Mike: "...The main reason AGI-ers invariably end up dealing with blocks
worlds, is because all their systems, logical, maths, geometry, and
computing, presuppose a world made of uniform elements – uniform blocks. –
uniform numerical, logical and formal/geometrical units.  And this
permeates the theorising of AGI-ers at every level. They have blocks
hammers so they presuppose the world is made of blocks nails..."

Todor:

Many are "stiff", that's right, but not all, and you're classifying wrongly
some of those that you're citing.

What physics proves is that world *is* made of "blocks", and what biology
proves is that your sensory system is built of "blocks" - finite amount of
receptors, and a finite amount of muscles, with finite precision. You just
have to have resolution  that is high enough (it has two dimension, I've
already told which) and you'll have the "endless variety".

*Let me generalize your profound confusions once again and discuss about
the pseudo-general intelligence of yours and of most of the humans...*

--- You don't understand levels of generalization and resolution. You see
everything as "different", no matter of resolution or abstraction (there's
no resolution in your mind).

As of the picture with the "chairs":
http://www.listbox.com/member/archive/post_content.html?post_id=20121023200141:FAEBD39A-1D6D-11E2-AC2A-F39E858D62D4&cid=1C2821765FE74D60B9EE0E5A63117AFB@MikePC

Well, these are not chairs, this is an image, where particular one-color
(black) spots can't be recognized on a background (it's a contrast,
adjacency/continuinity thing). You say that they are "chairs", this is the
top-down  part of naming things, you say "these are chairs".

1. "Chair" and anything that is generalized (words are generalizations,
except the names) has some "general/common" parts, features, details,
"substances" - that's the essential features upon which the word was
coined. (Recall my generalization of what a "chair" visually is.)

2. And... many inessential features, "decorations", details - that are
present in some samples and lack in others, which may or may not be
recorder as fixed samples, "endless variety". You may put anything in this
set, that's simple supervised labeling, not bottom-up generalization.

I've already discussed this issue - what you and the
top-down-and-bottom-up-generalization-deprived persons are doing is trying
to generalize un-generalizable taught (forced) labeled details, which
apparently lack enough common "basic" features in order to be "easily" and
unambiguously groupped and distinguished from other "things".

They can be distinguished only in forced labeled groups, "this elephant is
a stone, this toy-car is a stone, this octopus is a stone" (say, because
they all are gray) etc. - OK, if you're taught so, you may call them all
"stones", and your partner will know that you mean either of an elephant, a
car or an octopus, but given the word itself she won't be capable to
understand which one of them do you mean, neither she would be able to make
inferences about the qualities of any of the items, based on qualities of
any other, except their colour which is the same. The "physical" semantic
value for someone lacking that particular randomly selected set is close to
zero.

That's similar case to what small children often do - generalizing using
inessential features that just has taken their attention.

*So you fail to generalize those things, because it's impossible and they
are not general, and then you conclude that generalization is impossible,
there are no patterns, etc.*

The fact is that you don't understand what's the pattern, you see a bunch
of random details.

*--- Mike, you don't have a clue about maths, as others have pointed out...*

I've already tried telling, then gave up continuing, but in fact maths is
happily dealing with your "blobs" and "patchworks" for centuries, they are
called contours, curves, curvilinear integrals, integrals, functions;
generally your "endless variety forms" graphically are curves, parametric
curves and parametric surfaces, any of them and the families of them can be
defined mathematically and they are defined, and all of them can be grouped
on some mathematically measurable feature, and by altering the values of
the parameters all instances will be produced.

Such features are:

- closed or open
- number of inclusions (groups, subgroups))
- length of the curve (curvilinear integral)
- area or volume (double or triple integral). Is it continuous or has
interruptions, what are the slopes here or there, etc.
- relative length of the curve to the radius, to different radii, to area
etc.
- angles between segments having particular features of the above

Colours and their distribution can be treated the same way, differences in
the contrast are dealt with differential equations.

Besides, those on your pictures are simple functions, thus simply
recognizable, but you don't know what a function is.

*-- You talk about creativity, your friend Detusch talk about chess, how
creative a human player is, blah-blah...*

I suppose you both are artistically-deprived (if not disabled) - creativity
in arts is in fact the same like in chess, it's just the average people who
are creativity/artistically "disabled" in the part of producing it
themselves. They cannot understand how art is created, they don't
understand and don't manage the rules of those other "games" - those rules
go beyond their cognitive capacities. Music, drawing/painting, photography
(composition, light, contrast), dance, creative writing, sculpture, acting,
juggling...

It's all so trivial, and mathematically elementary. What do you people
don't understand about art, creativity and about the creation of pieces of
art or new inventions? Consult yourself with lectures, materials, papers,
and you should now understand all of it. If you fail, it's your brain fault
and your *pseudo-general intelligence. *

In fact , the reason why intelligence appears so "magical" in general lays
is in the pathetic* *cognitive capacities of the average humans.
Intelligence is trivial, but humans are not smart enough. By the way, the
general intelligence of humans is a myth, more specifically what's called
"general" is not that "general" and is not symmetrically general.

As I've discussed here also, human cognition is asymmetric, people are
largely passive recognizers, not creators/producers, and their failure to
understand and to produce super elementary things, in addition to their
associations with emotions makes those appear as extraordinary.

 One "generalist" here may claim that it's "a trade-off between precision
and generality", but I'll say that this is a bullshit, I challenge his
intelligence "generality".

What "precision" you need for drawing, or for playing a piano. It's
elementary perspective transformation, layers, and trajectories. In the
case of music - it's hitting super elementary sets of adjacent or
otherwise-related keys , making scales, chords, arpeggios, melodies. People
are amazed how does Beethoven composed music while deaf? Well, even if you
are deaf, and you are given a piano keyboard, what can you do with it?  Eat
it? Melt it? Cut it? No, you can press the keys at different velocity, hold
it for different length of time, at different coordinates, in different
successions, you can repeat those, you can use all of your fingers at once,
do it in different scales, combine, variate, transpose, change the
directions etc. That's the "magic" of musical composition and it's
mathematically *elementary. *

The great composers are just great mathematicians (in average humans scale)
who have conscious access to the musical data (pitches, times, ticks, their
relations) and much better working memory than their audience.

If physical motion is blamed, such as "fast coordinated motion requires a
lot of practice" - well, do it slowly, can you?... But no, average people,
including "generalists" can not deal with music or with drawing, or dance.
What's "so precise" here, what amount of cognition is needed for just a
bunch of a small set of possible motions done in particular way and
synchronization with music or other "clocks"?

So what's "general" about average human mind or your own mind's
intelligence, if you can't generalize and understand such obvious things in
their native domain? (Includes maths, soft sciences etc., you should deal
with ALL with the same mind, otherwise you're a *"pseudo-GI"*, or a *"partial
GI"*)

I will conclude: It's not that GI is complex, it's that average humans'
minds are not "complex" and general enough, the AGI will be intrinsically
more general than a human mind, it will be in fact more of a "general
intelligence" than "human level" (average human, poor "talents", or a few
specific talents).

So average people or ones who don't understand art or creativity may accept
the simplicity of it only when you slam a thinking machine that will
outperform them right into their faces. That's what any of us who want to
create real GI should do...


*---> Todor "Tosh" Arnaudov <---*
*
-- Twenkid Research:*  http://research.twenkid.com

-- *Self-Improving General Intelligence Conference*:
http://artificial-mind.blogspot.com/2012/07/news-sigi-2012-1-first-sigi-agi.html

*-- Todor Arnaudov's Researches Blog**: *http://artificial-mind.blogspot.com



-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968
Powered by Listbox: http://www.listbox.com

Reply via email to