Friday, May 4, 2012

Ayayay! AI/HCI and II/PSI/PSR

This is post is actually a WIP, but I wanted to get some thoughts down regarding II/PSI/PSR and AI/HCI before they evaporate from my mind. So if you see the untied-up ends and other bizarrities, fear not -- they are just ideas for now.


We are witnessing “one of the most profound changes in communication since the invention of cuneiform and the clay tablet -- the technology which first made it possible to imprint symbols of human vocal messages and to transport them physically from sender to receiver” (Cathcart & Gumpert, 1986). This change in communication is not gradual -- it is exponential. As we race toward the Technological Singularity, the term “mass communication” has become a hangover buzzword. Mass Communication no longer exists under the traditional definition except during major events such as 9/11 and the Superbowl. Media effects have become much more audience-member-specific and no longer can be assessed in terms of effects on the masses as easily as they were when we had fewer media options. Our interactions with mediated people and events are primarily imaginative and deeply personal. It is difficult to assume that a message will have a uniform effect on the aggregate. For these reasons, “mediated communication” may better describe our individual interactions with messages coming to us through traditional mass-media.
This can be explained as a result of a larger variety of media outlets and the human tendency toward selectivity. As for the role of selectivity, research has uncovered many motivations: reinforcing beliefs, especially voting intentions (Lazarsfeld, Berelson, & Gaudet, 1968), alleviating cognitive dissonance (Festinger, 1957), mood management through TV viewing choices (Bryant & Zillmann, 1984), various political orientations such as knowledge and partisanship (Johnson, Bichard, & Zhang, 2009) to name but a few. Work by Freedman and Sears (1965) suggested other plausible explanations for selective exposure phenomena, namely information utility and contextual/lifestyle circumstances that predispose people to exposure to some types of messages. For example, smokers tend to see more cigarette ads than non-smokers -- not because they were looking to advertisements to alleviate dissonance regarding their habit, but because points of cigarette purchases often have advertisements for cigarettes. Our lifestyles invariably lead to exposure to information that supports them. As a result of Freedman & Sears’ (1965) work, the theory’s development in regard to mass media was temporarily slowed. However, cable diffused in the 1980s followed by the Internet in the 1990s, effectively creating an explosion of media choices. As Potter (2004) tells us,
We live in an environment that is far different than any environment humans have ever experienced, and the environment changes at an ever-increasing pace. This is due to the accelerating generation of information, the sharing of that information through an increasing number of media channels, and the heavy traffic of media vehicles traversing those channels. Messages are being delivered to everyone, everywhere, constantly, We are all saturated with information (p.7).
Selective exposure again became a theory of interest for explaining media choices in the early part of the twentieth century.
Adding to the complexity of the study of mediated communication is the existence of artificial intelligence (AI). Once a sci-fi staple made famous by HAL 9000, C3PO, and even Arnold Schwarzenegger, it is now something we deal with every day. Increasingly AI is becoming the go-between, inhabiting the world between us and the information we seek. A simple Google search puts AI to service. More importantly, it is becoming “smarter” and better-able to respond to us. As authors such as Ray Kurzwweil (FIND A CITE) argue, it is becoming smarter and faster exponentially.
Several authors have argued that search engines not only alters brain function (Small, Moody, Siddarth, & Bookheimer, 2009), but might make us stupid (Carr, 2008)! Others argue human, rather than technological determinism, and claim the Internet exists as a result of who we are. “Is stupid making us Google?” (Bowman, 2008); Is the Internet just another product of larger currents in civilization, such as a human tendency toward formation of networks (Castells, 2004)? Considering media dependency theory (Ball-Rokeach & Defleur, 1976) it is likely that as we continue to adjust to the company of AI we will grow more dependent upon it to accomplish many common tasks.
As we become more dependent on AI, it becomes even more necessary to develop a framework for understanding how humans communicate with it. Somewhere in the middle ground between psychology, interpersonal communication, and mass communication is the study of parasociability. Originally, parasocial research dealt with exploring our relationships with television personalities (personae), and parasocial interaction and relationships were defined as “one-sided.” Through imaginative processes we tend to make parasocial relationships two-sided and use them for such purposes as gaining self-understanding and keeping up with the personae, much as we do in our imagined interaction with real people (Madison & Porter, 2012). Parasocial interaction and relationships were originally described as one-sided interaction (Horton & Wohl, 1956). An on-screen personae communicates a message or behaves in a certain way, and the audience responds in a manner that suggests we are responding to a real person. Such interaction might be yelling at football players on television, recoiling in disgust at a horror movie, or even laughing in response to a joke. This demonstrates that a viewer, by watching television, plays a role that is different from other roles they may play in other areas of life.
Technology has changed parasociability by opening up relationships with mediated personae and delivering to us countless interactivity options. With the advent of social media, parasociability is better-described as two-sided. Fans of actors, athletes, and politicians have the opportunity to actually receive responses from personae. With the advancement of video game complexity and narratives, parasocial interaction is clearly two-sided. Video game designers purposely place parasocial cues in video games. A narrator may, as Horton & Wohl (1956) described, coach a viewer through a relationship using gestures, body language, and carefully-chosen scripts. This trend in video gaming is likely to continue (Bopp, 2006).
Most people are not yet ready to accept AI as an actual human being, but we certainly have 2-sided interaction with everything we do with AI. Many representations of NPCs in video games might be described in the same way as other fictionalized characters: more like “good neighbors” than close friends of relatives (De Backer, Nelissen, Vyncke, Braeckman, & McAndrew, 2007).
The study of parasociability provides a useful basis for understanding our relationships with AI. It’s not quite a person, but we tend to interact with it as if it were a real person. Imagined interaction theory, which has largely been studied through a functionalist perspective (i.e. Honeycutt, 2003, 2010) and has specific pertinence to the study of human-AI-Interaction.
Before exploring AI through a parasocial lens, it is necessary to define it. What is it? Is it a practical tool? Is it a personality? Is it just a set of algorithms that take code, process it, and create new code? Are we ourselves anything more than just a set of algorithms that take code, process it, and create new code – with physical features?
A question that mass communication researchers should be asking is “How does AI as a middleman change the influence of information – journalism, public relations, advertising, political communication – on individuals?”
Arguably, using parasociability to negotiate with AI began with video games. I myself recall attributing human characteristics to various human, alien, animal, and robotic representations of opponents (Non-Playing Characters; NPCs) against whom I battled, mesmerized for hours on end. The Atari 2600 and original Nintendo brought these representations into my home and as a child and later a teenager found myself learning to negotiate with (read: destroy) them. Several joysticks met their fates as I slammed them down in frustrated parasocial interactions.
Video games taught me about personality. Just as different people move in different ways, so too do video game opponents. The Pitfall alligators were harmless when their mouths were closed. When their mouths were open they had a tendency to devour the representation of myself in a single bite. Living in Louisiana, I quickly learned that alligators are not completely harmless, but they are certainly less dangerous when their mouths are closed. Later as an adolescent I learned how not to impress women as I explored the landscape of seedy characters in Lost Wages , a.k.a. the Land of the Lounge Lizards, and home to “Leisure Suit” Larry Laffer. One does not approach a woman with the words “Hey baby” and playing the role of Larry showed me exactly why. Plus, the words “hey baby” could lead to contracting a veneral disease.

No comments:

Post a Comment