495200+ entries in 0.372s
ascii_field: but
that i knew precisely what must be done
to make a computer a
tolerable
thing, most of a decade ago.
ascii_field: did not say
that it is a definite guide (to what? l0l)
mircea_popescu: be happy if you manage
to ever actually make it, or at least nudge
towards it. 25 or 55.
ascii_field: mircea_popescu: largely because i am not mircea_popescu and not in control of
the variables.
ascii_field: sure, could have
tripped and fell
to death in
the bath.
ascii_field: and i could have easily built all of
this by 25 y.o. if martians had landed and given me
the means.
ascii_field: only 'useful'
to
the extent a lathe from newton's
time was useful
to creating
today's.
mircea_popescu: but
this is good news : new stuff is cheaper
to build
than retooling old stuff.
mircea_popescu: yeah, for none of
this is any of
the current
toolchain useful, from plant
to emacs.
ascii_field: most of it does not need
to be addressible at all
mircea_popescu: incidnetally - would you also have
the selective inhibition design ?
mircea_popescu: but i have no doubt
this is correct, chiefly because
that';s how
the brain works.
mircea_popescu: ie, you'd end up with half
the space or more
taken up by indexes
mircea_popescu: if you do
that
the current addressing schemes would never work
ascii_field: mircea_popescu: what'd you have it do << a number of
things, which resolve
to 'the contents ending up where needed by dependency-graph without being asked every
time'
ascii_field: bits expensively stored as charge on si ought
to be available for computing asynchronously.
ascii_field: memory as a device
that just sits
there, is a 20th c. idiocy.
ascii_field: (before anyone spits back
the old paper re:
the fact of c-gates implemented on xilinx fabric - go and see how many of
them you can fit. and how much room left for interconnects.)
gabriel_laddel: ascii_field: o. I
thought
that you could if
they have an entirely known design?
ascii_field: gabriel_laddel: not exactly. even if you could secure an infinite supply of
the particular fpga, it
turns out
that
they are not actually very general. in
that you cannot efficiently implement muller c-gates in
them.
mircea_popescu: which is what "not leaking" is, in mathematical
terms.
mircea_popescu: and no, it's not an argument against abstraction period. it's an argument against using abstractive methods outside of
their domain of definition
gabriel_laddel: ascii_field: iirc your argument was "it isn't worthwhile because if you succeed you'll not have enough of
them
to do anything interesting with"
mircea_popescu: gabriel_laddel for one
thing, natural languages can not be modelled in
terms of computer langs.
ascii_field: gabriel_laddel:
to avoid rehashing ancient
thread for a fifth
time, stuck at
the realization
that reversing xilinx is futile.
gabriel_laddel: I can't remember your exact words, but I
think you said in here once "my
time would have been better spent putting
together a demo
than blogging"
ascii_field: and it cannot be done under
the
thumb of vc sc4mz0rz
ascii_field: you are doing another
thing, spray-painted
to resemble 'it'.
ascii_field: gabriel_laddel: don't delude yourself
that you are doing 'it'
gabriel_laddel: ascii_field: we've already discussed
this. I'm for doing it piecemeal because
there are no other options.
ascii_field: gabriel_laddel:
the
type of funding necessary cannot be 'found'
gabriel_laddel: ascii_field: Sometime in
the next 48 hours I'm going
to drop some information in here regarding
the #b-a distro, and when finished with
that move on
to finding funding again.
ascii_field: (short version - it is
too much like
the proverbial 'mars voyage' - counter-economic leap.)
ascii_field: i will confess
that it
took me retardedly long
time
to understand
that my hypothetical rebellion against
the soup cannot be funded or organized for some very fundamental reasons
gabriel_laddel: mircea_popescu: actually, re-reading your last message - isn't
that really an argument against abstraction period?
ascii_field: i've been fighting a largely losing battle for many years, attempting
to convince folks
to imagine - in
their spare
time -
that a computer is not necessarily von neumann and
the 50 yrs of shit soup
gabriel_laddel: ^ I never would have
thought of presentation systems unless I happened upon ascii_field's blog.
ascii_field: 'imprinting' not in
the strict zoological sense, but in
that of having encountered a concept during 'formative years' when a fella is figuring out wtf 'a computer is'
gabriel_laddel: mircea_popescu: nonsene. We know any programming language
that doesn't allow
the operator
to manipulate its own ast programmatically is going
to waste
the operator's
time.
ascii_field is a believer in sapir-whorf
to
the extent
that he very much observed
the braindamage caused by imprinting on certain programming concepts, in a great many people.
mircea_popescu: gabriel_laddel you're on very slippery ice sheet
there. i'll save you
the egg and point out
that you can't form an apriori representation of why X model is no good.
gabriel_laddel: because some of
the $random-african-gruntlangs don't have any means
to describe
the future or past.
ascii_field: what i am speaking of is
the effect i observed where folks learn
to like fundamentally uninteresting concepts.
mircea_popescu: for one
thing, contrary
to common belief, well used holes get better, not worse. woman is not made out of soap.
ascii_field: gabriel_laddel:
there is also such a
thing as
the mental equivalent of stretched arse-rape sphincter
gabriel_laddel: Urbit is probably a good example for what I
think stas is discussing. You can get paid very well so long as you are very active doing nothing.
mircea_popescu: there's a guy who learns and his learning is wasted effort.
there's a different guy who learns for sport.
gabriel_laddel: ascii_field: it is palpable. << Yep. Let's say, for example
that you learn perl. Nothing you learn has any use outside of... writing perl. If you learn javascript, java, C++, scala etc. idem.
The web and
these related "technologies" are going away. All of
the effort poured into learning
them is wasted.
ascii_field: i'm not yet sold on
the fact of
that boy's existence being more fun
than american jail.
ascii_field: some sort of work
that does not demand
too much mental presence
ascii_field: this is why i often say
that in buenos aires i would have
to learn shoemaking or
the like
mircea_popescu: ascii_field
that gets extremely easier once you fuck his gf. you notice
the subtle difference between
the same object as he perceives it, and as it is.
ascii_field: the real reason i will not compete with
the 16 y.o. is not stamina per se, but
that i cannot
trick my brain into actually caring
mircea_popescu: see, and
this is a FINE example of what i mean by "leak like a sieve". your observation was in fact accounted for in my model. because i hadn't leaked it. because it doesn't leak
there.
ascii_field: and
to
this well-known maxim i will add mine, 'giving a fuck about
the problem is worth 100 iq points'
mircea_popescu: perhaps
the cpu also gets
tired, but
the rate of loss due
to
that has not yet been apparent, at least
to me.
mircea_popescu: so : better mental models, less domain lossage - > mega shrinkage of
time and effort required.
mircea_popescu: if you crash less often, you need less computer
time
to do
the same work.
mircea_popescu: another approach
to it is
to point out
that human brains leak like a fucking sieve.
ascii_field: 'the eagle never lost so much as when studied from
the crow' or what was it
mircea_popescu: that's a good approach
to explain it : imagine you can disqualify bad
teachers earlier. well...
that's definitely saving
time.
ascii_field: the spiffiest
tires will not
turn 'trabant' into 'mazerati'
ascii_field: but
this is a rather separate component of effective intelligence
ascii_field: 'knowing what parts
to skip' is -the- wunderwaffen.
mircea_popescu: anyway. i wouldn't want
to go back
to being 20 anymore
than i'd want
to re-take kindergarten.
mircea_popescu: but im not going
to because i see wtf
that is freom a mile away, including
the gender of
the author and her parenting errors in her own kids.
mircea_popescu: and yes, probably
the 22 me would have gotten
through
the fucking with phase in seven weeks and it'd
take me 9 now if i
tried it.
mircea_popescu: ben_vulpes much deeper
than
that. like "i readily see
THE REAL REASON
this is so, so i can skip eighteen paragraphs worth of confused "explicatory" nonsense
mircea_popescu: of course
this could be backed in my case by a solid wall of "i don't have
to do anything i don' want
to do"
ben_vulpes: "this notion is strongly similar
to
these other notions i'm already familiar with, so i'll integrate it more quickly"?
mircea_popescu: depends ben_vulpes.
the problem chiefly is
that as
the mental models get much better, learning speed collapses.
ben_vulpes: mircea_popescu: is
there a difference between how smart/dumb one is and how quickly or slowly one learns?
ascii_field: i - age
two weeks every
time my alarm clock rings.
☟︎ mircea_popescu: im i
the only guy
too absorbed in self-love
to notice
this phenomenon ?
ben_vulpes: did applied
thermofluids during
that
time. could feel
the slowdown hit.
mike_c: i
think you are unqualified in
that field.
the wrong job is pretty mind numbing, especially if you don't see
the path away from it.
mircea_popescu: list of
things slavegirls
think aforehand "kill
the mind" ? 559876985975 items.
ascii_field: high-paying job
that kills
the mind is a suicide.