Close
When you subscribe to Furtherfield’s newsletter service you will receive occasional email newsletters from us plus invitations to our exhibitions and events. To opt out of the newsletter service at any time please click the unsubscribe link in the emails.
Close
All Content
Contributors
UFO Icon
Close
Irridescent cyber duck illustration with a bionic eye Irridescent cyber bear illustration with a bionic eye Irridescent cyber bee illustration
Visit People's Park Plinth

Dreamlogs [2005]

05/05/2005
Dan Waber

prototype (test release) project Dreamlogs – by Christophe Bruno

My first visit to Dreamlogs by Christophe Bruno left me more than a little confused, but interested enough to do a bit of digging to try and figure out what was going on. There’s not much in the way of concrete instruction as to how the ‘idea association engine’ is intended to be used, though, for a project in the ‘testing stage’ this is not an insurmountable problem–particularly for a piece of web art, where it’s not unsurprising for fully realized pieces to come with no instruction at all beyond the feedback one gets from clicking around and trying things.

So I clicked around and tried some things.

On the way I learned that Dreamlogs are “The new world order” and “confront the temporality of a dreamed life with the infinite dimensionality of speech” and “an idea association engine”. “They propose another way to surf on the Internet, by disentangling the discourses that have interlaced over time” and “a new interface to the Global Text: a non-utilitarian and non-local alternative to search engines”. Cool, I’m up for that. Now, what do I do?

Choose my target language…Check.

Choose a sentence or part of a sentence, ok, put some words in…Check.

Optional: put more words in to “modulate” my search. Hrm. Since I don’t know exactly what that means in this context (though the “+” is a good clue), I probably don’t need it, yet.

Go!

Magic happens in the background, and, then I see poured onto the page a bulleted list of text excerpts from what I presume are web pages, each bulleted item showing my word(s) highlighted in the text. OK.
I’m with you so far. Now what? Read some new instructions on the left side. So…I can save it now (seems silly), end the process and start again (good to know but not yet) or do what I’m supposed to do, which is click on any words I see anywhere in the bulleted list items.
When words are clicked they appear in the new search box, right under a display of the current (previous) search(es). OK. So, I’m doing a little free associating. Cool. Clicky clicky. Clicky. Clicky Go!

“No (or very few) results were found, click here to relaunch”

Ooops. Must of done something wrong. Hrm. OK, relaunch.

Work my way back to where I was and this time be sure to pick fewer new words to search on. OK. I get it, I guess. Well. Sort of. Where’s this going?

Let’s think about this. I can put any search terms in and I get a slice of the pool of context that surrounds my terms. I can then pull new terms from that context (so they are, by definition, related, at the very least by proximity), and grab new slices of (at least proximally) related context. What for? I need to check out some of the related materials and see if I can get a handle on what’s going on here.

Back out to the first inside page. Lots of ways to go. FAQ of limited practical use. Paints the concept in brush strokes so broad I still don’t know what I’m supposed to be doing. Maybe there is no “supposed to be doing”. Eeep. Possible I suppose.

Back out to the first inside page. Let me do some reading on Christophe Bruno. Wow, cool. A ton of extremely intense art. Claims the earliest piece of search engine art. Excellent. I spend three hours checking out other projects Christophe has done or been involved in (my favorite so far: The Google AdWords Happening, but that should not suggest I only found one thing to enjoy, I thoroughly enjoyed the bulk of what I found).

OK, so lots of search engine explorations. Dreamlogs is certainly a logical extension of that. I’m still missing something, though. I do some more reading in and around the Dreamlogs site and reread the FAQ to see if I can puzzle out some usage clues from it.

I gather I’m supposed to be adopting a discursive position and working to its opposite position in a perhaps fixed number of steps. Why would I want to do that? Nevermind, let’s try doing that.

Maybe I’m just not properly wired for free associating out of a limited pool of context in order to bridge opposing discursive positions. Maybe I’m overthinking the whole thing and I need to just have some fun with it. Oh, what’s this? View Current Dreamlogs.

Oh, okay, there’s a whole other thing happening here. After I get my steps all taken, a blog is created on the fly that contains an entry for each of my steps. The entry is composed of an image, an excerpt
from a web page, a “read more” link to the actual page in its natural habitat, a “posted by” credit, and two numbers I can’t really figure out: Interest, and, Correlation With Previous Post. I can guess what they might signify, but, I can’t figure out how, in the context of this whole experience, I’m supposed to use or influence or be influenced by them. Alright, so let’s make some and save them and see what they look like as blogs.

I made two. In the first I started with a beginning of the day phrase and figured I’d try to make it to the end of a day. Not exactly an opposite discursive position, but, not a purely random surf, either.
I’d learned that longer phrases had a tendency to return no results, and single words return a huge amount of results, so I settled in for three to four word phrases and generally kept to phrases that were naturally occurring mostly intact within the blocks of context I was presented with. When I saved it, I learned that the posts appear in descending order, last first and first last. Not a major problem, but, would have been nice to know at the outset (if I’ve been paying attention when I looked at the other finished Dreamlogs, I could have figured that out, so, partially my own fault).

In the second, knowing that the final piece was going to be presented in reverse order, I set out to compose a poem in reverse order that would (as near as possible) end as it began. The phrases I selected to further search on were now chosen in an interactive fashion, visually scanning the context pool while attempting to build lines that would stand on their own and be readable as a poem in either direction (at the line level of granularity). Instead of pulling phrases out of the context, I was pulling single words from all over the page in the order I wanted them to appear. Totally do-able, but my Correlation With Previous Post scores dropped to ridiculously low numbers, as did my Interest numbers. Maybe it’s like golf, and I’m supposed to score low! Or maybe not.
So, it all boils down to: I haven’t been able to figure out what to do with Dreamlogs. In a way, it’s like doing collaborative writing exercises with someone else, only in this case the someone else is a search interface to the text of the web. I can see it being in improvement over a human in the range of context to associatively draw from, but that means trading relevancy for number of results, quality for quantity. I can see it as a way to get at associations and connections that might not normally present themselves, but I don’t see how it’s an improvement over myriad other generative techniques, including but certainly not limited to just garden variety surfing. I do, however, love the building of the blog to house each finished Dreamlog, and, to me, viewing the finished Dreamlogs of others was more interesting than the process of building them myself.

Main image by Anne Helmond.