AI Goes from Disruptive to Imperative
/Publishers have looked on as machine learning technology has developed—but now it’s time to leap.
by Mike Groth
This article first appeared in the Publishers Weekly Frankfurt Show Daily on October 16, 2019.
Over the course of the last five years, AI, natural language processing (NLP), and machine learning (ML) have been much talked about, as well as trialed and tested, in the publishing industry. These technologies are often the focus of panel discussions at conferences such as this one, discussions that have illustrated how AI could be used for a variety of purposes: discovery, peer review, bestseller predictions, and, perhaps most importantly, improving publisher efficiency.
But, while a handful of publishers are experimenting with AI in various ways, and many are happy to hear about the benefits, the technology is still far from reaching a tipping point. And one of the first things publishers want to know when considering adopting AI-powered workflow solutions is exactly what can be gained—and how much can be saved—by throwing AI and other machine learning technologies into the mix of traditional editorial and production workflows.
What we at Cenveo are finding is that a task that used to take several days is now possible to complete within hours. Whether it involves automating a workflow to apply pre-set rules, fast-tracking high-quality manuscripts to the composition stage, or avoiding mandatory stops in different editorial departments—a 30%-40% reduction in the overall time to publication is possible. For tasks that tend to be the most labor-and time-intensive, such as copy editing, this could dramatically improve editors’ productivity.
And yet, as we have seen in questions asked during the webinars and panels Cenveo has recently hosted, there is still some uncertainty among publishers surrounding AI and NLP technology. Some are concerned that the quality of editorial work will be sacrificed in a quest for speed; others wonder how many jobs may be lost if the work is farmed out to computers; and some publishers are still not sure how the technology actually works.
But machine learning and AI are ready to jump from the theoretical to the actual, and to become real, transformative solutions for publishers. For example, today’s technology can track what actions copy editors and other professionals perform over time, analyze the aggregate data, and make predictions based on this human behavior. It can accurately automate the detection of mistakes in grammar, spelling, punctuation, word choice, and style, and recommend corrections based on the rules it has learned from real-time analysis of many users on a daily, or even hourly, basis. Most importantly, huge amounts of data can be processed in just a matter of seconds. And the more a machine-learning system is used, the better it gets.
For instance, Cenveo’s Smart Edit began with inputs from various sources—dictionaries, editor experiences, and specific publisher standards. But it can improve with further inputs from a range of academic, trade, and educational publishers, nonprofit organizations, and other professional publishing bodies using our editorial services. Such widespread and increasing usage means the system could learn about and apply commonalities across all user actions, resulting in higher-quality automation.
To date, many publishers remain curious about implementing AI tools on a significant scale. We get it. Change is always difficult. And common misconceptions persist around the technology’s perceived disruptive nature, along with a reluctance to shake up models and structures that have been in place for decades. As a result, the true potential of this innovative technology has yet to be unleashed and we are still in the testing-the-water phase.
But the time has arrived for publishers to cast away these misconceptions and to learn more about AI’s power to transform their day-to-day processes. In order for machine learning to realize its true potential, much greater adoption needs to take hold. And by understanding the significant benefits to publishers and editors, the closer we will get to a faster, high-quality publishing model.
Michael Groth is Director of Marketing for KnowledgeWorks Global Ltd. (formerly Cenveo Publisher Services). Email him at mike.groth@kwglobal.com. To read more in the PW Show Daily about Smart Suite, digital learning trends, and Cenveo at the 2019 Frankfurt Book Fair, click here.
