I was halfway to a restaurant I was meeting a friend at when my car’s directional wouldn’t turn off. As I walked into the restaurant a few moments later, I was holding down my phone’s home button.
“Call Dad,” I told my phone.
Siri confirmed my command and seconds later, it was ringing my Dad/on-call mechanic.
A few months before that, I was having a dispute with a friend about how many Oscars a specific movie had won.
“Okay, Google,” my friend called to their phone across the room, preparing to settle the argument without leaving the couch.
This ability to glean information with a single click—or less—and a simple voice command is just one example of the abilities of artificial intelligence (AI) and automated technology. In recent years, machines have become increasingly capable of performing tasks once reserved for humans.
These advances allow industries to integrate automated, artificially intelligent technology into their work to enhance efficiency and accuracy. Medicine, journalism, and PR have seen advances in their industries as a result of AI and automated technology. These advances, in turn, have shaped the roles of human employees in specific and important ways.
A recent New Yorker article, written by Siddhartha Mukherjee, discussed how AI and machine learning is shaping the diagnostic process in the medical field.
Mukherjee writes of the four-step diagnostic process he learned during his clinical rotations in the 1990’s, which involved examining a patient’s history and conducting a physical exam, building a list of potential causes, administering preliminary tests and questions, and ordering definitive tests to confirm a diagnosis.
In January 2015, Sebastian Thrun, a computer scientist, became interested in diagnosing cancer before symptoms became apparent, as diagnosis after the appearance of symptoms often means the cancer has progressed to a more aggressive, serious stage. Thrun wondered if artificially intelligent technology could assist with this early-stage diagnosis.
Computer-aided diagnosis had become common in some medical tests, such as mammography, but these machines do not become smarter after looking at more scans. Thrun aimed to build software that would replace outdated rule-based technology with technology that could learn from each subsequent photo submitted for evaluation and diagnosis.
Thrun developed software that was fed images of cancerous skin growths alongside benign ailments, such as acne and rashes. The machine was able to correctly identify cancer 72 percent of the time, six percent higher than the average of two dermatologists who were tested alongside the software. It was Thrun’s hope that this software could develop to accept photo submissions of skin growths and produce a diagnosis for the average patient.
When asked by Mukherjee about how this technology could affect human diagnosticians, Thrun stated that he was interested in amplifying and aiding, not replacing, human intervention.
During the 2016 Rio Olympics, the Washington Post introduced an automated bot, Heliograf, to write tweets about the outcome of the game’s events.
Heliograf went on to cover the US election, writing articles with such accuracy, the content was generally undecipherable from writing produced by humans. It works by being plugged into a data source, VoteSmart.org in the case of the election. Once it has identified relevant data, it implements information into phrases from templates created by editors. Additionally, it can send tips to reporters to alert them of discrepancies that may be worthy of human investigation.
According to an article from Wired, Joe Keohane states that this technology could be useful for two reasons. First, it would grow the Post’s audience by churning out many articles on specific, niche topics for a small but far-reaching audience, without expending a lot of human labor.
Their second goal involved increasing newsroom productivity. This technology frees valuable, experienced reporters from spending time on simple pieces that can be completed by an automated template. In November 2016, for example, Heliograf produced 500 articles on various local and national races. Rather than spending time on a multitude of basic articles, reporters have time to write pieces that require critical analysis and human consideration.
In an era of ever-disappearing writing and reporting jobs, journalists would be justified in their skepticism of an automated tool that can produce work as effectively as a human. Keohane states, however, that Shailesh Prakashthe the Post’s CIO and VP of digital product development, and Jeremy Gilbert, the Post’s director of strategic initiatives, “take pains to stress that the system is not here to usher reporters into obsolescence.”
Instead, as the goals above reference, it has been implemented to free reporters from time-consuming, mundane writing, thereby allowing them to devote their time to more in-depth, thoughtful work.
Why does all of this matter to PR?
The above examples, specifically the one about medical automation, do not immediately appear related to PR. Although PR professionals rely on media and journalism for numerous PR functions, it doesn’t necessarily matter to industry professionals who is doing the writing, as long as media coverage is produced.
This trend of automation is both relevant and important to PR. Not unlike industries such as medicine and journalism, automation is reshaping and changing tasks integral to PR.
Automated tools like Heliograf may enhance relationships between PR professionals and reporters. Shrinking newsrooms had made successfully pitching stories to journalists difficult, as the small number of employed writers were being inundated with an avalanche of pitches, but with their increased free time, facilitated by automated writing, PR practitioners may find reporters more available, leading to stronger relationships.
Media monitoring, for example, began as a slow-moving process that required humans to clip relevant news coverage and create a book of the collected clips. The process evolved into automated software that scours online, social, and traditional channels and delivers results almost instantly.
Similarly, PR measurement tools have evolved into automated processes that collect data and place it in reports, which allows professionals to understand performance and perception of their organization.
Although the influx of automation is apparent, this technology isn’t rendering traditional PR professionals irrelevant. Even with the increased possibilities of automation and AI, PR professionals are still needed to analyze the information gathered from automated tools. The breadth of data collected by these tools offer no value to organizations without human analysis conducted to review trends, sentiment, and patterns.
Additionally, the PR industry is based on creating relationships between organizations and their intended audiences. Even the most advanced technology cannot connect with humans as effectively as another human. The PR industry needs professionals to produce thoughtful, engaging campaigns to ensure the success of their organization.
Despite the apparent strides in automated technology and artificial intelligence, human involvement remains crucial. As the technology continues to evolve, humans will still be necessary to produce quality analysis, as well as work that resonates with people.