
  69. 3:32:29 PM+LowInformationVoterAs some people have died, some beliefs in some of the people belonging to 'Islam' have been modified, artifacts have changed, etc.
  70. 3:32:51 PM+commieThe example I recall was linear causality = weight on a spring, and the relation between the spring length and the weight, so that yuo may end up with "double the weight and double the length of the spring"). The weird example he used for non-linear causality was pulling on your lip. A small force = large change, but then you hit a kind of "wall" and no matter the force exerted, you can't stretch your lip
  71. 3:32:57 PM+commiefurther.
  72. 3:33:07 PM+commieDeLanda, in his talk, says most systems are non-linear anyway
  73. 3:33:37 PM+LowInformationVoterThis approach to sociology seems similar to symbolic-interactionism.
  74. 3:33:55 PM+LowInformationVoterLet me finish this thread real quick and then I'll engage your most recent post.
  75. 3:34:01 PM+commiek
  76. 3:34:15 PM+LowInformationVoterIn machine learning, systems like neural networks are effective because they simultaneously analyze objects at multiple levels of perspective.
  77. 3:34:46 PM+LowInformationVoterSo, in a picture of a face, parts of my model are looking for straight lines, other parts are looking for things that use straight lines in simple shapes, things that use that, things that use those, etc.
  78. 3:35:05 PM+LowInformationVoterI think this is what Deluze is essentially advocating for as our paradigm of sociology.
  79. 3:35:37 PM+LowInformationVoterFor some types of analysis, we can think of Islam as this one big thing in Africa/ME/Asia, e.g. how it's interacting with Christianity and a border, e.g. in Turkey.
  80. 3:35:55 PM+LowInformationVoterBut if you increase the resolution, it's not a single object but multiple objects working together.
  81. 3:36:09 PM+LowInformationVoterAnd for a very fully analysis you'd have to go all the way down to the person level, and perhaps beyond.
  82. 3:36:19 PM+LowInformationVoterHumans cannot perform this level of analysis with our brains evolution gave to us.
  83. 3:36:35 PM+LowInformationVoterEven SOTA machine learning algorithms can sometimes struggle with this on complex data sets.
  84. 3:36:42 PM+LowInformationVoterI'll pause there.
  85. 3:37:00 PM+LowInformationVoterAnd, of course, that 'object', complex at multiple levels, is changing, too.
  86. 3:37:09 PM+commiehttps://www.youtube.com/watch?v=0wW2l-nBIDg I found this video helpful
  87. 3:37:31 PM+LowInformationVoterDo you think I am missing anything essential in Deluze's assemblage theory model?
  88. 3:38:04 PM+LowInformationVoterI agree most systems are non-linear.
  89. 3:38:21 PM+LowInformationVoterHumans seem to intuitively apply linear thinking because it's computationally cheap and reasonable effective.
  90. 3:38:43 PM+commieI don't know enough to comment on that, I'm still reading it. I do think Deleuze was going far beyong sociology though, since these machines are also the basis of biological and non-biological processes as well.
  91. 3:39:09 PM+LowInformationVoterYes, I think the structure I described is extremely generic.
  92. 3:39:23 PM+LowInformationVoterBasically, anything that can be modeled as a network can be analyzed with this approach.
  93. 3:39:45 PM+LowInformationVoterI think organizations doing SOTA data science on enormous, detailed networks should produce outcomes that startle us.
  94. 3:39:53 PM+LowInformationVoterThey advertised in 2014 that they could change elections.
  95. 3:39:59 PM+LowInformationVoterIn 2016, they changed the American election.
  96. 3:40:15 PM+LowInformationVoterThey should startle us because they can do something our brains simple cannot do.
  97. 3:40:36 PM+LowInformationVoterPerhaps no human who has ever lived has had the ability to do a reasonable multi-level, dynamic analysis of a complex network like 'America' or 'Islam'.
  98. 3:41:02 PM+LowInformationVoterThey = Facebook
  99. 3:41:08 PM+commieI believe this has spun off some modifications to evolutionary theory which I'm interested in looking into
  100. 3:41:26 PM+LowInformationVoter(They deleted those documents after it was revealed their platform was used by Russians, among others, to change the American 2016 presidential election)
  101. 3:41:26 PM+commiehttps://books.google.com/books/about/Developmental_Plasticity_and_Evolution.html?id=iBkQyA2PkxEC
  102. 3:41:48 PM+LowInformationVoterInteresting.
  103. 3:41:58 PM+LowInformationVoterI find your curiosity refreshing.
  104. 3:42:10 PM+LowInformationVoterTruly it can be said you are not a low information voter.
  105. 3:42:28 PM+LowInformationVotercommie: Do you think all people should have equal votes in democracy?
  106. 3:43:00 PM+LowInformationVoterI think it's pretty obvious this is not optimal, but it would be hard to devise a better system that was stable in a population that includes subpopulations that hate modernity, mathematics, and education.
  107. 3:43:21 PM+commieI am not sure I am settled on the question of democracy itself (not to say I support authoritarian systems). i.e. I think it is based on a faulty premise of individualism and representationalism which I am increasingly skeptical of.
  108. 3:44:04 PM+LowInformationVoterThat seems reasonable.
  109. 3:44:15 PM+LowInformationVoterIndividualism does seem a problematic construct.
  110. 3:44:40 PM← HansMcfarland has quit (Read error: Operation timed out)
  111. 3:45:46 PM+commiegotta run, heading to a meeting tonight
  112. 3:45:53 PM+LowInformationVoterTake care, comrade.
  113. 3:46:02 PM+LowInformationVoterUnder his eye.
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!