I am sincerely amazed how the Spam People are smart and inventive, and it makes me want to read all the spam I receive all day. I think I already mentioned this witty blog  that aggregates and cuts up spam on an every day basis in my recent "Cheap Art Text" blog entry : Spam Today! (http://spamtoday.blogspot.com/).

I have been receiving for a while these messages which subjects are either "Waiting for news, my friend!" or the simple "[Re]" without subject. Today I opened one of the latter kind. I have to mention that I am on Gmail, so my email list page shows the start of the content message. Here is how this one goes:

The body message of "Vicky Austin/ Re[20]:" seems to be "As long as they played fair, you can't miss it..." etc. like a normal, non psychotic email conversation. Until now, I knew it was a spam trick, but had never opened this kind of messages. When I opened it, I was surprised to not see these words at all but an image advertisement for Rolex: this is no surprise, but the words first displayed as the body message weren't in the body at all!! Well, where did these words go?? I am really interested to learn what kind of trick code they used to hide them.

This is my last spam (from a great friend that writes me every day: The Ultimate Online Pharmaceutical) from today, which is literally an evolutive piece of history of poetry - from Post-Modernist to code poetry with a very touching Romantic haiku in the end.

Vlieagra - $3.3
Levitrra - $3.3
Cialsis - $3.7
Imitrhex - $16.4
Frlomax - $2.2
Ultrram - $0.78
Viowxx - $4.75
Amrblem - $2.2
VaIiaum - $0.97
Xanmax - $1.09
Sowma - $3
Meriwdia - $2.2


visit our website


Best regards,
Online Pharmaceuticals

hjuyerhgw UlNeXVldVx1EV3FVXlVcXRxQW1g=


A black plum is as sweet as a white.
After clouds a clear sun.
Thinking is not knowing

----------------------------------------------------------- 

 And this one: visual poetry! even more! conceptual! I love it! I lmost want to invite Elsa to Gmail!

The people from ELGG really don't like spam. Here is what one of the administrators says about spam prevention: "Elgg now has two prongs to its attack. The first is a master blacklist of terms people cannot include in their comments; for Elgg.net we've set this to include things like "-casino" and "v1agra". People trying to post those to the site will receive an error message. The terms on this list are actually regular expressions, so we can match more complicated terms if necessary." (http://elgg.net/bwerdmuller/weblog/5151.html).

Watch out -gamblers and elderly p30pl3 (3$p3c14lly those who use l33t)! No commenting on Elgg for you!

Here are the tags that relate to spam in the ELGG community, through semantic linking:

Automatic tag suggestion:

spam-poetry
comment spam
Spam & malignant hackers
spam; dishonest people

You can see that thanks to my previous contributions in "Spam Poetry" this tags are illuminated by the pleasure of art. Also the people who dislike spam on Elgg are these:

You see there are not that many and there is still hope for a spam revolution! 

My friend Chris and I are telepathically connected. A few days ago he blogged on his Got Hacked! weblog about his interest about his own spam poetry. He is the proud owner of some very heart breaking romantic verses, by the famous “XP PRO, OFFICE 2003 AND ALL AT ONLY $12-60 EACH, WE GIVE U LICENSE next”.

Go check it out! http://gothacked.org/rmbblog/?p=12

  (spam fan, not Chris)

So my friend jUStin was nice enough to host my poems lost in space after the crash of my website on the Univ. of Buffalo server. Many thanks. Find them here on the plantarchy website (the site for his magazine): http://www.plantarchy.us/campom/campom_page.html

About  jUStin katKo: "extraordinaire young poet from Ohio" said Loss Pequeno Glazier, whose performance I video-ed here: http://www.vimeo.com/clip=51346. To sum him up, I'll cut & paste some deleuzian stance that he sent me the other day and that I find more than relevant: [T]here is no language in itself, nor are there any linguistic universals, only a throng of dialects, patois, slangs, and specialized languages...Language is, in Weinrich's words, 'an essentially heterogenous reality.' There is no mother tongue, only a power takeover by a dominant language within a political multiplicity.

(from the "Blue Book" Scheme, that is like an horoscope because it predicts the future he told me).

Really active review-maker, publisher and thinker at Plantarchy.us http://www.plantarchy.us.

Performer, sound/video/other stuff poet on his website at http://www.users.muohio.edu/katkojn/index2.htm

Here is jUStin with a cape in the video Drone (oh no i couldn't find the video on the website!):

 
I am writing a little post for the IDC blog about Power of Words (http://elgg.net/camillepb/weblog/8419.html) and a similar artist's project (by Christophe Bruno - Google Art) involving  data classification and speech analysis. To approach these matters I feel it's necessary to know a little bit about Semantic Web. What follows is the little I know.

 
-------------------

Before all I want to link here this course website by my blog teacher (Trebor if you're reading this, I am advertising you - to what audience? I don't know), that I wish I could have taken: Cultural Theory and Database art (http://molodiez.org/dms420/course3.htm)

-------------------

Semantic web is connected to the emergence of Web 2.0, a new generation of information tools implemented in famous social software as Flickr and del.ici.ous. But these tools are not only means of classification, they are means of analysis as well. We've come a long way since the link was considered as the most meaningful tool for organizing data and structure the meaning of text.

Theodor H. Nelson (in his book about Xanadu: Literary Machines) elected the link as the primary capability of the docuplex: from hyperfile (as a simple storage system based on linking) to "thinkertoy" (a computer screen facility that allows the detailed intercomparison of complex alternatives), the link is processed in many ways:
- through time (evolution)
- through space (tree structure)
- through concept: different models of links

The link is part of the document and of the writing, it participates to the definition of text as an "evolving, Protean structure" (Nelson, 2/17). But the link remains an abstraction of an individual choice (the author figure proposes a pathway through the document). Whether it expresses subjectivity or the vested interest of one group, the link works on a singular mode: you-click-you-go. It is based on an attribution: someone/something attributes a path to a document: to some extent, it is a form of authorization, the same kind of second-hand authorization that a quote has. Also it involves the same sort of definition of authority that post-modernism (through cut-up and reappropriation techniques) has experimented.

Here I can't resist posting one of Nelson's book's nice graphs, perhaps the nicest one. His graphs are deliciously useless. It tells a lot in my opinion about the abstraction of the link. Links cannot be visualized; even the tree structure, which is commonly used for representing the principle of linking, is not totally adequate because how do you limit the branching? Nelson's trees are completely crazy - I'll post one soon when I take more pics of Literary Machine's drawings (by the way this book is rad everybody with a slight interest for the internet should read it).

Google's search engine is still based (though it is much more complex I imagine) on the idea of linking that follows Nelson's stance: "Follow Any Link Wherever it May Go". Such program, based on contingency, has two consequences:
- free search, pleasure or boredom of being lost, surrealistic textualities or shock of meanings (cf. Googlism)
- oriented search, bought by companies ("Related Search", "Sponsored Links") or determined by clever internet guys (bomb-linking? how do you call that again?)

These type of tools can be used in a critical way via an act of extraction that has to do with a logic of spontaneity, an internal code that borrows the functions of algorithm but not in a systematic/programmative way (see Barthes about that, but I have to look for the reference), as a friend pointed out to me in a recent mail.

The Semantic Web project(s) (cf. wikipedia entry: http://en.wikipedia.org/wiki/Semantic_web).  The research in data-semantics tries to concretize and systematize the relationship to data, and to think about these paths in terms of "roads" of information. The approach is more concerned with the group as community rather than the individual itself or the corporate group ( = 1 body: a means to an end). What is interesting is that idea of community only requires sharing as a praxis, not the sharing of something in particular (same values/objects, etc.). Subjectivity is taken into account, but only to the extent that it meets other subjectivities.

The tag is one of the tools that replaces the link: it uses the link technics as well, but channels its arbitrary through statistical analysis (could I say that it used the logics of deduction instead of induction? I don't know if that's appropriate? I was thinking about something contrary to inference). Cf. my entry about del.icio.us and flickr tags: http://elgg.net/camillepb/weblog/8419.html

In the perspective of Steve Cayzer, who is developing a semantic document method of classification and implemented it in what he calls "Semantic blogging", it is all about extending the web thanks to a system of "intelligent annotations". He uses a selection process that has a lot to do with the cloning phenomenon in biological cell systems (artificial immune systems).  A distinction is made:
- classification with textual features: "Successful classification of the documents relies on attributes of the document, also known as features (...) Once the important features have been selected, then the data must be represented in a way that is suitable for use by the classifier system".
- classification with semantic features: "Taxonomies describe a structured way in which various information and classifications can be viewed, with each branch of the taxonomic tree having some relationship with the parent and siblings. They are a simple version of an ontology, which has been described as a 'formal specification of a shared conceptualization'. That is, ontologies are concerned with methods for knowledge representation, and taxonomies are one such method"

I guess the keyword here is "representation": in terms of criticism, artists use that power of representation that statitistics has (the science of diagrams...) to analyze data.

The textual features are the first extension of the ontology of the link (the link is part of the definition of the text according to T. H. Nelson): the nature of the document is grasped through a "Find"-type search (extraction of words and expressions).

The Semantic features jump over this and create their own ontology: they create a position, a coordinate that is flexible enough to change when data changes. A link can be broken; on the contrary a flickr popular tag grows as more photos include the semantic feature, and if for some reason all the users that used to tag their photos with the word "blah" decide to delete these photos, the tag disappears. In that sense it is a lot more organic than a link. This classification is based on the principle of a personalized, user-friendly method that allows a multiple ranging and sorting of data : it is also called "multiple mapping".

Steve Cayzer's links:
- "An artificial Immune System Approach to Semantic Document Classification" (pdf doc): http://www.hpl.hp.com/personal/steve_cayzer/papers/030901Sema
- "Semantic Blogging and Decentralized Knowledge Management": "Tapping into the structured metadata in snippets of information gives communities of interest effective access to their collective knowledge" http://portal.acm.org/citation.cfm?doid=1035134.1035164
- Semantic Blogging Demonstrator; http://www.semanticblogging.org/semblog/blog/default/

ha ha ha ha

(I posted this on the IDC blog earlier tonight, but why not post it here too?. Also, this post was based on this other one: http://elgg.net/camillepb/weblog/8419.html)

Research in semantic web and in text processing provides some very useful tools of classification that happen to become tools of criticism, as recent new media works have demonstrated recently.

This encounter of semantics and statistical analysis is already a fruitful method implemented in social softwares such as del.icio.us or Flickr, under the form of tags.

Tags are an alternative to traditional data classification as a real-time visualizer of how language is used in communities through keywords: a keyword (and the ideas and objects behind it) has a reality only to the extent that it is picked up frequently. The more people choose a keyword to tag their pictures or their bookmark, the more likely the tag will appear on a "popular tag" page. From there, it will be picked up more frequently due to exposure; thus it will grow bigger among other tags.

The access to information in these social softwares is determined by trends and rhetoric. It is always a surprise for me to find out that a picture of my cat playing with my swimming suit and that I tagged "bikini", as been visited 10 times more than any other of my exquisite pictures. See del.icio.us most popular tags and Flicker's tag page.

Two important blogs have recently accounted for such experiments in semantics and statistics, but with a critical perspective, with an application to the rules of web searching, or to the rhetorics of political speeches.

On the Information Aesthetics blog, the visualizer project called "The Power of Words, a Text Analysis of Political Discourse During Times of Crisis" by MMEDIA: a textual analysis visualization of keywords mentioned during famous speeches (ranging from G.W.Bush to W. Churchill). The visual display breaks down the rhetoric, takes the words out of context, & treats them at face value in order to analyze the breakdown of content. Each group of metaphors (e.g. decline, controversy, war, imagination) is color-coded, & sized based on frequency.

On We-Make-Money-Not-Art, an interview of French net-artist Christophe Bruno who defines his work as "diverting global symbolic structures like Google search engine or the blogosphere [using] language as a medium". His has gained recognition for the Ad-Word Happening project (rewarded at the 2003 Ars Electronic festival). In this interview he asks "What is speech at the "age of access", at the age of globalization and "taylorization of discourse"?". Here are a few quotes:

"It's clear to me that the history of Internet goes from utopia to dystopia. It started with a hope about sharing ideas, sharing media, free speech etc. and ended (well it's not over of course) with the commodification of language. On the political point of view, there is a clear will, from any power, to lock some of the libertarian aspects of the Internet. But what interests me are the economic dynamics that are in play here and how they interact with the political and social context. For instance, the relation of Google to free speech is very interesting: in fact free speech is the precondition for them to track and analyse the intimacy of the millions of bloggers. In other words Google's ethics of free speech (although they sacrificed this rule in China) lies in its economic dynamics, they need it to optimise their adwords/adsense system. (...)  I called this loop mechanism between control and spectacle, the "Taylorisation of speech". (...)

Google.art, let's call it like that, brings its own questions about the promethean myth of the separation between man and machine, which is nothing but the long term question: what is the Subject of speech?"

Read the whole interview here and find the AdWords Happenings there.

Christophe Bruno AdWords Happenings plays with the rules of  the Sponsored Links service proposed by Google. He wrote little " spam poems" in the ad boxes that appear selectively to the user according to his personal search.

Clicking on these links would of course redirect the user to Bruno's website. Bruno then collected enough data to draw tables rendering the values of a number of keywords: their price relatively to their use (you click, he pays). After being rebuked by Google for not playing the game of advertisement, Bruno was then able to figure some of the rules of what he calls a "generalized semantic capitalism".

This is interesting in the light of Theodor Adorno's definition of the jargon: "The jargon has as its disposal a modest number of words which are received as promptly as signals" (The Jargon of Authenticity). The idea of signal points out to this other idea that a message is encoded. This encoding can be seen as the digital form of jargon: the rules of encoding do promote something, but it is not see-through, thus it can easily become a medium of control (cf. Benjamin's critique of the use of cinema and radio under authoritarian regimes). Adorno was criticizing the social values and political implications behind the jargon, and Christophe Bruno, in the era of spam, is interested in the economic values:

"One of the most interesting fact is that we have reached a situation in which any word of any language has its price, fluctuating according to the laws of the market."

These experiments play, to some extent, with the idea of early wittgensteinian idea of language as image - a proposition shows the structure of what it is stating, the factuality of it. But in these cases the artists deal with a language that has been devoided of factuality (let alone of reality): what they state is just an empty structure, a mere rhetoric. In these two works, the rhetoric hides a very complex structures of vested interests that gain efficiency through a visual manipulation of words. The text dimension of the Internet becomes literally a space in which you find your path, or where you are forced into routes, through smart visual manipulations.

Both Bruno and MMEDIA work on unraveling these hidden structures.They seem to take the idea of fact, by interpreting factuality in discourse, as an ideological event. The notion of event is indeed highly ideological in the sense that it is based on social conventions and political or economical rules. The Wikipedia entry for "Event" sums it up: "A significant occurrence or happening", i.e. an arbitrary meaning meshed into language and action. Our textual navigation is determined by these pseudo-events, thus shaping our habits.

Just giving more information about this artist who coined (I think) the great word GOOGLE ART. It fuels my passion for spam poetry!

The AdWords project mentioned in my previous post is the most famous one, but there are other projects related to GOOGLE ART. Here are 2 of them. I will come back on him because he's cool and smart.

"Non-Weddings" (http://www.unbehagen.com/non-weddings/)

When you click on "Celebrate a non-wedding", a request is sent on the internet that brings back pictures randomly taken among those related to the chosen names (or words). You can of course type the names or the words you want and celebrate their non-wedding.

Non-weddings was the first artistic Google Hack for Image Search.

Here is a non-wedding for my blog teacher Trebor Scholz and his partner Jennie (I think that's her name):

 "B.L.O.O.D. F.O.R. S.A.L.E" (http://www.iterature.com/bloodforsale/):

"Blood for Sale" is a wireless adaptation of my very first net.art piece, "epiphanies" (inspired by James Joyce's definition of the epiphany). It features the pervasive invasion of language by financial globalization. As James Joyce did 100 years ago, I walk through the city of Dublin, but with a Wi-Fi PDA, recording encountered advertisements of company logos or brands into an administration interface via the wireless network. These inputs are sent to a program on a server that uses search engines (Google etc.) to fetch sentences related to the input from the web. These sentences are known as "sponsored epiphanies".

Today I listened to a conference by Katherine K. Hayles held at HUMlab in September 2005,  downloadable here: http://blog.humlab.umu.se/archives/20050921_hayles.mp3

HUMlab is a humanities IT environment at Umeå University in Sweden, and has an interesting blog that is mostly centered on the Web 2.0 techniques, and which also is a platform for all what they propose in terms of streaming, e tc. http://blog.humlab.umu.se/

 
Katherine Hayles: here is her homepage: http://www.english.ucla.edu/faculty/hayles/. She has a lot of things on the internet so googling her would lead to many audio conferences and texts.

Here are a few non-critical, non-exhaustive and non-documented notes on the conference, which was based on the first chapter of then to-be-realeased My Mother Was a Computer. I think these notes might be important to understand text generation processes.

Hayles expresses her main concern, related to what she calls "the computational regime". Her perspective on the subject starts with the question : "What is computation?". At a basic level, the act of computation is what happens in power to the desktop computer. The desktop computer itself is an instantiation of computational processes, but the idea of computation is much broader than this, it isn't limited to the computer: it can manifest itself within a wide range of media. All it requires is a small set of elements: a code (0/1), and, in the case of the digital computer, this set is composed of a small number of logical operations: the criteria seems minimalist but actually enables to to a lot of things.

Hayles recalls Steve Wolfram who was working on computation with paper and pencils (cf. the book A New Science). He was trying to model the process of "cellular automata". The initial state of a cell is binary (on/off). But its evolutionary behavior depends on how the neighbors cell are (on/off) too. Depending on that, a simple computation will generate complex patterns.
Patterns are spontaneous and emergent phenomena that come out of the cells' interaction with one another. A pattern is not predictable, there is nothing in the algorithm. You have to run the program (in Wolfram's case, the simulation), in order to actually see what happens.
Wolfram came up with a rule he called 1.1.0.: it is a principle of computational equivalence: "any behaviour not obviously simple (i.e. any complex behaviour) in any system, can be thought of as corresponding to a computation of equivalent sophistication"
From there, Hayles goes on making 3 interlocking claims:
1/ all complex behaviour can be simulated computationally
2/ simulations are computationnally irreducible (to an equation for example). One has to run the simulation: it is informationally resistant to any further compaction or condensation. If you want to predict future behaviors, a formula is required: this is precisely what lacks in computation.
3/ these behaviors are not only simulated but also generated by computational operations/cellular automata (controversial claim)

She points out that an universal Turing Machine could be simulated using the rule 1.1.0.

The general field of simulation does not prevent a question about physical reaction: how could some kind of physical activity could be generated by a cellular automaton mechanism?
Quantum mechanics are underlaid by a cellular automaton process. Hayles then expands on Friedkin's Digital Philosophy, and draws hypothesis to imply from his theories:

- the universe is digital in nature, not continuous >>  time & space thus would be discrete and not continuous:  generated by the incremented cellular automata. Our human threshold of perception is too low to perceive this discrete structure, but if the comparison to our perception of how a movie works may help us understand: we watch a movie as a continuous stream of data, but actually this stream is cut into little pieces of discrete units which movement goes too fast for us too grab.
- from then on, reality might be a software run by universal computation; but by definition, software can never know/be conscious of the hardware on which it runs. Thus we should never know. Here there is a shift from physics to metaphysics: we are fighting with a claim (the universal computation) that we can not test.
- there are testable claims: time & space as discrete rather than continuous.

The world might be then run by large permutations: Hayles tries to define this by calling this computational universe a "possibility space". The current and expanding research on simulation ( the science of the 21st century for Hayles) is trying to make this "possibility space" a "searchable space" (to find and normalize a new paradigm).

She evokes the classical problem of the observer : how do you posit a subject in a scientific experimentation, in order to have an objective view of the phenomenon? The most critical moment of this hard question was reached in the quantum physics experimentations. But in simulation, there is no problem: you simply model the observer as a function in the system.

Code becomes the discourse system of nature of today, just as mathematics were said to be the "language of nature" in the 17th-century. The process is different because our model is not based on calculation anymore.
17th: mathematis is the langage of nature; but here it's not calculation

--> all this (universe as computer): an overdetermined metaphor? (from a cultural critique point of view). But it is an inevitable metaphor in our computationally intense society.

But these scientist are talking literally, and we should take that seriously. Hayles proposes to see this model as an intermingle of metaphor and means.

Deleuze: the virtual is an infintie repository of resources that can dynanically become the actual but the realisation of the actual never depletes the virtual
--> relation of juxtaposition
--> like the "possibility space". Objects are epiphenomenon: we don't live in a world of objects anymore, they are not central to our behavior anymore.

So here is me with Textorizer (http://textorizer.whatfettle.com)

and with a color ASCII image convertor (http://www.degraeve.com/gif2txt.php)

;

mmh, none of them is really convincing. I should try with other pics.

In the perspective on the few posts I did on semantic web and "semantic capistalism", here is a reblog (with the comments) from my fav weblog Information Aesthetics (http://infosthetics.com/archives/2006/03/subject_tag_news_heat_):

a weighted tag cloud that uses a heat map approach to visually distinguish the most popular tags & news subjects. see also newzingo tagcloud & mood news.
[guardian.co.uk|via benhammersley.com]

While a nice concept, this implementation isn't really a heat map--it's just a table with a pretty background.

The "heat" of each cell has nothing to do with the popularity of each news subject. Each cell has a different color based on its position in the table, with no relation at all to the popularity. Which makes it useless if you want to distinguish between news focused around one or two subjects and news more evenly distributed among subjects.

The page uses JavaScript to fill in the table, however, so it wouldn't be that difficult to tweak the script into producing a real heat map.

As it stands, this is a good example of how to misuse color in a visualization.

Trevor Lohrbeer . 17 March 2006

Everything2 (http://everything2.com/index.pl?node_id=13522) has had a heat map for a while, although its colors also seem to be hard-coded. Not sure what metric it would measure anyway.

Grant Hutchins . 19 March 2006

um yeah... this kills both those:
http://www.marumushi.com/apps/newsmap/newsmap.cfm

patricia . 23 March 2006

UPDATE, MONTHS LATER: this project, Bumplist, should be checked if you're interested about mailing-list communities. 

++==++==++==++==++==++

 

  

http://www.thing.net/eyebeam/msg00069.html


I was delighted to find a few other of Jodi's interventions on the mailing list. There are definitely more than only spam poetry. Each of them is reprocessing e-mail threads structures of the listerv.  By doing so, they transform the texts (defined as "a text is authorized/has an author") in data and make new (illegitimate) texts with them. I think they are really smart and beautiful. I have gathered all the intervention mails on 1 page here: http://elgg.net/camillepb/files/-1/4759/jodi_eyebeam.ht

Here is a little selection of what they do. It is important to go actually check the mails in the context of the threads they are embedded in (threads in mailing lists are usually indicated by a cascade of subject lines in "re:" or in which a theme is reformulated) if one wants to grasp the meaning of what a very literal type of informational art - or art as a system of data processing - can be.

 

The famous french collective T.A.P.I.N. (based in Lyon, France) runs the xerox magazine called BoXoN (a colloquial word that means both "brothel" and "havoc"), and, for the first time since 1997, they decided to distribute for free the sold-out issues - pdfs dowloadable.

BoXoN covers the field of french visual and performance/sound poetry. They now have a cd with each issue, and have connected the world of e-poetry since a long time ago: some digital works are on display on the website. They also dedicated a few issues to foreign poets.

You'll find everything here: http://tapin.free.fr/

 2 great blog art projects:

1/ jimpunk's blog (from 2003) http://jimpunk.blogspot.com/ 

More jimpunk at http://www.jimpunk.com and htpp://www.jimpunk.org 

There is not too much info on the internet about them. I suspected them to be the Jodi guys, but apparently, they're French. I found a description of their career in Google's cache (the actual link doens't work anymore), so I built a mirror page as a relay of this lost information: http://elgg.net/camillepb/files/-1/4846/jimmpunk_CV.htm

2/ coupons≠coupons at http://couponscoupons.blogspot.com 

check it out! 

What if Gmail ad system doesn't get what you're talking about? What if they cannot find a keyword relevant enough to match with their ads database? I guess they process some random ads. Or maybe they have a special database that is searched into according to what they consider as your own nonsense. Maybe thay have psychologist bots that scan you and make a personality test in order to add some relevancy to the personlized ad system.

In my case, I saw a couple of ads come back regurlary, accompanying mails that, I reckon, made a little sense, or maybe no sense at all. Thus, Gmail seems to have classified me into an hypnotized schizophrenic mail user whose specializes into typing stuff, and needs a job. Typing stuff is indeed what I do, needing a job maybe, yes, soon, after the delirium is over.