If you ask Elon Musk, artificial intelligence (AI) should be feared. “I don’t think most people understand just how quickly machine intelligence is advancing,” Musk said on stage at Vanity Fair’s New Establishment Summit: The Age of Innovation. He also sponsors open AI, “a non-profit AI research company, discovering and enacting the path to safe artificial general intelligence.” Despite Musk’s misgivings, AI is infiltrating just about every corner of our lives and the digital content industry.
The AI Year in Review
2017 was the year AI became a household name—just ask Alexa, your new voice-activated assistant. “We saw enterprise use cases emerge in text analytics, natural language processing, voice recognition, image recognition, and, of course, structured data analysis in a variety of domains,” says Kashyap Kompella, a contributing analyst at Real Story Group. “I’d say in 2017 we reached the tipping or inflexion point when it comes to enterprise adoption of AI technologies.”
Why is AI exploding now? The answer isn’t simple. Or is it? Kompella says, “The terms ‘convolutional neural networks’ or ‘recurrent neural networks’ may sound intimidating, but deep learning is not. Just as the moniker ‘Big Data’ played its part in accelerating the adoption of a set of data-analysis techniques and technologies, the rebranding of neural networks and other machine-learning approaches as ‘deep learning’ is helping speed up enterprise adoption of AI technologies. That, and, of course, the availability of data and advances in computing techniques and processing capacity.”
Much of the conversation about AI in the digital media context has been around automating marketing functions, but it’s not the only application that we saw in 2017. Michelle Manafy is the editorial director at Digital Content Next (DCN), a trade organization dedicated to serving the needs of high-quality digital content companies. Manafy keeps a close eye on DCN’s member companies, and she saw plenty of AI-related action last year. “A number of publishers automated formulaic and high-volume writing using natural language processing,” she says. “The Associated Press is generating thousands of stories about U.S. corporate earnings and sports. The Washington Post is also using this kind of ‘robo-journalism’ to generate brief, timely news updates.”
Chatbots were also all the rage in 2017. Marketers love them, but so do publishers. For instance, Manafy says, “The Weather Channel uses its AI-powered chatbot to answer questions about the weather. It can also answer more complex questions like what to wear today—whether you are going running in your neighborhood or to a board meeting in another state.”
While Manafy isn’t exactly on board with Musk’s dire warnings about AI, she has seen some cause for concern in 2017. “An unfortunate trend I see in AI is that of algorithmic bias. Essentially, we flawed human beings either build our AI with our own biases, or our algorithms learn them from our collective behavior. This can cause far worse problems than filter bubbles. It is a common misconception that machines will be neutral. For better or worse, they are only as good as we make them.”
That being said, the worst fears of some haven’t been realized. “There’s a lot of fear about the jobs that AI will take away from journalists (and other professions). So far, though, AI is being used to augment media coverage and free up human reporters to take on more complex reporting and writing projects,” says Manafy.
A Look Ahead at AI
AI just came into its own in 2017. In many ways, we can expect to see familiar trends continue. Kompella says, “Because of deep learning, voice recognition is more accurate and powerful now. Expect to see voice interfaces [added] to more apps. …”
Manafy agrees: “I believe that voice-based interfaces offer an amazing opportunity to leverage AI. … I hold out hope that AI can also help detect fake news and identify other disturbing trends we see in terms of media manipulation online.”
Kompella points to a trend that may be completely new to some. “AI-inside—going forward, this is going to be a big trend. Be it devices or applications, they will have ‘smart’ or ‘cognitive’ capabilities. Some examples: think Amazon Echo, Apple EarPods, Google Pixel ?Phone. … On the enterprise side, Salesforce Einstein CRM brings some AI capabilities to its CRM [customer relationship management] apps.”
As hot as AI may be in 2018, publishers still seem to have a few blindspots. “AI is used by other industries for predictive analysis—and I don’t hear about media companies doing this,” says Manafy. “I hear about sentiment analysis, in which they are monitoring comments on social and traditional media for both their organization and that of brands they are working with. But predictive models can be used to target advertising, subscription, or membership offerings. The analytics identify consumer patterns and project the potential outcome. One of our members, the Financial Times, uses predictive analytics to correlate revenue to content usage and conversion rate to engagement.”
She also says personalization—the ultimate goal for any publisher—could use some help from AI. “Content recommendations are one of the low points of the web experience. I’d like to see publishers work on developing artificial intelligence that truly learns from consumers’ content consumption and provides them with real, valuable, substantive recommendations. This will not only increase time spent on your site, it would have the added benefit of improving your relationship with the reader.”
While Musk certainly sees a fantastic, and possibly frightening, future for AI, Kompella sees more mundane use cases in 2018. “The mind may conjure up spectacular examples, such as self-driving cars, when you think of AI. But in the digital content industry, the examples are going to be less spectacular,” he says. “In fact, in the content space, you could even say that we have not been using the term ‘AI’ per se, but been tinkering with intelligent search and discovery, advanced personalization, real-time recommendation engines, behavioral targeting, dynamic ads, marketing automation, and much more. The digital content industry is an early adopter in that sense.”
In other words, our experts think AI will mostly be used to enhance familiar experiences. It will be employed to make your life better in ways you may or may not notice.
* * *
Check out EContent's webinar on AI Driven Personalization for more information.