Contentpepper® Blog

AI - benefits and risks for editorial work

2019 June 07 at 11:36AM

Once the stuff of dreams, it served as the far-fetched fodder for Hollywood and the Sci-Fi entertainment industry, and formed the basic premise for movie classics such as AI. Artificial intelligence, since become ever more real in our day-to-day life and work. AI has evolved so much in recent decades that it can now replicate many human perception and action processes. In an increasing variety of industries, human work processes are successfully aygmented, or even replaced by, machines and programs - even in journalism and editorial work. But what advantages does AI bring to an editorial team in the average working day? What are the possible risks? And what should editors and content managers look for when choosing to be for or against AI?

Notable in the last three years, has been the development and integration of AI systems with increasing success into editorial work. Bloomberg News, Forbes, The Times - large media companies have already incorporated AI into a variety of editorial tasks. From spelling checks to personal data analysis, from individual referrals to the reader to writing articles - the latest AI tools are transforming the way news is written and distributed.

The Bots behind financial reports

Bloomberg has pioneered the process, using a program called Cyborg to turn financial reports and numerical data into understandable news content. The reality is that increasing amounts of financial content are being created and published by bots. In addition to Bloomberg, the Washington Post uses its Robot Reporting Program Heliograph to analyze current financial trends and create short news articles. In the first year after Heliograph was launched, the automated writing system wrote, processed and published a total of 850 articles.

But now AI tools have penetrated the media world outside financial news, and they are using these to better meet the needs and individual wishes of their readers and end consumers. In the UK, the Times and the Sunday Times have developed an analysis service called James, which analyzes reading habits and provides individual recommendations to readers based on this kind of data.

AI – is this trend still the way to go?

Despite the proven usefulness of integrating AI into routine daily editorial work, many editors are still suspicious of this trend, not least for fear of job cuts and loss of quality. Proponents, however, are convinced of the opposite. Their view is that implementation of AI should not replace journalists and editorial staff, but give them more time and support. While they pass on data analysis or the writing of shorter, factual messages to bots, reporters grant themselves more time to focus on investigative work, accurate content and storytelling. In addition, Times, Bloomberg & Co. report on the very good quality of the automated writing and text generation systems. In addition to objectivity, correct sentence structure and terminology, the systems also score points in terms of varied vocabulary.

Nevertheless, creativity and the intellectual work are always in the hands of reporters and journalists. Even while AI is helpful in daily editorial work, its regulated work processes cannot be equated with that of reporters, and it lacks the capacity to address more complex topics. This is before one considers the technical and processing errors.

In addition to the aforementioned risks, other factors require consideration when making a decision for or against AI. On the one hand, content must be optimised – created in a way that makes it readable and editable by AI systems. That's in addition to the existing optimising content for SEO. On the other hand, editors rely on suitable databases, which the AI can search and filter, to extract suitable content.

Conclusion

Although Artificial Intelligence in recent years has made important and helpful developments in performing rudimentary editorial work, in summary, many technologies are still not living up to the expectations. While bots and automated writing and text generation systems can be helpful in working with factual and numerical data, they still lack the capacity to handle more complex topics. Nonetheless, AI tools can be beneficial when it comes to analyzing reading habits and developing personal trends. Smaller topics and articles can be authored and edited quickly and well, and news distribution enhanced far more widely. All in all, today's Artificial Intelligence can serve as a support tool to save editing time and enable the human editors to focus on important topics and relevant storytelling. However, deeper AI involvement in the editorial work is technically still considered too risky.

Want to see more? Find insights from the publishing industry in our free White paper „Publishing in the Change – Digitisation, Crossmedia and Technology in the publishing sector“.