27 April 2006
Pro-life priorities
human history
mapped in true perspective
would foreground the disconnect
between the will of the people
and the acts of the state
(which disconnect
has reached an alltime max)
but the microcosmic symbol
of this disconnect
that moves me most
is the universal sentimentality
over pregnancy and babies
weighed against the
insane cynicism
of pollution-for-profit
(the growing embryo
being the most vulnerable point
in the chemical ecosystem)
mapped in true perspective
would foreground the disconnect
between the will of the people
and the acts of the state
(which disconnect
has reached an alltime max)
but the microcosmic symbol
of this disconnect
that moves me most
is the universal sentimentality
over pregnancy and babies
weighed against the
insane cynicism
of pollution-for-profit
(the growing embryo
being the most vulnerable point
in the chemical ecosystem)
19 April 2006
Heraldic barcodes and the semantic web
my phrase 'robot wisdom'
was always meant to imply
science-logic ('robot')
tempered by art-intuition ('wisdom')
asserting that the social
sciences
fail
because they fail to meet
novelists' standards
of realism
and now
the first ontologies
for the semantic web
fail too
because they lack
poetry
which is intuition's metric
for psychological realism
but here's an antidote
or at least a try
at meeting them halfway:
re-format your
ontologies, for classifying
photos
music
news
to show them graphically
iconically
heraldically
as human-readable 'barcodes'
that put first things first
human priorities before
machine priorities
for music, foregrounding
the genre
the beat
the stress/soothingness
for pix
the prettiness
the subject
the degree of artifice
(eg photoshop)
for news stories
the place
the category
(disaster, biz, bizarre)
the interestingness
heraldic representations
putting 1st things 1st
but leaving room for tweaks
and judging
first and last
by their aptness
for the eye
was always meant to imply
science-logic ('robot')
tempered by art-intuition ('wisdom')
asserting that the social
sciences
fail
because they fail to meet
novelists' standards
of realism
and now
the first ontologies
for the semantic web
fail too
because they lack
poetry
which is intuition's metric
for psychological realism
but here's an antidote
or at least a try
at meeting them halfway:
re-format your
ontologies, for classifying
photos
music
news
to show them graphically
iconically
heraldically
as human-readable 'barcodes'
that put first things first
human priorities before
machine priorities
for music, foregrounding
the genre
the beat
the stress/soothingness
for pix
the prettiness
the subject
the degree of artifice
(eg photoshop)
for news stories
the place
the category
(disaster, biz, bizarre)
the interestingness
heraldic representations
putting 1st things 1st
but leaving room for tweaks
and judging
first and last
by their aptness
for the eye
17 April 2006
Zipless blogging on the Tree of Life
data on hard drives
reflects the lifestories
of those who put it there
and anticipates the lifestories
of those who may retrieve it
and the Web is
a network of hard drives
holding data
chunked as webpages
the browser an interface
for exploring it
so the surfer, seeking completion
of a story-goal
contemplates a webpage
extracting clues
about which clicks
should yield which goals
and when successful
may seek to bookmark the page
or to blog it for others
using the minimal number
of mouseclicks and keystrokes
reflects the lifestories
of those who put it there
and anticipates the lifestories
of those who may retrieve it
and the Web is
a network of hard drives
holding data
chunked as webpages
the browser an interface
for exploring it
so the surfer, seeking completion
of a story-goal
contemplates a webpage
extracting clues
about which clicks
should yield which goals
and when successful
may seek to bookmark the page
or to blog it for others
using the minimal number
of mouseclicks and keystrokes
16 April 2006
Webjay, flickr, and the zipless blog
mining flickr
for good pix
is an awful lot like
mining webjay
for good tunes
above each flickr pic are
icons for 'fave' and 'blog'
and 'fave' is quicker
but the only available view
of a person's faves
is 36 tiny thumbnails per page
no rss
no fullpage with 'next' button
(there's the flash slideshow option
where you can pause and mark faves
but it's clumsy for this)
so i opted to start a new blog
and use the 'blog' icon
which demands
an extra step at the start
to choose your blog from a menu
and at the end
to go back to the picpage
and you're stuck with squinting
at the tiny thumbs
and opening each one
that shows promise
and because flickr's pages
are loaded with formatting
they're slow to display
all of which is a drag
on 'zipless blogging'
but even with squinting
the 'faves' pages are useful
in exactly the way
Webjay playlists are useful
you find someone whose
tastes look similar
as a startingpoint for exploration
(you can even pick a pic
or tune you like
and see who else 'favorited' it)
with Webjay playlists
you can send the whole list
to iTunes or Winamp
and check the ones you like
skip the ones you don't
and when Webjay fixes the bugs
that garbage your playlist
you'll be easily able to
aggregate your faves
and while giving each tune
a fair shake
is a lot slower than
scanning a page of thumbs
it would be quite comparable
to browsing with a 'next' button
if flickr offered that view
for others' faves
for good pix
is an awful lot like
mining webjay
for good tunes
above each flickr pic are
icons for 'fave' and 'blog'
and 'fave' is quicker
but the only available view
of a person's faves
is 36 tiny thumbnails per page
no rss
no fullpage with 'next' button
(there's the flash slideshow option
where you can pause and mark faves
but it's clumsy for this)
so i opted to start a new blog
and use the 'blog' icon
which demands
an extra step at the start
to choose your blog from a menu
and at the end
to go back to the picpage
and you're stuck with squinting
at the tiny thumbs
and opening each one
that shows promise
and because flickr's pages
are loaded with formatting
they're slow to display
all of which is a drag
on 'zipless blogging'
but even with squinting
the 'faves' pages are useful
in exactly the way
Webjay playlists are useful
you find someone whose
tastes look similar
as a startingpoint for exploration
(you can even pick a pic
or tune you like
and see who else 'favorited' it)
with Webjay playlists
you can send the whole list
to iTunes or Winamp
and check the ones you like
skip the ones you don't
and when Webjay fixes the bugs
that garbage your playlist
you'll be easily able to
aggregate your faves
and while giving each tune
a fair shake
is a lot slower than
scanning a page of thumbs
it would be quite comparable
to browsing with a 'next' button
if flickr offered that view
for others' faves
13 April 2006
Compassion on the Tree of Life
for every possible story
on a person's radial ontology
consider the cost of
making that story come true
as well as the cost of
not making it come true
and the higher each cost
the brighter that story shines
but also
costs that can be deferred
we'll light in blue
while undeferrable costs
will be red
so when bright red
costs of INaction
outshine bright red
costs of action
it's time to get movin'
but when costs of inaction
are just bright blue
the red costs of action
may keep you down
when we get to know another
we begin to learn these colors
on their ontology
perhaps we can even suggest
actions they haven't thought thru
but we have to care
to empathize with their costs
and the socialist theory of government
is that where government
can efficiently reduce
the majority's costs
it should attempt that
on a person's radial ontology
consider the cost of
making that story come true
as well as the cost of
not making it come true
and the higher each cost
the brighter that story shines
but also
costs that can be deferred
we'll light in blue
while undeferrable costs
will be red
so when bright red
costs of INaction
outshine bright red
costs of action
it's time to get movin'
but when costs of inaction
are just bright blue
the red costs of action
may keep you down
when we get to know another
we begin to learn these colors
on their ontology
perhaps we can even suggest
actions they haven't thought thru
but we have to care
to empathize with their costs
and the socialist theory of government
is that where government
can efficiently reduce
the majority's costs
it should attempt that
I'm an idiot about... mp3 players
i read dealnews regularly
waiting, among other things
for an mp3 player to drop down
to my price bracket
and so far
the cheapest by far
are the ones with no memory
that expect you to add
a memory card
but usb thumb drives
are cheap
so why can't a cheap
mp3 player module
plug into a cheap
thumb drive?
waiting, among other things
for an mp3 player to drop down
to my price bracket
and so far
the cheapest by far
are the ones with no memory
that expect you to add
a memory card
but usb thumb drives
are cheap
so why can't a cheap
mp3 player module
plug into a cheap
thumb drive?
12 April 2006
Stories in interface-AI
i like the effect
of quoting poems in my blog
raising the tone
but i don't run across
enough hightoned quotables
so i thought i'd try
random joycequotes
eg yesterday:
"Pause. If we were all suddenly somebody else. Far away a donkey brayed." (randomJAJ)
but this collapsed two paragraph breaks:
"Pause. / If we were all suddenly somebody else. / Far away a donkey brayed." (randomJAJ)
and on rereading it this morning
i realised the slashes
(or something) were needed
and added them
now this simple edit
is semantically byzantine:
should the rss feed signal a change?
should the rss item be redated from
created-time to updated-time?
suppose i want to discuss the change
(as i'm doing now)
shouldn't i
ideally
be able to link
before and after views
eg waybacked at archive.com?
but the
story
this edit tells:
person writes text
person dislikes poetry of re-read text
person edits formatting
is a special case of:
person writes text
person dislikes X about re-read text
person edits X
which is one of the
most basic 'stories'
in general wordprocessing
and only becomes controversial
complex
because a publication-step
was added
between error and edit
semi-live publication
being characteristic
with blogs
of quoting poems in my blog
raising the tone
but i don't run across
enough hightoned quotables
so i thought i'd try
random joycequotes
eg yesterday:
"Pause. If we were all suddenly somebody else. Far away a donkey brayed." (randomJAJ)
but this collapsed two paragraph breaks:
"Pause. / If we were all suddenly somebody else. / Far away a donkey brayed." (randomJAJ)
and on rereading it this morning
i realised the slashes
(or something) were needed
and added them
now this simple edit
is semantically byzantine:
should the rss feed signal a change?
should the rss item be redated from
created-time to updated-time?
suppose i want to discuss the change
(as i'm doing now)
shouldn't i
ideally
be able to link
before and after views
eg waybacked at archive.com?
but the
story
this edit tells:
person writes text
person dislikes poetry of re-read text
person edits formatting
is a special case of:
person writes text
person dislikes X about re-read text
person edits X
which is one of the
most basic 'stories'
in general wordprocessing
and only becomes controversial
complex
because a publication-step
was added
between error and edit
semi-live publication
being characteristic
with blogs
11 April 2006
MemeTracker
i subscribe, by rss
to a handful of
del.icio.us tags
like 'googlebase'
that very occasionally
deliver something useful
but mostly just
repeat and repeat
as multiple posters
rediscover the same
page/meme
and similarly, i consider
how memes recur in waves
passed around
forgotten
rediscovered
passed around again
so imagine
what you'd have to add
to delicious
or to technorati, etc
to track this
a database
not just of urls, but of
identical, or just comparable
content-items
that may appear under
many different urls
that may trace an evolving story
that will have a meta-history
of who linked when
and will have tags
that may evolve, themselves
as the meme's appeal evolves
(eg new parody reviving interest
in old original inspiration)
so you can do a background check
on a meme that's new to you
and easily see where it's been
and how to frame it best...
to a handful of
del.icio.us tags
like 'googlebase'
that very occasionally
deliver something useful
but mostly just
repeat and repeat
as multiple posters
rediscover the same
page/meme
and similarly, i consider
how memes recur in waves
passed around
forgotten
rediscovered
passed around again
so imagine
what you'd have to add
to delicious
or to technorati, etc
to track this
a database
not just of urls, but of
identical, or just comparable
content-items
that may appear under
many different urls
that may trace an evolving story
that will have a meta-history
of who linked when
and will have tags
that may evolve, themselves
as the meme's appeal evolves
(eg new parody reviving interest
in old original inspiration)
so you can do a background check
on a meme that's new to you
and easily see where it's been
and how to frame it best...
07 April 2006
Human factors on the Tree of Life
by substituting radial
concentric circles
for a triangular hierarchy
in imagining my universal ontology
i gain the significant dimension of
distance from the center
which i mapped first
onto how common, in the real world
a given story-concept was
and second onto how
simple or complex
but now i want to re-map
that significant dimension
onto how much cost
or how much benefit
the realization of a story
would cost or benefit
the story-realizer
so in the 'benefits view'
stories with small benefit
will lie closer to the center
and stories with great benefit
farther out
and in the costs-view
small costs are close
and great costs farther out
and planning
means finding cost-stories
that lead to greater-benefit stories
now when a designer refines a tool
the goal is to maximize benefits
and minimize costs
and in software tools
the costs are mostly linear
sequences of keyclicks and mouseclicks
and when good design
makes a benefit cheaply accessible
for the first time
via a series of cost-clicks
below some unknown threshold
you have a killer app
a visicalc
a mosaic
an itunes
concentric circles
for a triangular hierarchy
in imagining my universal ontology
i gain the significant dimension of
distance from the center
which i mapped first
onto how common, in the real world
a given story-concept was
and second onto how
simple or complex
but now i want to re-map
that significant dimension
onto how much cost
or how much benefit
the realization of a story
would cost or benefit
the story-realizer
so in the 'benefits view'
stories with small benefit
will lie closer to the center
and stories with great benefit
farther out
and in the costs-view
small costs are close
and great costs farther out
and planning
means finding cost-stories
that lead to greater-benefit stories
now when a designer refines a tool
the goal is to maximize benefits
and minimize costs
and in software tools
the costs are mostly linear
sequences of keyclicks and mouseclicks
and when good design
makes a benefit cheaply accessible
for the first time
via a series of cost-clicks
below some unknown threshold
you have a killer app
a visicalc
a mosaic
an itunes
06 April 2006
Analysis and ontology
i'd like to elevate
the humble html tag
<ul>
(unordered list)
to a post of honor
in the philosophical realm
where i'll re-dub it
'analysis'
after a rule of thumb
that emerged as i compiled
my timeline of knowledge representation
where i found that
any domain of knowledge
could be advanced
by proposing an
unordered list
of components of that domain
ie, an analysis
which lists i tried to quote
as illustrations
for the timeline
so a hierarchy is built of
nested
analyses
and an ontology is a hierarchy
that tries to span a domain
but all universal ontologies
are failures
because complex composites
can't follow
a simple hierarchical analysis
the humble html tag
<ul>
(unordered list)
to a post of honor
in the philosophical realm
where i'll re-dub it
'analysis'
after a rule of thumb
that emerged as i compiled
my timeline of knowledge representation
where i found that
any domain of knowledge
could be advanced
by proposing an
unordered list
of components of that domain
ie, an analysis
which lists i tried to quote
as illustrations
for the timeline
so a hierarchy is built of
nested
analyses
and an ontology is a hierarchy
that tries to span a domain
but all universal ontologies
are failures
because complex composites
can't follow
a simple hierarchical analysis
02 April 2006
Musical barcode?
suppose you wanted
to scan, by eye
100 songs you'd never heard
looking for ones you might like
so you'd need a way to represent
their Pandora-style 'musical DNA'
as patterns of color
like: a rectangle
that's longer if the song is longer
that has a pink stripe
where there's female vocals
blue for male
lightcolored for sweetvocals
dark for harsh
closespaced vertical stripes
for fast-beat parts
wide for slow
subtle rhythms
distinguished, somehow
from ordinary ones
some kind of repeating symbol
for looping sequences
something messy for glitchy
primary colors for ordinary harmonies
bluesy colors for bluesy harmonies
subtle for subtle...?
to scan, by eye
100 songs you'd never heard
looking for ones you might like
so you'd need a way to represent
their Pandora-style 'musical DNA'
as patterns of color
like: a rectangle
that's longer if the song is longer
that has a pink stripe
where there's female vocals
blue for male
lightcolored for sweetvocals
dark for harsh
closespaced vertical stripes
for fast-beat parts
wide for slow
subtle rhythms
distinguished, somehow
from ordinary ones
some kind of repeating symbol
for looping sequences
something messy for glitchy
primary colors for ordinary harmonies
bluesy colors for bluesy harmonies
subtle for subtle...?
Transparent things on the Tree of Life
Nabokov:
a human branch eyes an object
at one thin layermoment
an imaginary line drawn
between eye and object
and the human's radial ontology
lighting up
with stories about that object
stories it's incarnating
in the present
or may incarnate
in the future
or may have incarnated
in the past
(whence N warns)
"When we concentrate on a
material object, whatever its situation,
the very act of attention may lead
to our involuntarily sinking
into the history of that object."
a human branch eyes an object
at one thin layermoment
an imaginary line drawn
between eye and object
and the human's radial ontology
lighting up
with stories about that object
stories it's incarnating
in the present
or may incarnate
in the future
or may have incarnated
in the past
(whence N warns)
Subscribe to:
Comments (Atom)