The Computerized Automation of Speaking and Writing
Imagine reading an automated news story that was composed entirely by a computer. (It’s likely that you already have.) Or imagine yourself sitting on a couch, conversing with an artificially intelligent psychotherapist who interacts with you from a screen across the room.
While most people tend to think that automation affects only certain sectors of labor (especially work performed in blue-collar professions), the computerized automation of communication will have a serious impact on diverse fields. Communication scholar Joshua Reeves of Oregon State University ponders the social and political impact of this transformation. What happens, he asks, if people increasingly rely on automated machines to carry out the socially essential work of communicating with one another?
Reeves, who for several years has been studying the increase in the automation of military technology, has been noticing similar trends in the civilian world. One example of this trend in everyday life can be seen in toys such as Hello Barbie dolls, which record and save the conversations of millions of American families. Critics of the doll have concerns about privacy, and also are concerned that Hello Barbie, whose intelligence is based on a massive cloud of surveillance-based data, will absorb inappropriate language and conversational habits. But Reeves argues that when it comes to automation of communication, there are broader social, economic, and political concerns as well.
The economic consequences of automated communication are already affecting people who work in fields that rely heavily on communication, including psychotherapists, personal assistants, college advisers, telemarketers, life coaches, bank tellers, and even teachers and professors. In fact, Reeves says, most people have already been exposed to automated discourse when ordering fast food, learning the positions of political candidates, checking bank balances, or making doctor appointments.
Automation and Immaterial Labor
The prospect of automation is nothing new. In fact, medieval Muslim inventors such as Al-Jazari designed and built a number of machines that imitated animal and human behavior—hydro-powered peacocks, musicians who entertained royal guests with music, and servants who washed nobles’ hands and served them food. As time marched on, French inventor Jacques de Vaucanson invented automated silk looms for the French silk industry. When it became known that his loom could replace human laborers, the silk weavers destroyed the machine and chased him out of town. These French workers, Reeves says, recognized the threat that was raised by automated machine labor.
The threat is as real today as it was in the 18thcentury. “The widespread circulation of automatic communicating machines gradually reduces—in real terms—the opportunity and impulse for cooperative human struggle,” says Reeves. As machines develop competency in interpreting and producing discourse, they are gradually taking over many domains of social life in which communication is of utmost importance. In one example borrowed from Sherry Turkle’s 2013 presentation to the American Association for the Advancement of Science, a robotic baby seal is designed to function as a conversational companion for older adults who suffer from dementia or depression. But Reeves argues that this device is depriving younger people of the communicative act of listening to their elders. Reeves quotes Turkle: “We are building the machines that will literally let [the elderly’s] stories fall on deaf ears.”
“By idealizing the machine, people become more impatient with the flaws and contingencies of human relationships,” writes Reeves. But communicative labor relies on the productive, spontaneous surplus of human communication to generate dissent and creativity. According to Reeves, the socially essential work of human communication is being “drained of its spontaneity, contingency, and politically creative potential.” In an era of automated communicative labor, those uniquely human qualities are destined for elimination, he notes.
Automation of Written and Oral Discourse
While blue-collar workers have been suffering the brunt of automated labor for some time, people in other fields of work also should be concerned about their fate, says Reeves. He examines the threats to communicative workers such as journalists. “Robo-journalism” has become commonplace. In March 2014, when an earthquake hit southern California, theLos Angeles Times was able to use an algorithmic discourse generator called Quakebot to break the news. While some are not worried that robo-journalism will take over the field, others disagree. The company Narrative Science estimates that 90 percent of news stories will be bot-generated by 2030.
With respect to oral discourse, trends in automation can be broken down into two general areas: simple interactive voice response systems such as Apple’s Siri or Microsoft’s Cortana, and more complex agents such as natural language chatbots, which are sometimes complemented by robotic bodies or screen-based representations of speaking faces. One of the most interesting talking heads, Reeves notes, is EHeBby, which links facial expressions with spoken discourse and has the ability to detect emotion in the human voice interacting with it.
Reeves argues that as automated communication becomes more prevalent, people need to develop a stronger understanding of the challenges facing others in communication-oriented fields. While automation creates some opportunities to open doors to other forms of creative work, it also creates the potential for social alienation and labor displacement.