The SUICIDE BOX is one of our earliest information devices and demonstrates some of the strategies the Bureau has been developing around information.

The Suicide Box is a motion-detection video system, positioned in range of the Golden Gate Bridge, San Francisco in 1996. It watched the bridge constantly and when it recognised vertical motion, captured it to a video record. The resulting footage displays as a continuous stream the trickle of people who jump off the bridge. The Golden Gate Bridge is the premiere suicide destination in the United States: a 100 day initial deployment period of the BIT Suicide Box recorded 17 vertical events (during the same time period the Port Authority counted only 13).

In 1996, the Bureau released the video documentary "Suicide Box", documenting the deployment. When the video was exhibited at the 1997 Whitney Biennial of American Art (1997), the curatorial essay identified it as an imaginative work, "a seemingly factual documentary using a combination of real and fictionalised statistics and staged surveillance footage to create something utterly convincing .. completely plausible". A curatorial response implying not only the fictional nature of the BIT camera equipment, but also of the tragic information field in which it was operating. This review expresses a recurrent theme in artworld reaction to Bureau practice - the inherent suspicion of artists working with material evidence.

There are however a number of ways to interpret the Suicide Box data. In 2003, the Bureau launched the DI10, based on a 10 year analysis of the DESPONDENCY INDEX, a novel BIT economic indicator which takes the moving average of the suicide rate off the Golden Gate Bridge and indexes it to the Dow Jones Industrial Average.

One question frequently asked of the Bureau is, "do they correlate"? That is, does the rate of suicides from the Golden Gate bridge correlate with the Dow Jones Industrial Average? The answer is, of course, "no". But in that question is the cultural expectation that the Dow Jones Industrial does have comprehensive representational power: that it somehow should or could capture and represent the health of a nation. This expectation was amplified in the US directly after September 11th 2001, with the incessant reporting of the Dow Jones Industrial Average and the NASDAQ as if the drop respective economic indicators somehow measured the losses.

What the Bureau demonstrates with the DI10 is that the Dow Jones Industrial is radically exclusive information: a highly privileged yet very partial representation, an arbitrary economic performance of 50 companies which in no way could or does represent tragic social phenomena such as suicide in the Information Age.

In a recent covert news release, BIT released the DI10 data on a series of taxi cab screens in Manhattan. 400 New York City cabs have been equipped with an on-roof GPS information display, typically used for location-sensitive store advertising; or to display up-to-date financial or news tickers to pedestrians. BIT is hijacking this system for a number of information releases, including the street release of the DI10.

The DI10 was released for a number of minutes before it was shut down by Adaptmedia, the for-profit entity that equipped and is inhabiting these screens - creating a new public media space but dedicating it entirely to burger menus, ESPN, sports results and the NASDAQ.

What is of interest to the Bureau about this exhibition space is that it is a new form of public display, iconically for the display of information, but one that simply transposes and reiterates existing information priorities. The Dow Jones Industrial's fluctuations are reported and updated with minute attention across the media: in the Information Age this data is what is counted, whereas something like the suicide rate off the Golden Gate Bridge is not registered - literally does not count as information.

At the time of the Suicide Box installation, BIT was experimenting with a new form of corporate relations which we called inverse sponsorship. The project required a Silicon Graphics Onyx - a $40 000 computer - to watch the bridge and do the realtime video-processing. The Onyx was on loan from SGI on the condition that we "in no way associate Silicon Graphics with the project" which we of course did, we used their logo on everything and thanked them in the video credits.

That was 1996: in 2004 the same deployment is possible with a consumer-level camera, and high-speed wireless connection to upload the video data for centralised processing. So this kind of realtime monitoring for informational purposes - counting things that are not counted - is increasingly possible,

The Golden Gate Bridge is not merely a Californian tourist icon, it also marks the gateway to the Silicon Valley - just south of San Francisco - where the Bureau conducted further information-gathering research via agency of the bitplane.

The BITPLANE is a radio-controlled model airplane, equipped with a micro-video camera and transmitter. In 1997 it was launched on a series of sorties over the Silicon Valley to capture an aerial rendering. Guided by the live control-view video feed from the plane, the pilot on the ground was able to to steer the unit deep into the glittering heartlands of the Information Age.

Most of the corporate research parks in Silicon Valley are no-camera zones and require US Citizen status or special clearance for entry. The bitplane (citizenship undisclosed) flew covertly through this rarified information-space, buzzing the largest concentration of venture capital in the world, to return with several hours of aerial footage.

Bureau observations recorded that the bitplane was traversing and transgressing several strata of airspace on this sortie. Firstly aerospace: it is illegal to fly RC aircraft within 5 miles of an airport (there were 3 airports within 5 miles of the bitplane flight path including Moffat Field, a Lockheed Martin and NASA jointly operated military airfield). It is also not permitted to fly model planes in the Palo Alto area due to city ordinances on noise pollution. The bitplane additionally violated Federal Communications Commission regulations by transmitting live video on Cable channel 51, momentarily inserting bitplane gods-eye video into the regularly-scheduled television viewing of the households in its flight path. So the plane revealed in its wake multiple layers of policing and control over what seemed to be transparent airspace.

One of the issues the bitplane engaged with on this mission is the threat of the camera to information space. Cameras are not permitted into the corporate research parks in the Valley on the logic that visitors could steal intellectual property by taking photographs of it. This reveals an assumption about what information is - that it is something that you can take a photograph of, a thing - not the product of a community of expertise or social network of shared discourse, but something you could go in and steal with a camera.

This traditional view of photography, evidence and information reinforces a definition of information as property - and therefore that property laws can hold. The Bureau holds that information doesn't in fact exist or circulate in those sorts of forms.

Economic indicators are a format that the Bureau works in regularly as they are the native genre of information.

The HALFLIFE RATIO is an economic indicator that tracks the market value of sperm and the market value of ova, and compares the two. The Bureau has noted some striking anomalies. Keeping in mind that the market demand for sperm and ova is almost equal, the market price is very different. One type of doner [F] comes out about $38 000 down, per viable unit, and one type of doner [M] comes out about 3 cents per unit up*

Typically in economic formulations, if a good is ephemeral - if it has a use-by date - this will push the market value of that item up. This is not the case with sperm and ova. Sperm and embryos can be cryogenically stored, but ova cannot: not commercially or in any volume. There is also the risk of litigation, which is 12 times higher for women than it is for men; and the short-term, medium-term, longterm and "unknown" health risks, which are much higher for women than they are for men.

Additional analysis of the contracts reveals that contracts in sperm donation are all in the language of transaction. Conversely, the language used in ova transactions, specifically in the UK and Canada where it is illegal to be paid for donating ova but not for sperm, is the language of donation or philanthropy. The Bureau views this as a real and clear example of a very new technology - assisted reproductive technology - reproducing very old and familiar inequities.

The BIT halflife ratio collects this publicly available data on sperm and ova and transposes it from the statistical realm into an animated, much more "Discovery Channel" kind of genre.

Information is traditionally held to be an unmediated representation of a phenomenon. For example it is accepted that a there are good and bad journalists: that the good journalist reveals the facts but a bad journalist could be bribed. This belief ascribes an external relationship between power and information: that there is a truth and that either the journalist produces it truly or that they're under some kind of bias. What the Bureau has been working with is a different assumption, of a very internal relationship to power. That is that information and power co-produce each other, that power is not something that warps information, but produces it.

There is an idea that information is comprehensive, a-historical and true. The Bureau presents the evidence that information is partial, particular and subject to all sorts of political influences.

Voicechips are the cheap, ubiquitous speech-synthesis chips integrated into products to give them autonomous voice. Embedded in toys, trashcans, automobiles, phone networks, relaxation and emergency products, voicechips stand on the frontlines of human/machine relations. The Bureau has been investigating this cultural phenomenon, along with their silent corollary the speech recognition chip. The voicechips pose the question, now that our things can talk to us, what do they say? The speech recognition chip asks a similar cultural question: now we can talk to our things, what do we say?

The Bureau is building a database of voicechip samples and the human agents struggling to use them, including the whole set of computerised telephone voices that are scripted to refer to themselves as "I". The database is designed for archiving diverse audiofiles online and in the public domain; semantically tagged, searchable and open for study, remix and reuse.

BIT research into the speech patterns of distributed voices has included the development of our own audio hardware, the VOICEBOXES.

Instructions for use: 1) open to record; 2) to play: move, throw or touch the box ; 3) place appropriately . The voice box provides an unscripted opportunity to be heard. It records when opened and plays when moved or manipulated in any way. Because of the simplicity of the device it can be used in complicated ways that combine spatial positioning, temporal displacement, and verbal expertise. Interacting with these multiple dimensions provides a limitless number of combinations, in contrast to the predetermined menus of choice in other digital / 3D audio environments. They can be used one at a time, or many can be used to build a localised digital sound environment.

All of the video and audio samples the Bureau is collecting are open source, that is they are designed, collected and intended to be reinterpreted and reused. This is a very different idea of property, genre and relationship to authorship from the traditions inherited from news-documentary or fine-art production. BIT treats audio and video primarily as data - as can be seen with the Bangbang camera network.

BANGBANG is a BIT news network of sensor-activated webcams, installed for sustained observation. The network takes its name and its logic from the first Bangbang camera, a webcam triggered by the audio signature of gunfire or explosion. (The camera's audio sensor is tuned to recognise gunshot acoustics: whenever there is a BANG it will upload several seconds of audio and video to a permanent record; so it archives just the media associated with the bangs). The Bangbang cameras are designed for sustained installation, embedded in war zones and other areas of flash news attention.

Other deployments have included the SLAMCAM, an amplitude-triggered camera placed at an alarming San Francisco intersection, Division and Portrero, in 1998. The Slamcam watches the intersection and buffers video: whenever there is a screech or a high-amplitude impact noise, it will capture the event.

In 1999, the Bureau installed a radiation-triggered Bangbang cam at Fresh Kills landfill, Staten Island New York. An onboard dosimeter measured local field intensities from background radiation; every time the reading fluctuated above the EPA "safe" level, the Fresh Kills cam would capture 2-4 seconds of video. Official records report no radioactive waste at Fresh Kills, formerly the largest landfill in the world and recently closed for conversion to a recreational public space (before being reopened on September 12th 2001 to receive the World Trade Center debris). Fresh Kills is already marked as a parkland on all the maps, but despite official reports of no radioactive materials present, the Bureau's webcam was triggering constantly.

The design behind the Bangbang project proposes a kind of editorical logic that is legible. Contrast this with CNN. Broadcast news crews are despatched to "hot" news locations where they gather footage for the night's news headlines. The editorial logic - why you are seeing what you are seeing - is opaque. (How do they decide to go to Kosovo or East Timor that time and not other times? The editorial logic is illegible). In the case of the Bangbang network the editorial logic - the trigger algorithm - is available. Dumb logic: BANG - you see the footage.

This a kind of editorialisation that produces video for reuse. The clips are available in granular amounts, they are incremental and downloadable; you can choose to see the footage over weeks or months; it will auto-assemble for incorporation and interpretation in other formats. Documentary filmmakers, VJs, environmentalists and academics have used the Bangbang footage.

Seagulls have featured as an unexpected trigger at a number of BIT equipment installations.

In Suicide Box trials, the algorithm had to be adjusted to tune out the problem of spurious dive-bombing seagulls that would activate the box to record. The seagull factor emerged again at Fresh Kills, where initial viewing of the Bangbang video data showed seagulls appearing to trigger the camera there. Very little had been published on this phenomenon at the time, but it has since been observed also at Sellafield in the UK, that birds biologically concentrate radioactive waste and radioactive dyes.*

A hospital on Staten Island also monitors radiation levels via a sensor on the outside of their building - they use the same sensor and a very similar area to that which the Bureau was covering. The readout from the hospital sensor is the same data, but delivered as a conventional radiation plot, it looks like information; it looks scientific. The Bangbang camera instead outputs information in the commonly legible medium of video. The hospital was not seeing radioactive seagull phenomenon because this is not what they were looking for, but the Bureau was able to extrapolate this possible affinity from the visual media of the seagull triggered footage.

This is a claim towards the legibility and interpretability of information using open-ended, popular formats like video, as opposed to closed, scientised genres of information rendering. Open-ended formats can be reused and reappropriated; they enable diverse interpretations beyond the predetermined monitoring environment.

Another application developed from the Bangbang project is the FADE-TO-BLACK cameras. These are webcams that are oriented straight up watching the sky; they fade to black with the gradual particulate accumulation on their lens. (When the camera fades completely to black, a human operator will have to come along and clean the lens for monitoring to resume). Fade-to-black cameras are systematically triggered, taking a frame every 15 minutes, to archive in animation the ongoing movement of the sun across the sky as it gradually fades to black

The first installation of these cameras in New York was motivated primarily by diesel fumes. In 2001 the Bureau had one webcam on Broadway, Manhattan, and another one at the River Arts Center in the Bronx. In analysing the camera data after September 11th, the Bureau observed that even as the World Trade Center towers fell down and dust settled over the city, the fade-to-black cam in the Bronx, where all the incinerators are, still darkened faster than the one in Manhattan. This readout gives a very material, very visceral measure of actual environmental disparities - the concentration of incerators in in a low-income area - against the more newsworthy toxic event represented by the incredible amount of particulate matter generated as the towers collapsed.

The fade-to-black project is designed as a way to format a local relationship to air quality via material evidence. The issue of being able to have your own material evidence assigns a particular relationship to the information and to political differences around air quality - as opposed to consuming, for example, the Environmental Protection Agency's 62 page public report. This is a public document in terms of being publicly available - but who is reading it? In Bureau assessment, the artists in Manhattan and the mothers at the Bronx high schools are not. The fade-to-black information, rendered visibly as video clips, is more legible than the EPA report, it is functionally more available. And it gives an authority, it provides material evidence with which to claim that something should be done about air quality.

The Bangbang cameras, like other BIT systematic collection devices, are available to copy and build: they can be used, with credit, in other projects. It is not just the footage that is openly available but the information design, downloadable in PDF format from the Bureau website. These instructional kits contain design materials and information for building and deploying a number of BIT devices.

A subsequent BIT incursion into the field of material evidence is D4PA, or Design For Political Action - a set of projects the Bureau initated around the World Economic Forum at the Waldorf Astoria in New York, February 2002. The D4PA represented a number of BIT devices, including the BIT RADIO which was deployed at the WEF over a 3-day period.

The Bit Radio is an intermittent FM "autopirate" transmitter, covertly installed on this occasion inside the Waldorf Astoria Hotel. The unit was activated to broadcast over the top of local National Public Radio for 1 to 3 seconds every time the levels of airbourne asbestos in Manhattan fluctuated above what the Mayor described as a "safe" level. After September 11th, there was considerable concern over air quality as huge clouds of asbestos were pumped into lower Manhattan - not a region actually used to that kind of industrial loading of environmental toxins - along with thousands of gallons of burning diesel fuel that were stored in the Mayor's WTC command and control bunker, and a host of other volatile compounds and contaminants. The public concern about air toxicity was immediately answered by the Mayor Rudi Guiliani's announcement that the air was safe to breathe.

At the time of the WEF gathering, the Bureau saw this piece of sensor-gathered data on airbourne asbestos levels as information that literally unified delegates on both side of the police lines. We accordingly used it to trigger the Bit Radio transmitter to overcast a small bleep over the top of NPR radio WNYC on each asbestos alert. A tangible reminder that air that the delegates were breathing and the air the informal forum delegates outside the Waldorf Astoria were shouting in, was contaminated with the same material. The kind of discussion about globalisation going on within the forum of the WEF was by definition a discourse that obscures, downplays or under-represents the real, local material conditions that it's situated in.

In 2001, NPR along with the commercial radio lobby - the National Association of Broadcasters (NAB) - successfully lobbied Congress to oppose a proposal (since reinstated) to grant 2,000 new radio broadcast licenses to community groups around the USA, arguing that their broadcasts would "bleed into" established stations. Hence the Bureau chose an NPR station and its established audience to transmit over the top of. Transmitting in short bursts of data meant the FCC Enforcement Division would find it difficult to track the signal: they are looking for community radio - normally music or discussion - to shut down; not intermittent bleeps of information. The FCC can't lock onto this signal for long enough to be able to identify it, hence it takes on the quality of stealth transmission - a very simple technology but one the Bureau has found to be very effective.

After September 11th, there was a heightened level of police aggression measurable in Manhattan. Another BIT project almost-launched as part of D4PA illustrates the tension around street protests at that time.

The BIT ROCKET is an adapted hobbyist rocket, fitted with a micropower video transmitter and portable ground-based receiver similar to those of the bitplane. It is designed to be launched from the middle of a crowd or demonstration to capture a family aero-portrait - in this case the incendiary moment at the end of the march, the mass of bodies assembled outside the Waldorf Astoria. The rocket streams video while ascending to up to 600 feet in height, then parachutes gently back down to the crowd

The Bureau is developing headcount software for this rocket application. This is open source software which takes the video frames, assembles them into a compound image and does a headcount of the number of heads captured from aerial vantage. Again, addressing this issue of material or imagistic evidence, of information - in a situation where there typically will be a 10 to 100-times differential in what Indymedia will report as crowd attendance, and what Reuters or the police will calculate. The Bureau's system produces both a legible compound image and a live, peer-readable quantitative count.

The Bit Rocket did not get fired at the WEF gathering due to alarm from within the activist community that a launch could create an excuse for teargas attacks on the crowd. This self-policing was, by Bureau reckoning, the most alarming element of the event.

A later proptype was developed for February 15th 2003, the worldwide street protests against war in Iraq. February 15th is the day after Valentine's day and downtown Manahattan provides a surplus of red heart balloons available at half price. The rocket cam was reconstituted as an activist-nonthreatening red balloon cluster; and floated down 5th Avenue to Washington Square Park where it got entangled in a number of trees. The unit was transmitting video from altitude approximately halfway between the street protester point of view and the New York Times journalist helicopter vantage point. The video was streamed on a publicly accessible television frequency, producing a set of images that were available for circulation and reinterpretation.

In 2004, the Bureau chose to celebrate the first anniversary of the February 15th protests - the largest coordinated global street demonstrations in history - with another group of globally defined subjects, the globally wandering Graylag geese. To mark the occasion, we hosted a black tie cocktail event for humans and geese in Zeewolde (NL), where we attempted to share food.

Along with other communications initiatives between the human and nonhuman species present, a Bit Radio was installed to transmit goose emergencies into the local radio space of Zeewolde. Every time the geese squawked at a certain amplitude, it would transmit that squawk over the top of easy-listening Nordzee 95.9FM in the area - inserting attention to local conditions that are usually obliterated or underrepresented in mass media genres.

Another BIT interface for handling wild information is the Uphone, a phone-in-audio interface to the web. The first application of this system in 2003 was the ANTITERROR LINE.

The Antiterror Line is a project that addresses the increased incursions into civil liberties that have become ubiquitous in the post September 11th context: the arbitrary exercises of power that go on in the name of antiterrorism. Incidents such as being detained at Heathrow airport for over an hour for wearing a t-shirt with a gun on it by British Airways security staff concerned "for the well-being of the other passengers". Often these attacks are minor inconveniences. However as a distributed phenomonon they are worth assessing, which is what the Bureau has done with the Antiterror Line.

The Antiterror Line is a public audio database and national phone number (UK and US). The next time you are under antiterror attack you can dial the number discretely and use your phone to record events - or you can call in later to report an antiterror incident experienced. Within 5 minutes, the recording will upload to a publicly available, read-writeable database, marked by the date and time of the event. Later you can go to the site and annotate it, or not, or anyone else can - but it is there, compiling your testimony or experience with everyone else's. A database of civil liberties infringements that by themselves might be trivial but accumulated, accreted, amassed, may amount to something more.

In 2002 the US Attorney General John Ashcroft set up a similar line under the US Patriot Act, called the Terrorist Information Prevention System or TIPS. This was a hotline that recruited thousands of Americans, "TIPS volunteers", to report suspicious activity. They were volunteers because they were not paid to assume the the risks of reporting on their neighbours and friends; and of not knowing what would happen to that data. Reporting a suspicion can incur substantial personal risks, and in other informant programmes that the Federal and State governments run they both pay and carefully ensure the safety of informants. In this case, volunteers were just expected to tear at the social fabric of trust that underscores and underwrites daily interactions voluntarily.

The BIT Antiterror project was set up very much in contrast to Ashcroft's TIPS project and has a number of features that are different, crucially with respect to the structure of participation.

The central question is who gets to produce the information and who gets to interpret it. With Ashcroft's TIPS Line, only CIA or Federal agents get to interpret the information. In the case of the Antiterror line the database is public: anyone can listen to the calls as they are deposited in the online database; annotate the system, reuse the samples, or reinterpret them.

Who assumes the risk and who assumes the costs in both cases is quite different. In the case of the TIPS line, it is an 800 number - a free call - but affords no other protection or assurances. In the case of the Antiterror line you do have to assume the cost of a local phone call, the costs are distributed. This also marks the absence of that power asymmetry - that there's not a centralised agency that is bearing the costs or doing the interpretation or exclusively reusing the work.

The final project we will present here is another Uphone utility, the SPARROW REPORT LINE. This is an audio database which uses the vast proliferation of mobile phones to measure the catastrophic decrease in the population of sparrows in London and New York.

The cause of this drastic decline has not been determined - something between a 90% and 99% dive in the London sparrow population over the past 50 years - but it does suggest an indictment of the material conditions of our shared urban environment.

In 2003, the Bureau launched the Sparrow Report Line, with local numbers in sparrow-depleted East London and Manhattan. Citizens can phone in and record either directly (use phone as microphone) or verbally (spoken report), their sparrow sightings, non-sightings, theories, or expressions of regret. After uploading a recording the caller will be able to download a sparrow ringtone, to somehow replenish the urban soundscape of these chirps, or perhaps just to amplify the loss.

This project addresses a different context but a very similar structure of participation to the Antiterror Line. Anyone can upload a sparrow report, anyone can interpret the data. The interface is very simple and recognisable, you record after the beep. The open space of a 5-minute audio recording is designed for a higher resolution of participation than other similar data-gathering systems (surveys, questionaires, collective population counts): anyone who has a theory about the sparrows can upload it for public consideration.

In Bureau evaluation, the sparrow decline is an embedded/potential environmental catastrophe, a phenomenon that involves complex environmental interconnections with our own lived material conditions. The Bureau proposes that this open form of participation has a higher probability of finding some explanation, than delegating interpretation to any single community of expertise.

This set of projects from the first long decade of the Bureau demonstrates the politics of information that we are interested in: what counts as information and what doesn't; who produces the information, who interprets it; and what gets called truth or knowledge in the Information Age.

This document is a preliminary text for the forthcoming Bureau book, The Decade Report, the Bureau the First 10 Years; and will be further developed at this site.

Produced with moral, political and financial support from the Research Fellow programme of Media Design Research, Piet Zwart Institute, Willem de Kooning Academy Rotterdam. http://pzwart.wdka.hro.nl