With just a few days left before the end of 2017, it’s time for communicators to finalize their plans for the new year and familiarize themselves with the trends and developments that will dominate 2018. In recent months, AI and automated technology have made a near-constant stream of news, often covering stories that have important implications for communications professions.
From algorithms that make decisions for entire cities to creative campaigns fueled by data mining, machines are infiltrating many aspects of everyday life and are reshaping roles previously held by humans.
Check out our selection of some of the best recent posts on the developments of AI, automation, and data.
- Wired correctly notes that AI cannot do everything yet, and in many cases companies turn certain tasks over to humans. Expensify is one of those companies, but late last year it faced a PR crisis when this fact came to light. Why the problem? Expensify was designed to reduce the burden of filing expense reports and “other benefits submission processes.” The problem is privacy: users were not aware that on occasion people would enter data rather than automated scanning.
- In the last several years, AI became part of the decision-making process in many facets of life, including healthcare, the justice system, and internet use. People are beginning to question, however, how AI makes decisions, as the algorithms that power it can be easily biased. A report from Quartz outlines the measures taken and organizations formed in 2017 to discuss and research ethical implications of AI and the possibility of regulating this technology.
- James Vacca, a third and final term New York City Council member, introduced a bill in August that, once Mayor de Blasio signs into law, will create a task force that examines the city’s decision-making algorithms. The bill, which regulates algorithms that affect the lives of New Yorkers in countless ways including detecting Medicaid fraud and assessing the performance of teachers, aims to make this technology more transparent, accountable, and understandable.
Social Media and Communications
- Reuters is using AI to scoop its competitors on news stories. How? By using Twitter as a type of “global sensor” of news stories as they are happening. Information culled from Twitter is captured and then machine learning is applied to sift through the data, categorize the topic, and then write a headline and summary.
- In the last year, entertainment companies have started using their users’ data in creative campaigns. Spotify recently launched its third billboard ad campaign that pairs cheeky taglines with bits of customer data, including one that says, “Deliver burns as well as the person who streamed ‘Bad Liar’ 86 times the day Sean Spicer resigned.” Similarly, Netflix recently tweeted, “To the 53 people who’ve watched A Christmas Prince every day for the past 18 days: Who hurt you?,” referring to a holiday movie the streaming service premiered last month. The New York Times explores how consumers respond to data mining consumer information for these messages.
- New Facebook regulations will work to combat posts the platform refers to as “engagement baiting.” These posts, which encourage users to take action such as liking or sharing the post if they meet the content’s criteria, game Facebook’s algorithms to gain more exposure. The new regulations, assisted by a machine learning model that reviewed hundreds of thousands of posts, aim to remove these posts from users’ News Feed and allow them to engage with more authentic posts.
- Algorithms are battling for your attention, and companies are betting those algorithms will do a better job of knowing what you want to read than you do yourself. Does that sound unlikely? Nieman Labs reports on this development, and highlights what it could mean for publishers, noting: “When a post hits a user’s feed because of the keywords it contained, rather than because the user ‘followed’ the publisher, building and retaining a loyal audience becomes even harder.”
- Storytelling is a key method for PR professionals to share messages that resonate with audiences. Emerging AI technology now collaborates with people to provide insights to optimize the emotional impact of and audience engagement with a story. McKinsey and MIT experts studied how this technology analyzes current videos to learn how to tell the most engaging story.
- Workers across various industries feel anxious about what the rise of AI and automation means for the future of the workforce. This Harvard Business Review post explains that as machines evolve to carry out more tasks that once required humans, employees and businesses must adapt to capabilities that are uniquely human, such as creativity, curiosity, and emotional intelligence. Machines will continue to improve efficiency, but for companies and workers to remain competitive, they must consider how human talent will be necessary for well-rounded growth.
- We are past the holiday season, but this story squeaks in under the wire because it is illustrative of the types of questions we are seeing as software and AI get smarter and more responsive. Hot new toys pop up every year, and at the end of 2017, The Atlantic told us a tale of how humans wrote programs to snatch up all of the good toys on retailer websites, and then turned around and resold them—with a considerable markup of course—to desperate parents.
- According to a recent survey, one in three organizations have AI fully implemented throughout their business, while another 20 percent have partial AI implementation. Forbes explains that with over half of businesses experiencing some AI within their business, more human intelligence is needed to run this technology.
- Have you ever given any thought to the thumbnails that appear on the home screen of your Netflix account? Netflix certainly has. Machine learning experts and researchers with Netflix explain how the company uses machine learning algorithms to personalize the artwork associated with each title that appears on the main page of users’ accounts. Netflix’s experts outline the process of personalizing the content for each user, providing useful insight on how communicators can employ similar strategies to customize their brand for each of their customers.
- How can machine learning be used right now? For anyone who has followed healthcare issues over the past few years, one of the most vexing problems is how to efficiently convert to electronic health records, as it can be laborious and error-ridden to convert patient conversations to electronic files after a patient consult. Google might have an answer to that problem, using natural language processing (NLP). A new research paper from Google testing a NLP application found that the system converted speech to text with a solidly acceptable word error rate of 20 percent—even in conversations using complex diagnostic terms and drug names.
AI is Finding Its Voice
- Google has released a voice-generating AI that is now “indistinguishable” from humans. Quartz examines the advancement in text-to-speech that mimics human voice patterns, including details like stressing capitalized words and changing enunciation based on punctuation.
- In both 2016 and 2017, home voice-activated AI “assistants” like Amazon’s Alexa and Google Home were a hot gift item. Pew takes a look at breakdowns of survey data showing us who uses these voice assistants, what devices they use them on (mostly smartphones), and why they use them (mostly to keep their hands free). The technology is primarily used by those age 49 and younger, and of those who choose not to use the tech at all, most say they “just aren’t interested” in using voice assistants.