The term “avatar” derives from the Sanskrit word avatara, meaning “incarnation” and has been used to describe a deliberate descent of a higher being into mortal realms for special purposes. Avatar also has a long history of use in Hindu texts to characterize the human appearance on earth of various Hindu gods such as Vishnu, Shiva, and Ganesha. More recently, avatar has become a popular term in new media technology to describe the digital representation, human or otherwise, of a real person in a virtual world. This use of the term emerged initially as a text construct in multiple user dungeons (MUDs), text-based social gathering points and chatrooms that became popular in the early days of the Internet, and then later on to describe a two-dimensional icon used on Internet forums (Damer 1997; Fink 1999; Blackwood 2006) and other digital communities or the form of a three-dimensional human model used in computer games (Lessig 2000). For example, a person can use an avatar to represent himself or herself in a meeting held in an online virtual environment such as Second Life or similarly as a tutor in the course of conducting a distance learning activity.
Avatars can be designed to suit the “owner’s” choice of how they want to be represented, whether it be a realistic representation of their actual appearance, an idealized form, or even a surreal character or animal. They are operated in digital worlds typically using mouse and keyboard interfaces and are akin to a puppet controlled by a puppeteer. Users can interact through their avatar with other avatars that also represent other live users who are “present” in a virtual space. The most common mode of communication between avatars is text-based, whereby the user can communicate in shared digital worlds by typed speech that appears in close proximity to their avatar for others to read. In these interactions vocal prosody and body/face gesture-based expressions of emotion are typically limited, so a variety of text codes have evolved to serve this function. However, some digital environments can allow for basic keyboard control of a limited set of “universal” body and facial actions to serve the purpose of expressing emotion and for the enhancement of believability.
By contrast, an agent is a more complex software entity that “possesses” or retains a certain level of autonomy or even some level of artificial intelligence. In real life, an agent can be a representative who acts on behalf of other persons or organizations for a variety of purposes (e.g., publicity, real estate, FBI, etc.). In digital technology, an agent is often referred to as a software agent, which is an abstraction or a logical model that describes software that acts for a user or other program in a relationship of agency (Franklin & Graesser 1996). Software agents can be autonomous, exhibit artificial intelligence, such as learning and reasoning, and may communicate and cooperate to achieve an objective. Viewed in this fashion, an autonomous agent is a system situated within (and as a part of ) an environment that senses that environment and acts on it, over time, in pursuit of its own agenda so as to affect what it senses in the future (Franklin & Graesser 1996). Such agents are self-contained and capable of making independent decisions and taking actions to satisfy “internal” goals based upon their software and the sensed environment.
An intelligent agent is a software agent that exhibits some form of artificial intelligence that assists the user and acts on their behalf. In some literature, intelligent agents are also referred to as autonomous intelligent agents, which means they can act independently, and will learn and adapt to changing circumstances. With recent advances in digital technology, intelligent agents have become more common in everyday life. For example, buyer agents from shopping websites can explore a network (e.g., the Internet) to retrieve information from various venders about a product that a customer wants to purchase. These agents, also known as “shopping bots,” work very efficiently for commodity products such as CDs, books, and electronic components. In addition to finding a product at the lowest price, buyer agents can also offer users a list of products that the user might like to buy on the basis of what he or she is buying now and has bought in the past. Amazon.com is a good example of a site that employs this type of shopping bot.
In recent years, there has been growing interest in a particular form of intelligent autonomous agent – one that looks and acts like a human being. These software entities, often called virtual humans, can be designed to engage in conversation and participate in collaborative tasks while inhabiting a simulated environment. Such virtual humans have been created for use in a variety of applications, including primary education (Biswas et al. 2005), training (Swartout et al. 2006), and therapy (Marsella et al. 2003), as well as in marketing and entertainment. With continuing advances in virtual human technology, one could speculate on a wide variety of areas where their impact could be felt. History students could visit simulations of ancient Greece and debate Aristotle. Persons with social phobia could practice interpersonal interaction skills in the safety of their homes. Social psychologists could study theories of communication by systematically modifying a virtual human’s verbal and nonverbal behavior as they interact with other virtual humans. Virtual human teachers could provide engaging and tireless educational activities that apply pedagogical best practices anytime, anywhere, and at any pace that learners require.
However, the construction of useful and believable virtual humans is still a challenging task and requires multidisciplinary effort. Since a virtual human agent is designed to emulate a real human, people have high consensual expectations as to how it should look and act, and will be disturbed by or misinterpret even minor discrepancies from human norms. Virtual humans must act and react in their simulated environment, drawing on the computer science domains of automated reasoning and planning. To hold a conversation, they must exploit the full gamut of natural language research. To create virtual human bodies that can be controlled in real time requires significant computer graphics and animation knowledge. And designers of embodied virtual human agents must draw heavily on the disciplines of psychology and communication to create entities that can effectively convey nonverbal behavior, emotion, and personality. In this regard, the creation of a virtual human is still a daunting task, but research is progressing on multiple fronts. Although Asimov’s benevolent robots are still a distant fantasy, realistic interactive virtual humans will almost certainly populate our near future, guiding us toward new opportunities to learn, enjoy, and consume.
- Biswas, G., Leelawong, K., Schwartz, D., Vye, N., & the Teachable Agents Group at Vanderbilt (2005). Learning by teaching: A new agent paradigm for educational software. In Applied Artificial Intelligence, 19(3/4), 393 – 412.
- Blackwood, K. (2006). Casino gambling for dummies. Hoboken, NJ: Wiley, p. 284.
- Damer, B. (1997). Avatars: Exploring and building virtual worlds on the Internet. Berkeley, CA: Peachpit.
- Fink, J. (1999). Cyberseduction: Reality in the age of psychotechnology. Amherst, NY: Prometheus.
- Franklin, S., & Graesser, A. (1996). Is it an agent, or just a program? A taxonomy for autonomous agents. In Proceedings of the third international workshop on agent theories, architectures, and languages. Heidelberg: Springer.
- Lessig, L. (2000). Code and other laws of cyberspace. New York: Basic Books.
- Marsella, S., Johnson, W. L., & LaBore, C. (2003). Interactive pedagogical drama for health interventions. In U. Hoppe, F. Verdejo, & J. Kay (eds.), Artificial intelligence in education: Shaping the future of learning through intelligent technologies. Amsterdam: IOS.
- Swartout, W., Gratch, J., Hill, R. et al. (2006). Toward virtual humans. AI Magazine, 27(2), 96 –108.
Back to Exposure to Communication Content.