shalizi   165

« earlier    

Mixture Models from Cosma Shalizi
looks like a version that basically preceded the book chapter. Clear as always
Shalizi  gaussian-mixture-model 
april 2018 by gideonite
A Marxian Econophysics
looks fascinating, but no energy at the moment.
economics  physics  Shalizi 
november 2017 by gideonite
John E. Roemer, A Future for Socialism
In summary: This book makes a good case for a species of market socialism, and (here I risk prophecy) will have no influence at all.

For more on what the seemingly contradictory term "market socialism" means, read this review. Seems strange to reject the labor theory of value but perhaps this is necessary to do something which is eminently useful, and not merely a critique.
socialism  economics  Marx  Shalizi 
november 2017 by gideonite
In Soviet Union, Optimization Problem Solves You — Crooked Timber
Coming back to some of these ideas after getting some education in optimization, this article rings true in many ways and expresses some previous ideas in a language that I have come to enjoy. Just as the ACN (attention conservation notice) suggests, there are many references to optimization and mathematics, but little or no actual maths. Furthermore, while enjoyable reading, the article is not particularly optimized for reader comprehension. Many of the ideas could be expressed in a more straight-forward manner without loosing fidelity.

Solving a contrained optimization problem, but which one? The function itself is undefined. Is it linear? Is it convex? The objective is undefined. What exactly is the quantity that we are trying to optimize?

But suppose for a moment, suppose that the problem is linear. We would essentially need prices for all goods (linear coefficients). Furthermore, these prices would need to be indexed by space and time since exchanging an apple in Australia for an orange in Florida is not the same as if the apple was in New York. There are some bounds for linear programming problems which are polynomial in n, the number of products. But according to Alec Nove, in /The Economics of Feasible Socialism/, there were 12 million identifiable products in the USSR in 1983. Introduce spacial indexing and we're screwed because this will grow the number of products to the billions, which is no longer tractable in polynomial time.

But to return to the real problem --- finding the optimal plan, where plan means pricing.

Furthermore, there's the question of the labor theory of value. Coming back to Marx after all this time, it is refreshing to think of this idea once again. In it contains almost everything. The question of the uniqueness of labor as a commodity in the commodities vector /was/ discussed at length by Marx (IIRC) and so it is unfortunate that Shalizi overlooks this.

Marx's argument addresses this article directly and despite his best intentions, Shalizi misses this point. The point is to go beyond the labor theory of value. The entire exposition here is just another exposition of the fundamental contradictions between freedom and the labor theory of value.

There is a reason why labor is "traditionally" the universal product. It is central to the entire theory. It is discussed at length in the first chapter of Kapital and is not an arbitrary choice.

This problem with Shalizi's critique also comes out in the discussion of the ultimate objective for the optimization problem since it is tautological to maximize wealth defined in terms of a labor theory of value.

The idea of convex constraints in this context is also interesting since it has something to do with exchangability --- convex constraints say that as long as you remain below a certain quantity, any exchange is valid.

Finally, it refreshing to view things practically in terms of democracy. Since ultimately, plans are policies, and policies are watered down for the /demos/ to vote on them.

Effective-demand --- another interesting concept. Since the "demand signal" is confounded by the distribution of wealth. It is literally weighted by that distribution.

Fundamentally disagree that all massed human interactions (institutions) are alien in some form. Compare to other societies. Use sociology. Durkheim.

Indeed, what this shows is that planning is not an alternative to the market as a way of operating the optimization problem. But, this doesn't touch upon the system of capitalism.
optimization  economics  Marx  Shalizi 
september 2017 by gideonite
trees are harlequins, words are harlequins — bayes: a kinda-sorta masterpost
lol, gwern:
> What sort of person thinks “oh yeah, my beliefs about these coefficients correspond to a Gaussian with variance 2.5″? And what if I do cross-validation, like I always do, and find that variance 200 works better for the problem? Was the other person wrong? But how could they have known?
> ...Even ignoring the mode vs. mean issue, I have never met anyone who could tell whether their beliefs were normally distributed vs. Laplace distributed. Have you?
I must have spent too much time in Bayesland because both those strike me as very easy and I often think them! My beliefs usually are Laplace distributed when it comes to things like genetics (it makes me very sad to see GWASes with flat priors), and my Gaussian coefficients are actually a variance of 0.70 (assuming standardized variables w.l.o.g.) as is consistent with field-wide meta-analyses indicating that d>1 is pretty rare.
ratty  ssc  core-rats  tumblr  social  explanation  init  philosophy  bayesian  thinking  probability  stats  frequentist  big-yud  lesswrong  synchrony  similarity  critique  intricacy  shalizi  scitariat  selection  mutation  evolution  priors-posteriors  regularization  bias-variance  gwern  reddit  commentary  GWAS  genetics  regression  spock  nitty-gritty  generalization  epistemic  🤖  rationality  poast  multi  best-practices  methodology  data-science 
august 2017 by nhaliday
Taiping Rebellion - Wikipedia
The Taiping Rebellion or Taiping Civil War (simplified Chinese: 太平天国运动; traditional Chinese: 太平天國運動; pinyin: Taìpíng Tīanguó Yùndòng, literally "Taiping Heavenly Kingdom Movement") was a massive rebellion or civil war in China that lasted from 1850 to 1864 fought between the established Manchu-led Qing dynasty and the millenarian movement of the Heavenly Kingdom of Peace.[8][9]

never heard of this before
history  early-modern  china  asia  sinosphere  revolution  class  religion  theos  article  wiki  reference  war  pre-ww2  christianity  multi  shalizi  scitariat  books  summary  review 
june 2017 by nhaliday

« earlier    

related tags

**  aa  academia  accretion  acm  ai  analysis  anthropology  aphorism  article  asia  axelrod  bactra  bayes  bayesian  bayesian_cognition  bayesianism  behavioral-econ  best-practices  bias-variance  big-peeps  big-yud  bio  blogs  boltzmann  book  book_review  books  bounded_rationality  broad-econ  c:*  canon  capitalism  cartoons  cass_sunstein  causal_inference  chaos  china  christianity  christopher_norris  class  clustering  cmu  cog-psych  cognitive_science  commentary  complex_systems  complexity  computation  concept  confluence  cooperate-defect  core-rats  cosma  cosma_shalizi  cosmo  course  critique  cross_validation  cshalizi  cultural-dynamics  cultural_evolution  culture  dan_sperber  data-mining  data-science  data  data_analysis  dawkins  debate  deborarh_mayo  decision-making  deep-materialism  delong  democracy  dependence-independence  design  devonthink  donald_norman  dt  early-modern  econ  economics  education  efficiency  enlightenment-renaissance-restoration-reformation  entropy-like  entropy  epistemic  equilibrium  ergodic  error  errors  essay  europe  evolution  evolutionary_psychology  experiment  expert-experience  expert  explanation  exposition  extrema  fluid  franco  frequentist  freud  gam  gary_king  gaussian-mixture-model  gelman  generalization  generalized_additive_models  genetics  giants  gmt  graphical_causal_models  graphical_models  graphsmapstrees  group_selection  gt-101  gwas  gwern  have-read  henry_farrell  history  history_of_thought  homepage  iidness  inequality  info_mgmt  information-theory  init  institutions  intelligence  interdisciplinary  interests  intricacy  iq  iron-age  john_tyler_bonner  justice  kickstarter  knowledge  lacan  large-factor  lecture-notes  lens  lesswrong  letters  libertarian_paternalism  liberty  limits  links  list  lit  local-global  lol  machine-learning  machinelearning  markets  markov  marx  math  mathematics  mayo  mechanics  medieval  meme  methodology  microfoundations  mixing  mixture  ml  model-selection  model  modeling  modernity  moretti  mostly-modern  multi  multinomial  mutation  mysticism  n-factor  navier-stokes  nibble  nitty-gritty  notebooks  notes  nudge  old-anglo  optimization  org:bleg  org:mat  organistions  p:someday  paper  paul_krugman  pca  pdf  peirce  people  philosophy  philosophy_of_science  philosophy_of_statistics  physics  poast  policy  political_economy  political_philosophy  politics  polymath  postmodernism  powerlaw  pre-ww2  preprint  presentation  priors-posteriors  probability  prof  programming  psych-architecture  psychoanalysis  psychology  psychometrics  public-speaking  quixotic  quotes  r  rationality  ratty  reading  recommendations  reddit  reference  references?  references  regression  regular_expressions  regularization  regularizer  relativedistributions  relativism  religion  review  revolution  richard_dawkins  role_of_government  russia  s:*  samuel_bowles  science  scifi  scitariat  sde  selection  self-interest  self-organizing  similarity  singularity  sinosphere  social-norms  social  socialism  society  space-time  spatial  spock  ssc  stat-mech  statistical_learning  statistics  stats  steven_pinker  stochastic-processes  stochastic  structured  study  summary  sunstein  susan_haack  syllabus  synchrony  taxes  teaching  technology  temporal  the-great-west-whale  the-trenches  theos  thermo  thermodynamics  thinking  time_series  to:read  todo  top-n  troll  tumblr  turbulence  twitter  unit  university  war  welfare_economics  wiki  writing  🌞  🎩  🔬  🤖 

Copy this bookmark: