nhaliday + hardness   14

Moravec's paradox - Wikipedia
Moravec's paradox is the discovery by artificial intelligence and robotics researchers that, contrary to traditional assumptions, high-level reasoning requires very little computation, but low-level sensorimotor skills require enormous computational resources. The principle was articulated by Hans Moravec, Rodney Brooks, Marvin Minsky and others in the 1980s. As Moravec writes, "it is comparatively easy to make computers exhibit adult level performance on intelligence tests or playing checkers, and difficult or impossible to give them the skills of a one-year-old when it comes to perception and mobility".[1]

Similarly, Minsky emphasized that the most difficult human skills to reverse engineer are those that are unconscious. "In general, we're least aware of what our minds do best", he wrote, and added "we're more aware of simple processes that don't work well than of complex ones that work flawlessly".[2]


One possible explanation of the paradox, offered by Moravec, is based on evolution. All human skills are implemented biologically, using machinery designed by the process of natural selection. In the course of their evolution, natural selection has tended to preserve design improvements and optimizations. The older a skill is, the more time natural selection has had to improve the design. Abstract thought developed only very recently, and consequently, we should not expect its implementation to be particularly efficient.

As Moravec writes:

Encoded in the large, highly evolved sensory and motor portions of the human brain is a billion years of experience about the nature of the world and how to survive in it. The deliberate process we call reasoning is, I believe, the thinnest veneer of human thought, effective only because it is supported by this much older and much more powerful, though usually unconscious, sensorimotor knowledge. We are all prodigious olympians in perceptual and motor areas, so good that we make the difficult look easy. Abstract thought, though, is a new trick, perhaps less than 100 thousand years old. We have not yet mastered it. It is not all that intrinsically difficult; it just seems so when we do it.[3]

A compact way to express this argument would be:

- We should expect the difficulty of reverse-engineering any human skill to be roughly proportional to the amount of time that skill has been evolving in animals.
- The oldest human skills are largely unconscious and so appear to us to be effortless.
- Therefore, we should expect skills that appear effortless to be difficult to reverse-engineer, but skills that require effort may not necessarily be difficult to engineer at all.
concept  wiki  reference  paradox  ai  intelligence  reason  instinct  neuro  psychology  cog-psych  hardness  logic  deep-learning  time  evopsych  evolution  sapiens  the-self  EEA  embodied  embodied-cognition  abstraction  universalism-particularism  gnosis-logos  robotics 
june 2018 by nhaliday
Shtetl-Optimized » Blog Archive » Why I Am Not An Integrated Information Theorist (or, The Unconscious Expander)
In my opinion, how to construct a theory that tells us which physical systems are conscious and which aren’t—giving answers that agree with “common sense” whenever the latter renders a verdict—is one of the deepest, most fascinating problems in all of science. Since I don’t know a standard name for the problem, I hereby call it the Pretty-Hard Problem of Consciousness. Unlike with the Hard Hard Problem, I don’t know of any philosophical reason why the Pretty-Hard Problem should be inherently unsolvable; but on the other hand, humans seem nowhere close to solving it (if we had solved it, then we could reduce the abortion, animal rights, and strong AI debates to “gentlemen, let us calculate!”).

Now, I regard IIT as a serious, honorable attempt to grapple with the Pretty-Hard Problem of Consciousness: something concrete enough to move the discussion forward. But I also regard IIT as a failed attempt on the problem. And I wish people would recognize its failure, learn from it, and move on.

In my view, IIT fails to solve the Pretty-Hard Problem because it unavoidably predicts vast amounts of consciousness in physical systems that no sane person would regard as particularly “conscious” at all: indeed, systems that do nothing but apply a low-density parity-check code, or other simple transformations of their input data. Moreover, IIT predicts not merely that these systems are “slightly” conscious (which would be fine), but that they can be unboundedly more conscious than humans are.

To justify that claim, I first need to define Φ. Strikingly, despite the large literature about Φ, I had a hard time finding a clear mathematical definition of it—one that not only listed formulas but fully defined the structures that the formulas were talking about. Complicating matters further, there are several competing definitions of Φ in the literature, including ΦDM (discrete memoryless), ΦE (empirical), and ΦAR (autoregressive), which apply in different contexts (e.g., some take time evolution into account and others don’t). Nevertheless, I think I can define Φ in a way that will make sense to theoretical computer scientists. And crucially, the broad point I want to make about Φ won’t depend much on the details of its formalization anyway.

We consider a discrete system in a state x=(x1,…,xn)∈Sn, where S is a finite alphabet (the simplest case is S={0,1}). We imagine that the system evolves via an “updating function” f:Sn→Sn. Then the question that interests us is whether the xi‘s can be partitioned into two sets A and B, of roughly comparable size, such that the updates to the variables in A don’t depend very much on the variables in B and vice versa. If such a partition exists, then we say that the computation of f does not involve “global integration of information,” which on Tononi’s theory is a defining aspect of consciousness.
aaronson  tcstariat  philosophy  dennett  interdisciplinary  critique  nibble  org:bleg  within-without  the-self  neuro  psychology  cog-psych  metrics  nitty-gritty  composition-decomposition  complex-systems  cybernetics  bits  information-theory  entropy-like  forms-instances  empirical  walls  arrows  math.DS  structure  causation  quantitative-qualitative  number  extrema  optimization  abstraction  explanation  summary  degrees-of-freedom  whole-partial-many  network-structure  systematic-ad-hoc  tcs  complexity  hardness  no-go  computation  measurement  intricacy  examples  counterexample  coding-theory  linear-algebra  fields  graphs  graph-theory  expanders  math  math.CO  properties  local-global  intuition  error  definition 
january 2017 by nhaliday
Bottoming Out – arg min blog
Now, I’ve been hammering the point in my previous posts that saddle points are not what makes non-convex optimization difficult. Here, when specializing to deep learning, even local minima are not getting in my way. Deep neural nets are just very easy to minimize.
machine-learning  deep-learning  optimization  rhetoric  speculation  research  hmm  research-program  acmtariat  generalization  metabuch  local-global  off-convex  ben-recht  extrema  org:bleg  nibble  sparsity  curvature  ideas  aphorism  convexity-curvature  explanans  volo-avolo  hardness 
june 2016 by nhaliday

bundles : abstracttcs

related tags

aaronson  abstraction  acmtariat  ai  algebraic-complexity  algorithmic-econ  algorithms  analogy  announcement  aphorism  approximation  arrows  ben-recht  big-list  big-picture  big-surf  bits  causation  cheatsheet  circuits  coding-theory  cog-psych  communication-complexity  complex-systems  complexity  composition-decomposition  computation  concept  convexity-curvature  counterexample  course  critique  crypto  cs  curvature  cybernetics  database  deep-learning  definition  degrees-of-freedom  dennett  EEA  embodied  embodied-cognition  empirical  entropy-like  equilibrium  erik-demaine  error  evidence  evolution  evopsych  examples  expanders  explanans  explanation  exposition  extrema  fields  forms-instances  frontier  game-theory  generalization  gnosis-logos  graph-theory  graphs  hard-core  hardness  hmm  ideas  information-theory  instinct  intelligence  interdisciplinary  intricacy  intuition  jargon  knowledge  lecture-notes  lectures  linear-algebra  list  local-global  logic  lower-bounds  machine-learning  math  math.CO  math.DS  math.RT  measurement  meta:math  metabuch  metrics  mit  motivation  network-structure  neuro  news  nibble  nitty-gritty  no-go  number  objektbuch  off-convex  open-problems  optimization  org:bleg  org:edu  org:inst  org:junk  org:mag  org:sci  overflow  p:whenever  paradox  pdf  philosophy  popsci  problem-solving  profile  proofs  properties  psychology  q-n-a  quantitative-qualitative  questions  rand-approx  reason  reduction  reference  relativization  research  research-program  rhetoric  rigorous-crypto  robotics  sapiens  SDP  slides  soft-question  sparsity  speculation  strings  structure  summary  symmetry  synthesis  systematic-ad-hoc  talks  tcs  tcstariat  the-self  time  time-complexity  topics  tricki  UGC  unit  universalism-particularism  video  volo-avolo  walls  whole-partial-many  wiki  within-without  yoga 

Copy this bookmark: