środa, 23 lutego 2011

More list-monadic sundries


let concat_map f l =
  let rec cmap_f accu = function
    | [] -> accu
    | a::l -> cmap_f (List.rev_append (f a) accu) l
  in
  List.rev (cmap_f [] l)

let rec concat_foldr f l init =
  match l with
    | [] -> init
    | a::l -> concat_map (f a) (concat_foldr f l init)

let map_reduce mapf redf red0 l =
  match List.sort (fun x y -> compare (fst x) (fst y))
    (List.map mapf l) with
      | [] -> []
      | (k0, v0)::tl ->
        let k0, vs, l =
          List.fold_left (fun (k0, vs, l) (kn, vn) ->
     if k0 = kn then k0, vn::vs, l
            else kn, [vn], (k0,vs)::l)
     (k0, [v0], []) tl in
        List.rev_map (fun (k,vs) -> k, List.fold_left redf red0 vs)
          ((k0,vs)::l)

let rec update_assoc k v0 f l =
  let rec aux acc = function
    | [] -> List.rev_append acc [k, f v0]
    | (a, b as pair) :: l ->
      if a = k then List.rev_append acc ((k, f b)::l)
      else aux (pair :: acc) l in
  aux [] l

let cons e l = e::l

wtorek, 18 stycznia 2011

Metaphysics

I believe that:

  • other people exist and (some of them) have minds roughly like mine
  • the explanations of science (e.g. evolutionary) are factual should be taken at face value ("naturalistic ontological attitude")
  • the reality is based on a monistic incompletely intelligible (substrate) plane
    • "the map is not the territory, but the map is part of the territory"
    • there is no essential distinction between the observational knowledge and the logical knowledge, they are "limiting cases" of each other (one always only observes one's mental reasoning, or calculator outputs, or publications; Laplace's demon)
    • knowledge is fallible and approximate and (possibly) objective; there is no separate unknowable world of things in themselves even if/though we only know an infinitesimal amount of what exists; a complete effective description of what exists is not possible (attainable)
  • personal God does not make sense philosophically (the attributes are contradictory) (one can't be omniscient and be a person)
  • some ways of life are better than others
  • "what should I do?" for a particular person in particular context (which is what really matters) can be answered both rationally and by intuition (e.g. conscience)
  • persons are not ontologically primitive but are complex systems (Marvin Minsky on qualia) (i.e. it's not complicated person contents running on the basic element of personhood, it's the personhood that is complex -- but not necessarily complicated); yet there is something "looking simple" to empathically know in that we do have experiences (see non-eliminative reductionism) -- only that it is not simple to be someone capable of having given experience (and experiences are participatory)
  • people are (possibly) Turing-computable and the "problem of free will" is adequately solved by Daniel Dennett, Gary Drescher, Eliezer Yudkovsky and their ilk
  • categories of systemic metaphysics are useful conceptualisations but there is nothing privileged about them 
  • objects of mathematics exist, but non-constructive objects are grounded in constructive objects (have weaker ontological status)
  • nevertheless they "exist more" when they are more worth of pursuit
  • http://en.wikipedia.org/wiki/Neopragmatism#Background
  • "discovering" and "creating" stress different aspects of the same process. Creation is the (amount of) determination on the variable (pointing to what is to be created-discovered) attributable to the creator-discoverer and discovery is the (amount of) determination on the variable attributable to the object that instantiates it. Determination "is in the description", so for a somewhat disputable example, intuitionistically a mathematician is a creator, and classically a discoverer
  • we are responsible for discovering (creating) our values; we are responsible for pursuing goals in accordance with our values
  • the exact "lossless" complexity of value may be high, but there is a pattern: values promote versatile long-term efficiency of the society at achieving goals its free individuals might reasonably set out to
    • the above notion of efficiency embeds the "exploration principle" that diversity has an instrumental value, and "non-waste / irreversibility principle" that cooperation and negotiation have more instrumental value than killing
    • I'm not sympathetic with the "complexity of value" view; I'd say, of course, "we value complexity of what we value", and our practical desires increase in complexity; but the process of valuing is an intelligible open system that has low-complexity approximate descriptions (compare: the universe and its equations); see some posts at Becoming Gaia
  • (broad vs. narrow ethics) unlawful -- unethical "in the narrow sense" -- is that which infringes upon the free will of entities who do not so infringe other entities; failing to address (fulfill) one's responsibilities is unethical "in the broad sense"
    • not all will is free, it is possible to know another person well enough to perform coercive action that promotes the freeness of their will
    • as we grow stronger together these two views become more coextensive
  • pleasure is never a value, because it is a cognitive tool to navigate values (hedonism is a category error); similar comment applies to happiness, but the error there is less dangerous because happiness is an abstract notion and "I value happiness" is a tautology
    • the same argument covers the error of valuing individual experiencing (aka. indexical utility); the egocentricity of value is better explained by spatiotemporal discounting approach to dealing with uncertainty
    • in case someone hasn't grasped it yet: I'm saying that pleasure is a measurement tool like a thermometer, which can be broken or be tampered with, and there are other ways to infer what the temperature is, although still the thermometer is usually good at telling the temperature
  • this universe is not likely created by an agenty process. Why would they do it? it's not that I hate the creator: the universe is quite neat! but I think that the following self-defeating argument is invalid: "the ability to get better is good; therefore a better universe is not better because it has less room to get better". A higher existence would better spend its resources better, in more interconnected ways, on furthering its "highness". An agenty process is more likely to discover this universe, but this is inconsequential to us (the likelihood of inconsequence is strengthened by my monistic view of creation-discovery)
  • reiterating some points: everyone is guilty of not being as good as he or she could possibly be
  • pursuing in the intuitionistic direction to distinguish possibly impossible "possible worlds" from meaningless impossible "possible worlds" puts some ontological weight on intuitionism that classical logic lacks
  • past, present and future (or futures, according to quantum physics) exist (why care for something that does not exist?)
    • this point of view vindicates the ontological status of classical logic somewhat
  • persons are continuous over time, but a self is not extended over time (for more than five minutes, say) -- a self is a "tangent space" over a person, existing only for a single decision
    • not talking above about the psychological self, but about the proto-philosophical identity-bearing "I"
  • in "Permutation City", it was wrong for them to cut themselves off (their reality mass went back to "score zero" when they discorrelated themselves from our universe)
    • [I'll add more later]
    Of some of the points I'm less sure than of others. (I've removed some points from the initial post. The stroke-over text is essentially deleted but kept for continuity with other points or comments; it does not mean its negation is stated.)