Computer directed reporting are robo journalists becoming normalized

IMAGE CREDIT:

iStock

Computer-directed reporting: Are robo-journalists becoming normalized?

Media companies increasingly rely on artificial intelligence and machine learning to automate content production.

Author:
Author name
Quantumrun Foresight
January 18, 2023

With computer-directed reporting, many tasks traditionally done by humans, like curating a feed or writing basic news stories, can now be automated with the help of natural language processing (NLP) algorithms and artificial intelligence (AI). As these tools continue to develop, they will allow newsrooms to retrain staff or cut labor costs further.

Computer-directed reporting context

Bots are increasingly being used to automate news production and distribution. For example, within minutes of an earthquake striking California, robot reporter Quakebot can write an article and send it to a human editor at the Los Angeles Times, who then decides whether to publish the story. Sweden-based MittMedia increased real estate coverage from two articles per month to 2,000, all driven by a reporting robot. This surge in page views and subscribers can be attributed to the rise in content volume.

Algorithms, either directly programmed or trained on data using machine learning, are becoming more complex than ever. They can transcribe interviews rapidly, write news reports, investigate leads, and even curate stories to keep readers engaged. The most recent shift has seen AI tools becoming more visible in reporting and investigating roles.

Although it appears that most people don’t care if their reporters are human or robots, experts warn that as algorithms do more journalistic work, newsrooms must develop a stronger understanding of the technology’s advantages and disadvantages. A Reuters Institute survey of 52 countries found that 40 percent of news leaders believe robo-journalism, where AI writes stories automatically, will be an important industry trend by 2022.

Disruptive impact

The increasing adoption of NLP in journalism can go two ways: focusing on higher-value editorial content or aggressively automating nearly all processes to streamline operations and save labor costs. An example of the latter is UK-based Radar, the world’s only fully automated local news agency launched in 2018. The news company was supported by Google’s Digital News Innovation Fund.

The firm has a team of five journalists who have filed over 400,000 articles since Radar switched to a subscription model. What makes the agency interesting is that it uses publicly available datasets and feeds them into templated algorithms to create different news stories with various angles using these statistics. The idea is not to focus on investigative journalism but on creating daily stories based on numbers.

This business model might get more traction as more companies realize the value of creating custom AI algorithms to scale up content production. However, news organizations face an ethical dilemma when adapting AI systems for journalistic purposes. For example, how can editors detect bias in an algorithm? How can a natural language generator (NLG) error be fixed? Is AI accurate enough to know what information is important and new? Should AI be responsible for choosing what content is consumed by millions of people?

Answering these questions is critical because technology companies, not journalists, are responsible for much of the innovation in this area. Tech firm turned media company Knowhere News, for example, understands that its ML systems can only collect information from existing sources, such as press releases and social media posts. Thus, the points of view are limited or inherently biased, and objectivity is compromised.

Implications of computer-directed reporting

Wider implications of computer-directed reporting may include:

More startups providing AI solutions to newsrooms, including building landing pages and installing paywalls where appropriate. Reducing the cost of these services may make independent, local, or niche news and journalism financially viable again.

Major news corporations developing their automated systems and leasing the service to smaller news sites.

A deepening paradigm where content and news is real-time, plentiful, free, and covered from every angle. Content consumers will be overwhelmed with choice and depend on curated feeds for their news.

Investigative journalism will grow in importance and financial value as all other forms of news content will be commoditized. Real journalism may see a renaissance.

Increasing mistrust from readers who prefer traditional journalism.

Threat actors injecting disinformation and propaganda campaigns into algorithms without editors knowing about it.

Increasing public demand (or legislation) for news organizations to be upfront on how they use AI/ML in their systems and to give AI journalists clear bylines.