Monday, November 9, 2015

Welcome to the age of Robotic Reporters.

“August CPI a new high in the past 12 months.”
Is there anything special about the above line? It is most likely that you will say ‘no’. However, will you consider it a bit more special when you come to know that this is headline of a news story written by a robot?
Throughout its history, Journalism has been heavily influenced by technological innovations. Positive impacts of technology include increase in speed of production, information delivery, reduced cost, improved accuracy etc. However each technological improvement has brought about massive changes in news rooms and the lives of journalists. “The printing press put a generation of scribes out of a job, the telegraph sent couriers scurrying to find new employment” (American Journalistic Review, 2013). The latest innovation, that is all set to revolutionize the media landscape, are algorithms and machines with artificial intelligence. The question is simple “Could Robots be the Journalists of the Future?” (Guardian, 2013). Can thinking machines truly take the place of professional journalists? Does the answer lie in the term “professional” in the question itself? In her article Zelizer (1993) writes “seeing journalism as a profession... may restrict our understanding of journalistic practice, causing us to examine only those dimensions of journalism emphasized by the frame through which we have chosen to view them.” Through this research paper I will try to understand the proper frame to examine journalism and answer the question “Can an Algorithm Write a Better News Story than a Human Reporter?”(Wired)
We will use Humprey’s Marketing SWOT Analysis (Strength, Weakness, Opportunity & Threat) to evaluate the relative position of the new entrant, ROBOTS, in the news making business. This in turn will help us analyze as to whether or not human journalists should feel threatened by their robot-colleagues.
Strength:       The most evident strength of robot journalist is definitely its computational prowess. According to its manufacturers as well as users robots have proved to be effective in mining huge quantities of data, pinpointing trends and patterns in data. “Automated news may lack the stylistic flavor of the best news writing, but it substitutes a meticulous commitment to factuality possible precisely because it is not human”(Carlson, 2015). However, this term ‘factuality brings forth with it a new range of problems.
Facts, according to Tuchman (1978), are “the pertinent information gathered by professionally validated methods specifying the relationship between what is known and how it is known.” While it is possible for Robots to dig out trends and patterns of preexisting data, can it ever find facts on field. Can artificial intelligence ever execute the human elements of on ground journalism? Waite (2012) writes “journalism … reflects.. Human beings: flawed, complicated, emotional. It is precisely the things that can’t be defined in a programming language that makes us human.” A critic further lashes out Canwest News Service (2011) by saying “Journalism don’t just bundle data together. They bring human presence to reporting… like imagination, humor and critical thinking.”

Another strength of robot journalism is the claim that it “removes human biases” while generating meaning from complex data (Carlson, 2015). However this very claim becomes a bit foggy when we analyze the role of the “meta journalists” or “meta writers as visualized by Narrative Science. Automated journalism can never be fully automated till the time the robots can themselves go and cover events. Till that time the frames and data will still be fed in by the journalists. In that case the lack of “human bias” is not possible.
Another key strength of robot journalism, and perhaps a one that cannot be debated against, is its speed and cost efficiency.
Weakness:     Compositional Form. Kafka (2012) writes “the result isn’t elegant but it’s perfectly readable.” While manufacturers site the experiments conducted on readers where they could not distinguish between robot writing and human writing, journalists, journalists choose to see the lack of human insights and perspectives in the articles.
Opportunity:             Cost-Cutting objective of media firms: Given the writing speed and cost of operations, robot journalism may be much cheaper to companies in long term.
Formulaic nature of conventional news: According to experts Journalists follow preexisting templates to fit facts to story forms (Darnton, 1975). This in turn helps the robot journalists get disguised into the already formulaic content.
Threat:                       Dependence on humans for data, Set up costs.
So, asking the question again, is this the end of human journalism?
As clearly seen through the SWOT analysis, robot journalism has clear strengths but humans still have the talent edge.  At the moment, the software’s algorithms mine through data and find facts and trends, and combine them with historical data and other contextual information to form narrative sentences.
Journalists, especially in India, do not seem to be too perturbed by the technology being competitor. An article in a major Indian daily newspaper, Hindu, stated “some kinds of journalism jobs will be more vulnerable than others in the coming years. At the more secure end of the spectrum will be on-the-ground reporting. It’s hard to imagine interviewing and reporting to be anything but a human activity, especially involving those aspects of life that aren’t easily amenable to the form of structured data.” (Hindu, 2015)
However journalists should not be too complacent. Technology is advancing further more rapidly. The softwares by Narrative Science and Automated Insights, for instance, can quickly spot patterns and zero in on what matters. With time it may start to mimic human writing and get better at it. It is essential that human journalists always stay one step ahead. Perhaps one way of doing it is by framing, as suggested by Barbie Zelizer, journalism as an Interpretive Community. Journalists may “establish themselves as qualified to discuss a certain issue by … interpretation from a localized, particularistic viewpoint” (Zelizer, 1993). The article further goes on to explain the importance and use of journalistic discourses to generate meaning to their work. Such discourses and on ground evaluation can’t at present be done by their robot contemporaries.
 “The moot point is journalists have to constantly ask themselves: what are they bringing to the table?” (Srinivasan, 2014).

No comments: