|Art & Artificial Life International Competition |
| Home | Entry | Editions | References | Español |
|Pamphleteer: A Propaganda Robot for Cultural Resistance |
This paper presents research findings based on performance evaluation of Pamphleteer, a propaganda robot which automates the often dangerous practice of distributing subversive literature to the public. In field tests, Pamphleteer consistently out-performed human activists in quantity, scope, and efficiency, and also scored significantly higher on the cuteness obnoxious scale (COS).
Contestational Robotics, cultural resistance, public space, propaganda, subversive literature, activism, cuteness factor
Although the internet has become an effective tool of information dissemination, handing out literature in real world public environments remains the most effective means of reaching large numbers of people in a given locality. However, activist groups attempting to utilize this technique face three obstacles which often impede their effectiveness. First, market saturation of street leafleting causes activist messages to be lost in a sea of religious tracts and "all you can eat shrimp night" coupons, while pedestrians eschew interaction with anyone who appears to be promoting anything.
Second, due to limited financial resources , grassroots activist organizations often depend on volunteer labor to perform their leafleting activities. As a result, the activist leafleting is often performed by workers such as hippies, punk rockers, teenagers, and the elderly who wind up in the volunteer labor force because their personality defects make them unsuitable for gainful employment in the private sector. The activist labor force is notoriously flaky. This unreliable performance is compounded by the lack of any financial incentive to work efficiently.
Third, the privatization of public space has had a chilling effect on activist communications. As malls replace marketplaces and parking lots replace parks, distributing unsanctioned information is becoming an increasingly high-risk endeavor. Once protected by free-speech laws, activists now face fines, imprisonment, and bodily harm for distributing literature on what had previously been considered public property.
In response to this need, the Institute for Applied Autonomy undertook the development of a robotic solution which automated the often dangerous practice of disseminating subversive literature to the public. The proposed benefits of such a robot parallel those long touted by the military/commercial robotics industry: 1) An ability to operate in conditions deemed unprofitably dangerous for humans. 2) An ability to work long hours without need for 'break' periods. In addition, the project was guided by the principles of Contestational Robotics : namely that robotic systems designed for activist use must be inexpensive, easy to construct, and highly portable.
Pamphleteer is a humanoid robot designed in the tradition of American science-fiction film and ultra-cute Japanese toy aesthetics . The robots utility is not driven by technological sophistication, but rather by its aesthetic appeal, or "cuteness factor" (CF) . The robot's form is based on an exhaustive study of contemporary paradigms of cuteness such as children, kittens , and teddy bears, and is manifest in Pamphleteer's oversized head and eyes, and through the robot's short, chubby body. CF is further enhanced by the robot's full complement of costumes, which can be customized for particular applications. For example, Pamphleteer wears an athletic jersey and baseball cap at college functions, and is clad in a Santa Claus outfit for late-December events.
Pamphleteer is capable of limited speech and movement, which also contributes to CF. After several iterations, a high-pitched, pleasantly squeaky child-like voice was deemed an appropriate complement to the visual cues. Head rotation and sonar sensing technologies are employed to create believable movement, while allowing for only the most limited interaction.
Technically, Pamphleteer utilizes remarkably simple and cost-effective technology. The robots body is custom-built from aluminum stock, and uses a $30 Basic Stamp to monitor its front mounted sonar and control head-rotation motor and speech functions. Audio is recorded onto a $12 off the shelf voice recording chip available from most electronic stores.
While Pamphleteers internal systems are 'low-tech' by robotics standards, it appears that the impression of technological legitimacy can be achieved equally effectively with even less sophisticated technology . We are considering ways in which the current design might be made even simpler for the next prototype. Future versions of Pamphleteer may rely on cordless microphones, remote control systems, hidden tape recorders, or embedded midgets/children.
To test Pamphleteers effectiveness, field research was conducted on several separate occasions under differing terrain and weather conditions. In each case the robot and a human activist were deployed on similar street corners within the same commercial district of a metropolitan city. Each was given equally subversive literature for dissemination and no performance enhancing drugs were used. Performance was monitored by anonymous researchers dressed in civilian clothing, who also conducted exit surveys with participants as they left the experiment area.
Field studies have conclusively demonstrated Pamphleteer's effectiveness in engaging the public, with particularly notable success in reaching notoriously difficult populations such as the elderly and supervised children. Generally speaking, the robot is capable of distributing 23% more literature to 18% more people than his human counterpart, and is capable of performing for up to 6 hours without interruption, as opposed to an observed limit of 78 minutes for an unpaid human volunteer. We expect that the next generation prototype, which utilizes more powerful batteries, will further widen this gap.
While people were much more willing to interact with the robot than with human activists, the duration of these interactions was much shorter, which further contributed to Pamphleteer's ability to outperform the human. In aggregate, humans tended to interact with the robot for no more than 10.2 seconds, as opposed to an average interaction time of 3.45 minutes with human activists. Our hypothesis is that Pamphleteer is perceived as less intelligent than a human activist, and as a result, people are much less likely to engage it in conversation. This may also explain the observed difference in risk, calculated at 2 threats of physical violence against the human and 0 threats towards the robot. This is notable because it is possible to program Pamphleteer to be more verbally aggressive towards passers-by than human activists, even to the point of making derisive or lewd comments. We suspect the reason for this is that the behavior was mitigated by the robots overall cuteness, and may have actually enhanced public perception of Pamphleteer as a "fun" device. When passers-by were asked to rate the human and the robot on the cuteness-obnoxious scale (COS). Using a rating system in which 10 = "cute"; 1 = "obnoxious", human activists received an average COS score of 3.23, while Pamphleteer averaged an astonishing 8.56.
Our initial findings have been extremely encouraging. Now that we have demonstrated proof of concept, we are exploring additional functionality for our next prototype. In addition to improving Pamphleteers battery system (described above), we are incorporating a magnetic card reader into Pamphleteers head, which will enable the robot to receive drivers license and credit card information from users. We expect that this will enable Pamphleteer to collect petition signatures and to perform basic fundraising functions, such as selling bumperstickers and badges.
We also recognize that there is a need for additional testing. Having demonstrated Pamphleteers ability to outperform human activists in relatively safe environments (public streets), the logical next step will be to test the robots effectiveness is more contested space. Additional experiments are planned for such locations as shopping malls, government buildings, and corporate offices.
In conclusion, Pamphleteer is capable of reaching more people, more efficiently, with fewer risks than human activists. While much of this performance is attributable to the cuteness factor described above, we hypothesize that these results are also dependant on social and cultural factors particular to street activism.
In addition to issues of perception, human limitation, and contextual noise identified above, human activists are further hindered by the social marginalization which faces any obvious act of political discontent. The tendency of mass media and the general public is to characterize political activists as soft-headed, politically-correct drones (on the left) or violent, paranoid psychopaths (on the right). In addition, many activists choose to manifest their opposition to the status-quo through outlandish or unorthodox personal appearance, creating a social environment in which information originating from a position even slightly left or right of center is dismissed as "crazy" even before the message is received. In a media-saturated environment, messages originating anywhere other than "reliable"
sources, such as large media organizations, elected officials, or corporate publicity departments are ignored. Contemporary activists can no longer be solely concerned with the content of their messages, but now must equally consider both the manner and context in which these messages are received. In other words, in order to be heard, activists must utilize the form and content of the cultural institution which they intend to critique - by co-opting those institutions and/or appropriating their representation.
Robots, almost exclusively products of trusted military/industrial/entertainment institutions, provide an excellent real-world means by which this subversion may be achieved. When deployed into a public space, "Trojan horse" robots are perceived as representatives of these "safe" institutions, and as such, the public is willing, and often eager to engage them. Messages delivered by these devices are not suspect until they have been received and analyzed - providing an opportunity for activists to bypass social conditioning and impact the public consciousness.
1.Institute for Applied Autonomy and Critical Art Ensemble, 1999. "Contestational Robotics," ReadMe! ASCII Culture and the Revenge of Knowledge, Josephine Bosme, ed, (Nettime, Autonomedia: NY)
2.Sanrio, Inc. (http://www.sanrio.com/)
3.The Cute FAQ (http://www.ualberta.ca/~bfan/cute/def.htm)
4.Kingdon, J.S. 1980. "The role of visual signals and face patterns in African forest monkeys (guenons) of the genus Cescopithecus". Trans. Zool. Soc. London. 35, 425-475.
5.Atlas Robotics (www.atlasrobotics.com)