no-go   32

Applications of computational learning theory in the cognitive sciences - Psychology & Neuroscience Stack Exchange
1. Gold's theorem on the unlearnability in the limit of certain sets of languages, among them context-free ones.

2. Ronald de Wolf's master's thesis on the impossibility to PAC-learn context-free languages.

The first made quiet a stir in the poverty-of-the-stimulus debate, and the second has been unnoticed by cognitive science.
q-n-a  stackex  psychology  cog-psych  learning  learning-theory  machine-learning  PAC  lower-bounds  no-go  language  linguistics  models  fall-2015 
26 days ago by nhaliday
Complexity no Bar to AI - Gwern.net
Critics of AI risk suggest diminishing returns to computing (formalized asymptotically) means AI will be weak; this argument relies on a large number of questionable premises and ignoring additional resources, constant factors, and nonlinear returns to small intelligence advantages, and is highly unlikely. (computer science, transhumanism, AI, R)
created: 1 June 2014; modified: 01 Feb 2018; status: finished; confidence: likely; importance: 10
ratty  gwern  analysis  faq  ai  risk  speedometer  intelligence  futurism  cs  computation  complexity  tcs  linear-algebra  nonlinearity  convexity-curvature  average-case  adversarial  article  time-complexity  singularity  iteration-recursion  magnitude  multiplicative  lower-bounds  no-go  performance  hardware  humanity  psychology  cog-psych  psychometrics  iq  distribution  moments  complement-substitute  hanson  ems  enhancement  parable  detail-architecture  universalism-particularism  neuro  ai-control  environment  climate-change  threat-modeling  security  theory-practice  hacker  academia  realness  crypto  rigorous-crypto  usa  government 
april 2018 by nhaliday
If Quantum Computers are not Possible Why are Classical Computers Possible? | Combinatorics and more
As most of my readers know, I regard quantum computing as unrealistic. You can read more about it in my Notices AMS paper and its extended version (see also this post) and in the discussion of Puzzle 4 from my recent puzzles paper (see also this post). The amazing progress and huge investment in quantum computing (that I presented and update  routinely in this post) will put my analysis to test in the next few years.
tcstariat  mathtariat  org:bleg  nibble  tcs  cs  computation  quantum  volo-avolo  no-go  contrarianism  frontier  links  quantum-info  analogy  comparison  synthesis  hi-order-bits  speedometer  questions  signal-noise 
november 2017 by nhaliday
Europa, Enceladus, Moon Miranda | West Hunter
A lot of ice moons seem to have interior oceans, warmed by tidal flexing and possibly radioactivity.  But they’re lousy candidates for life, because you need free energy; and there’s very little in the interior oceans of such system.

It is possible that NASA is institutionally poor at pointing this out.
west-hunter  scitariat  discussion  ideas  rant  speculation  prediction  government  dirty-hands  space  xenobio  oceans  fluid  thermo  phys-energy  temperature  no-go  volo-avolo  physics  equilibrium  street-fighting  nibble  error  track-record  usa  bio  eden  cybernetics  complex-systems 
september 2017 by nhaliday
Correlated Equilibria in Game Theory | Azimuth
Given this, it’s not surprising that Nash equilibria can be hard to find. Last September a paper came out making this precise, in a strong way:

• Yakov Babichenko and Aviad Rubinstein, Communication complexity of approximate Nash equilibria.

The authors show there’s no guaranteed method for players to find even an approximate Nash equilibrium unless they tell each other almost everything about their preferences. This makes finding the Nash equilibrium prohibitively difficult to find when there are lots of players… in general. There are particular games where it’s not difficult, and that makes these games important: for example, if you’re trying to run a government well. (A laughable notion these days, but still one can hope.)

Klarreich’s article in Quanta gives a nice readable account of this work and also a more practical alternative to the concept of Nash equilibrium. It’s called a ‘correlated equilibrium’, and it was invented by the mathematician Robert Aumann in 1974. You can see an attempt to define it here:
baez  org:bleg  nibble  mathtariat  commentary  summary  news  org:mag  org:sci  popsci  equilibrium  GT-101  game-theory  acm  conceptual-vocab  concept  definition  thinking  signaling  coordination  tcs  complexity  communication-complexity  lower-bounds  no-go  liner-notes  big-surf  papers  research  algorithmic-econ  volo-avolo 
july 2017 by nhaliday
Predicting the outcomes of organic reactions via machine learning: are current descriptors sufficient? | Scientific Reports
As machine learning/artificial intelligence algorithms are defeating chess masters and, most recently, GO champions, there is interest – and hope – that they will prove equally useful in assisting chemists in predicting outcomes of organic reactions. This paper demonstrates, however, that the applicability of machine learning to the problems of chemical reactivity over diverse types of chemistries remains limited – in particular, with the currently available chemical descriptors, fundamental mathematical theorems impose upper bounds on the accuracy with which raction yields and times can be predicted. Improving the performance of machine-learning methods calls for the development of fundamentally new chemical descriptors.
study  org:nat  papers  machine-learning  chemistry  measurement  volo-avolo  lower-bounds  analysis  realness  speedometer  nibble  🔬  applications  frontier  state-of-art  no-go  accuracy  interdisciplinary 
july 2017 by nhaliday
What is the likelihood we run out of fossil fuels before we can switch to renewable energy sources? - Quora
1) Can we de-carbon our primary energy production before global warming severely damages human civilization? In the short term this means switching from coal to natural gas, and in the long term replacing both coal and gas generation with carbon-neutral sources such as renewables or nuclear. The developed world cannot accomplish this alone -- it requires worldwide action, and most of the pain will be felt by large developing nations such as India and China. Ultimately this is a political and economic problem. The technology to eliminate most carbon from electricity generation exists today at fairly reasonable cost.

2) Can we develop a better transportation energy storage technology than oil, before market forces drive prices to levels that severely damage the global economy? Fossil fuels are a source of energy, but primarily we use oil in vehicles because it is an exceptional energy TRANSPORT medium. Renewables cannot meet this need because battery technology is completely uncompetitive for most fuel consumers -- prices are an order of magnitude too high and energy density is an order of magnitude too low for adoption of all-electric vehicles outside developed-world urban centers. (Heavy trucking, cargo ships, airplanes, etc will never be all-electric with chemical batteries. There are hard physical limits to the energy density of electrochemical reactions. I'm not convinced passenger vehicles will go all-electric in our lifetimes either.) There are many important technologies in existence that will gain increasing traction in the next 50 years such as natural gas automobiles and improved gas/electric hybrids, but ultimately we need a better way to store power than fossil fuels. _This is a deep technological problem that will not be solved by incremental improvements in battery chemistry or any process currently in the R&D pipeline_.

Based on these two unresolved issues, _I place the odds of us avoiding fossil-fuel-related energy issues (major climate or economic damage) at less than 10%_. The impetus for the major changes required will not be sufficiently urgent until the world is seeing severe and undeniable impacts. Civilization will certainly survive -- but there will be no small amount of human suffering during the transition to whatever comes next.

- Ryan Carlyle
q-n-a  qra  expert  energy-resources  climate-change  environment  risk  civilization  nihil  prediction  threat-modeling  world  futurism  biophysical-econ  stock-flow  transportation  technology  economics  long-short-run  no-go  speedometer  modernity  expert-experience 
may 2017 by nhaliday
Futuristic Physicists? | Do the Math
interesting comment: https://westhunt.wordpress.com/2014/03/05/outliers/#comment-23087
referring to timelines? or maybe also the jetpack+flying car (doesn't seem physically impossible; at most impossible for useful trip lengths)?

Topic Mean % pessim. median disposition
1. Autopilot Cars 1.4 (125 yr) 4 likely within 50 years
15. Real Robots 2.2 (800 yr) 10 likely within 500 years
13. Fusion Power 2.4 (1300 yr) 8 likely within 500 years
10. Lunar Colony 3.2 18 likely within 5000 years
16. Cloaking Devices 3.5 32 likely within 5000 years
20. 200 Year Lifetime 3.3 16 maybe within 5000 years
11. Martian Colony 3.4 22 probably eventually (>5000 yr)
12. Terraforming 4.1 40 probably eventually (> 5000 yr)
18. Alien Dialog 4.2 42 probably eventually (> 5000 yr)
19. Alien Visit 4.3 50 on the fence
2. Jetpack 4.1 64 unlikely ever
14. Synthesized Food 4.2 52 unlikely ever
8. Roving Astrophysics 4.6 64 unlikely ever
3. Flying “Cars” 3.9 60 unlikely ever
7. Visit Black Hole 5.1 74 forget about it
9. Artificial Gravity 5.3 84 forget about it
4. Teleportation 5.3 85 forget about it
5. Warp Drive 5.5 92 forget about it
6. Wormhole Travel 5.5 96 forget about it
17. Time Travel 5.7 92 forget about it
org:bleg  nibble  data  poll  academia  higher-ed  prediction  speculation  physics  technology  gravity  geoengineering  space  frontier  automation  transportation  energy-resources  org:edu  expert  scitariat  science  no-go  big-picture  wild-ideas  the-world-is-just-atoms  applications  multi  west-hunter  optimism  pessimism  objektbuch  regularizer  s:*  c:**  🔬  poast  ideas  speedometer  whiggish-hegelian  scifi-fantasy  expert-experience  expansionism 
march 2017 by nhaliday
Redistributing from Capitalists to Workers: An Impossibility Theorem, Garett Jones | EconLog | Library of Economics and Liberty
org:econlib  econotariat  spearhead  garett-jones  economics  policy  rhetoric  thinking  analysis  no-go  redistribution  labor  taxes  cracker-econ  multi  piketty  news  org:lite  org:biz  pdf  links  political-econ  capital  simulation  operational  dynamic  explanation  time-preference  patience  wonkish  study  science-anxiety  externalities  long-short-run  models  map-territory  stylized-facts  s:*  broad-econ  chart  article  🎩  randy-ayndy  envy  bootstraps  inequality  absolute-relative  X-not-about-Y  volo-avolo  ideas  status  capitalism  nationalism-globalism  metabuch  optimate  aristos  open-closed  macro  government  proofs  equilibrium 
february 2017 by nhaliday
6.896: Essential Coding Theory
- probabilistic method and Chernoff bound for Shannon coding
- probabilistic method for asymptotically good Hamming codes (Gilbert coding)
- sparsity used for LDPC codes
mit  course  yoga  tcs  complexity  coding-theory  math.AG  fields  polynomials  pigeonhole-markov  linear-algebra  probabilistic-method  lecture-notes  bits  sparsity  concentration-of-measure  linear-programming  linearity  expanders  hamming  pseudorandomness  crypto  rigorous-crypto  communication-complexity  no-go  madhu-sudan  shannon  unit  p:**  quixotic 
february 2017 by nhaliday
Science Policy | West Hunter
If my 23andme profile revealed that I was the last of the Plantagenets (as some suspect), and therefore rightfully King of the United States and Defender of Mexico, and I asked you for a general view of the right approach to science and technology – where the most promise is, what should be done, etc – what would you say?

genetically personalized medicine: https://westhunt.wordpress.com/2016/12/08/science-policy/#comment-85698
I have no idea how personalized medicine is supposed to work. Suppose that we sequence your entire genome, and then we intend to tailor a therapeutic approach to your genome.

How do we test it? By trying it on a bunch of genetically similar people? The more genetic details we take into account, the smaller that class is. It could easily become so small that it would be difficult to recruit enough people for a reasonable statistical trial. Second, the more details we take into account, the smaller the class that benefits from the whole testing process – which as far as I can see, is just as expensive as conventional Phasei/II etc trials.

What am I missing?

Now if you are a forethoughtful trillionaire, sure: you manufacture lots of clones just to test therapies you might someday need, and cost is no object.

I think I can see ways you could make it work tho [edit: what did I mean by this?...damnit]
west-hunter  discussion  politics  government  policy  science  technology  the-world-is-just-atoms  🔬  scitariat  meta:science  proposal  genetics  genomics  medicine  meta:medicine  multi  ideas  counter-revolution  poast  homo-hetero  generalization  scale  antidemos  alt-inst  applications  dimensionality  high-dimension  bioinformatics  no-go  volo-avolo  magnitude  trump  2016-election  questions 
december 2016 by nhaliday

related tags

2016-election  aaronson  absolute-relative  academia  accuracy  acm  adversarial  aesthetics  age-of-discovery  ai-control  ai  algebraic-complexity  algorithmic-econ  alt-inst  analogy  analysis  antidemos  apollonian-dionysian  applications  archaeology  architecture  aristos  article  asia  automation  average-case  baez  beauty  big-peeps  big-picture  big-surf  bio  bioinformatics  biophysical-econ  bits  boaz-barak  bootstraps  bostrom  bounded-cognition  broad-econ  c:**  capital  capitalism  chart  chemistry  china  circuits  civilization  clarity  climate-change  cocktail  coding-theory  cog-psych  commentary  communication-complexity  communication  comparison  complement-substitute  complex-systems  complexity  composition-decomposition  computation  concentration-of-measure  concept  conceptual-vocab  concurrency  consilience  contradiction  contrarianism  convexity-curvature  coordination  counter-revolution  course  cracker-econ  critique  crypto  cs  cybernetics  cycles  data  death  deep-learning  deep-materialism  definition  dennett  density  detail-architecture  digitalspy  dimensionality  dirty-hands  discrete  discussion  disease  distribution  domestication  dynamic  dynamical  early-modern  econometrics  economics  econotariat  eden-heaven  eden  efficiency  electromag  elite  ems  energy-resources  enhancement  entropy-like  environment  envy  epistemic  equilibrium  error  essay  estates  europe  evolution  examples  expanders  expansionism  expert-experience  expert  explanans  explanation  exposition  externalities  fall-2015  faq  fiction  fields  finiteness  fluid  flux-stasis  food  forums  france  french  frequency  frontier  funny  futurism  game-theory  garett-jones  gedanken  generalization  genetics  genomics  geoengineering  germanic  giants  gilens-page  government  gravity  ground-up  gt-101  gwern  hacker  hamming  hanson  hard-tech  hardware  hetero-advantage  hi-order-bits  high-dimension  higher-ed  history  homepage  homo-hetero  humanity  ideas  ideology  illusion  inequality  inference  info-dynamics  information-theory  insight  instinct  intelligence  interdisciplinary  interests  internet  intricacy  iq  islam  iteration-recursion  janus  jindal  labor  language  learning-theory  learning  lecture-notes  len:long  lens  linear-algebra  linear-programming  linearity  liner-notes  linguistics  links  london  long-short-run  lower-bounds  luca-trevisan  machine-learning  macro  madhu-sudan  magnitude  map-territory  math.ag  math.ds  mathtariat  measure  measurement  mechanics  medicine  meta:medicine  meta:science  metabuch  mit  models  modernity  moments  multi  multiplicative  muslims  nationalism-globalism  naturality  network-structure  neuro  news  nibble  nietzschean  nihil  nitty-gritty  nogo  noise-structure  nonlinearity  objektbuch  occam  oceans  off-convex  open-closed  operational  optimate  optimism  org:biz  org:bleg  org:econlib  org:edu  org:lite  org:mag  org:mat  org:nat  org:sci  organizing  overflow  p:**  p:whenever  pac  papers  parable  parasites-microbiome  parsimony  patience  pdf  performance  pessimism  philosophy  phys-energy  physics  pigeonhole-markov  piketty  poast  policy  polisci  political-econ  politics  poll  polynomials  popsci  power  prediction  preprint  presentation  probabilistic-method  proof-systems  proofs  proposal  pseudorandomness  psychology  psychometrics  q-n-a  qra  quantitative-qualitative  quantum-info  quantum-money  quantum  questions  quixotic  quotes  randy-ayndy  rant  ratty  realness  reason  redistribution  reference  reflection  regularizer  relativity  relativization  research-program  research  retention  rhetoric  rigor  rigorous-crypto  risk  robust  roots  rough  s:*  sapiens  scale  science-anxiety  science  scifi-fantasy  scitariat  security  sex  sexuality  shannon  shift  signal-noise  signaling  simulation  singularity  skunkworks  smoothness  social-science  sociology  soft-question  software  space-complexity  space  sparsity  spatial  spearhead  speculation  speed  speedometer  stackex  state-of-art  state  status  stock-flow  stories  street-fighting  structure  study  stylized-facts  summary  synchrony  synthesis  talks  taxes  tcs  tcstariat  technology  temperature  the-bones  the-self  the-trenches  the-world-is-just-atoms  theory-practice  thermo  thick-thin  thinking  threat-modeling  time-complexity  time-preference  time  traces  track-record  trade  transportation  trivia  trump  unit  universalism-particularism  usa  vaclav-smil  video  visual-understanding  volo-avolo  web  west-hunter  whiggish-hegelian  whole-partial-many  wigderson  wiki  wild-ideas  within-without  wonkish  world  x-not-about-y  xenobio  yoga  zones  🎩  🔬 

Copy this bookmark:



description:


tags: