Semantic Restructuring is the pursuit of enlightenment, enlivenment, empowerment through the creative re-arranging of the building blocks of meaning. For a better description, Start Here.
| Sun | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
| 1 | 2 | 3 | ||||
| 4 | 5 | 6 | 7 | 8 | 9 | 10 |
| 11 | 12 | 13 | 14 | 15 | 16 | 17 |
| 18 | 19 | 20 | 21 | 22 | 23 | 24 |
| 25 | 26 | 27 | 28 | 29 | 30 | 31 |
| Months | ||
|---|---|---|
| Jan | Feb | Mar |
| Apr | May | Jun |
| Jul | Aug | Sep |
| Oct | Nov | Dec |
Bateson, books, cogling, context, CPB, embodiment, framing, I Ching, paradox, perception influence, prisdem, semantic punctuation, sensation, techniques, unconscious
From today's spam:
matter at the wrong end of its body it would have been beautiful thus to in that all the atoms of Democritus large and small alike were
Which I quickly posted to Cafe Blue as the following found poem (with ugly tags a-showin')
<blockquote> <pre> matter at the wrong end of its body it would have been beautiful thus to in that all the atoms of Democritus large and small alike were </pre> </blockquote>
Makes me think it's the spammers who first will pass the Turing test, at least if the measure becomes blank verse rather than teen chat. In fact, I'd aruge the spammers have, in fact, succeeded in passing a version of the test, each time someone deems it neccessary to write a new program to filter spam because the spammers succeeded at defeating the old one, since one of the things a tough spam filter does is tries, by the fewest steps neccessary, to establish whether or not the content is human generated or machine generated. Where the gaming world drives processor and video card innovation, it is the spammers and those who crusade against them who are doing the truly profound work in artificial intelligence as expressed in language production.
There's some real question begging that goes on with regards to the Turing test. An apochryphal story goes that the creator of Eliza, a dim relative of the chomsky-bot, indeed fooled a fellow researcher such that Eliza's programmer received a nasty call in middle of the night complaining about his "recalcitrant research assistant." (I'm having a little trouble finding a link to that just now...) But in such a case the computer isn't credited with a win on the Turing test, rather the human is credited with a failure. He *should* have been able to tell. But that leads to the circular nature of how we've defined intellgence in this context and the value of truly blind studies. It's one thing to try to detect that an interlocuter is a computer program when you're told up front only one of the two chatters you're on with is human. But what if there were no such warnings?
Of course the net is a wonderful medium for exploring this kind of stuff, because it has so many variants that are text only. The task of writing a program to handle intonation and facial expression as well as word choice...well, luckily most artificial intelligence folks (ahem) know that such matters aren't part of intelligence at all, and are only barely important to communication. (Yes, these are the same guys sportin' wood over xml because it allows endlessly tweakable semantic tagging; go figure.) Chat is the ideal place to play Turing, because it's got that realtime flavor and it's strictly textual.
Still, what we have is an ethnocentric definition of intelligence being foisted off on a machine designed to model the human mind while in fact being nothing at all like the human mind (unless you're of the opinion that the mind operates by adding enormous numbers of binary figures, in which case you might look up Luria, Pietsch, and Pribram.) We define the target based on our prejudices, then tell the tester to try and cause the program to fail. Let's not point out that pre-verbal infants would fail such a test. Even an Einstein, a Gates, a Bush.
If you're wondering, "Why'd he use use the bare pre and blockquote tags above?", well, they're human readable, ain't they? Oh, and here's a quick "gimme" link to chomsky-bot, and one more, just in case you're too lazy to click through to the fascinating "How does it work" link (By the "American Chinses Menu" principle, viz. One from Column A, One from Column B, etc.)
[]
static link
writebacks: 0 (writeback = trackback +/- comment)