23 Nov Journalism: AI for Publishing
Advances in AI and Deep Learning developments provide multiple opportunities in several sectors, including journalism and publishing.
Newspapers are using AI to automate many activities in the journalistic production chain: data collection and verification, production of stories and graphics, publication, automatic setting of appropriate tags for each article, etc.
The three key concepts
First of all, three different concepts must be explained: Automated Journalism, NLG Journalism (Natural Language Generation Journalism) and Artificial Intelligence applied to journalism.
Automated Journalism is a computer science specialization that aims at entirely replacing human journalists with algorithms that can take content and rewrite it according to criteria set by humans.
NLG Journalism, on the other hand, refers to that computer science functionality that aims to use language-generating algorithms to help some professional journalists generate content, increasing their communicative capabilities (e.g. through multi-language conceptual writing).
By contrast, Artificial Intelligence applied to journalism is a wider and all-encompassing concept. AI is a ‘Science‘ and thus has no limitation of purpose or specific domain when applied to another discipline of human activity. It is based on reasoning exercised by giants, such as Shannon or Turing, and has a very close bond with symbolic calculation and the Number Theory’s essence. Therefore, it also embraces the problems of automatic journalism and automatic language generation, but is not limited to these two aspects.
Artificial Intelligence and Journalists
Currently, articles written by AI are limited to relatively simple and stereotypical topics (stock market results, takeover announcements, etc.). When AI does not write articles itself, it can also help human journalists with jobs that are too intricate to handle, such as long-form articles, in-depth analyses and investigative journalism. Moreover, helping them allows them to focus on producing high value-added content.
One extremely valuable use case is the automated transcription of interviews, which can save human journalists untold work hours. Although the results of AI transcription are rarely flawless, the few mistakes the software makes can easily be corrected by a human editor.
Tailor-made content for users
AI not only influences the news that is written, but is also able to control the articles that users ‘browse’. In the same way as Facebook‘s algorithm works, AI technology can keep track of the articles subscribers visit on a news website, learning their behavior and preferences (e.g. how long they spend reading each article).
The more a news outlet knows about its customers, the more relevant content it can display on the website or in a weekly newsletter, providing a personalized experience that helps encourage user interaction and discourage attrition. According to a survey by digital media company Digiday, 70% of digital publishers affirm they personalize content for visitors.
Integrating Artificial Intelligence into the world of information is a crucial process, which could lead to the development of faster and more reliable journalism. We all make decisions based on the information we read, as ordinary citizens, as professionals and as companies. The faster we find this piece of information, the more time we will have to evaluate and mitigate our choices, especially in view of the risks that the next decade seems to bring.