Interactions - March/April 2012

from humans. The system went from perceiving the world, to deciding what to do to accomplish its goal, to modifying the world by following and posting—just like a physical robot [ 9].

The sense-think-act paradigm is just one tool from traditional robotics that we can bring to socialbots. Because socialbots act only in digital spaces, they bring a unique opportunity to apply artificial intelligence and robotics theory in a world of near-perfect truths. The other human agents with which socialbots interact may be difficult to predict, but they are easy to observe.

With such a short history, computational social psychologists and others studying social robots would be wise to learn from the decades of literature produced by their peers studying physical robots. Techniques in robotics sensor fusion may provide insight into how to combine multiple sources of information about the social graph, giving deeper insight into people’s connections. Concepts from swarm robotics may give way to armies of socialbots acting in concert, aiding each other toward their collective goal. Autonomous agents are autonomous agents, and whether they are operating in the physical world or the digital world, we’ll be seeing a lot more of them in the future.

Greg Marra (@gregmarra) has a degree in electrical engineering with a focus in robotics at Olin College and now works on Google+.

3. Securities and Exchange Commission. Findings Regarding the Market Events of May 6, 2010; http:// www.sec.gov/news/studies/2010/marketevents- report.pdf 4. Mirvish, D. The Hathaway Effect. The Huffington Post (Mar. 2, 2011); http://www.huffingtonpost.com/ dan-mirvish/the-hathaway-effect-how-a_b_830041. html 5. Gerber, A.S. et al. Social pressure and voter turnout. American Political Science Review 102, 1 (2008); http://journals.cambridge.org/action/display Abstract?fromPage=online&aid=1720748 6. Dillow, D. Tired of repetitive arguing about climate change, scientist makes a bot to argue for him. Popular Science (Nov. 3, 2010); http:// www.popsci.com/science/article/2010-11/twitter- chatbot-trolls-web-tweeting-science-climate-change-deniers 7. Asimov’s Three Laws of Robotics; http:// en.wikipedia.org/wiki/Three_Laws_of_Robotics 8. Boshmaf, Y., Muslukhov, I., Beznosov, K., and Ripeanu, M. The socialbot network: When bots socialize for fame and money. Annual Computer Security Applications Conference (Dec. 2011); http://lersse-dl.ece.ubc.ca/ 9. See Gepetto’s Army, SXSW 2011; http://www. slideshare.net/gregmarra/gepettos-army-creating-international-incidents-with-twitter-bots 10. Pearce, I., Nanis, M., and Hwang, T. PacSocial: Field test report; http://pacsocial.com/files/pacso- cial_field_test_report_2011-11-15.pdf 11. Orcutt, M. Twitter bots create surprising new social connections. Technology Review (Jan. 23, 2012); http://www.technologyreview.com/ web/39497/

plistic bots are able to reliably
generate some type of statisti-
cally significant change in the
behavior of an entire social group,
one might imagine that swarms
of these bots might be designed
to shape or reshape communities
on a very large scale, in what we
might call “social architecting.”
Early experiments have pro-
duced fascinating results. More
recent experiments have gener-
ated bots that are able to “super-
charge” the rate of new friendship
growth between users in a social
group on Twitter. These bots facili-
tate new human relationships by
introducing users to one another
and exposing them to content
from others with whom they do
not usually interact. This tends to
boost the rate of friendships that
emerge between human users
in those networks by a statisti-
cally significant margin, and the
effect is sometimes quite large.
In one case, a bot was able to
boost the regular rate of connec-
tion growth to more than three
times the normal rate [ 10, 11].

However, the future fate of this technology remains an open question. The ethical considerations of such research will become ever more subtle and complex as the socialbots themselves become ever more sophisticated. However, nearly all innovative technologies represent double-edged swords. While we should be mindful and wary of potential abuses of such technologies, we should also be open to the potential benefits that may arise through new opportunities promoted by socialbots.

ABOUT THE AUTHORS

Tim Hwang (@timhwang) is man-
aging partner of the Pacific Social
Architecting Corporation, a Bay
Area research and development
firm working on technology to
enable precise, large-scale auto-
mated social shaping.

Ian Pearce (@peeinears) is a researcher and developer specializing in applications at the intersection of social psychology, anthropology, and computing.

Max Nanis (@x0xMaximus) is a computational biologist whose research focuses on modeling protein interactions and macromolecular structure visualization.

Conclusion These articles represent a preliminary stage of research and development in social robotics. It is worthwhile to consider the future scale and application of socialbots as a more advanced technology. Insofar as even sim-

March + April 2012

ENDNOTES:

1. Wright, A. The social life of robots. Communications of the ACM 55, 2 (Feb. 2012), 19-21. 2. Lauricella, T. and McKay, P. Dow takes a harrowing 1,010.14-point trip. Wall Street Journal (May 7, 2010); http://online.wsj.com/article/SB10001424052 748704370704575227754131412596.html