By Jenny Lee WIRED Korea
Working as a journalist many times involves sitting on the floor near a power outlet holding a laptop in his lap while he writes up stories.
For journalists, time is very valuable. They are constantly operating on tight deadlines all the while meeting the demands of journalism--the exigencies of production and the need for stories that are fair and accurate, yet compelling.
To ease their burden, artificial intelligence (AI) is being integrated into newsrooms more and more, not just for speeding up research and accumulating data, but also for converting data into narrative news text in real-time. Slowly but surely, it is transforming the industry and the future of journalism.
Last week, South Korea’s Yonhap News Agency published its first machine-written article on weather, the first time for a South Korean news outlet to pick up AI tools for content.
At the heart of its program, developed in partnership with game publisher NCSOFT Corp, are AI techniques, such as machine learning, which enables it to analyze huge data sets, recognize patterns and make predictions without needing explicit programming, and natural language processing (NLP), which helps it understand, assess and manipulate human language.
“The launch of this system is a first step toward creating an algorithm that produces completely automated articles and selects images that are meaningfully related,” said Choi Jonghyun, a professor at Gwangju Institute of Science and Technology (GIST).
Yonhap News Agency and NCSOFT, which has a division devoted to AI, struck a deal in 2018 to conduct joint research on AI technologies that could be applied to news services.
Major news outlets around the world have already adopted machine learning tools to improve efficiency of news-writing processes and slash costs.
While Cyborg helps Bloomberg by turning company earnings reports into news stories in a quick and efficient way, the Associated Press uses automation for coverage of 10,000 minor baseball leagues games annually, the Washington Post for elections and the Los Angeles Times for earthquakes and homicides. Forbes also uses Bertie, which provides to journalists first drafts and templates for news stories. And the list goes on.
Because of the formulaic nature of automation, robot reporters are most suited for topics like financial reports, sports recaps, weather, real estate analyses and earning reviews where numbers and numerical figures take precedence over sophisticated narration.
Although the media going fully automated won’t happen anytime soon, the availability of machine learning tools is spawning fears that these robots could strip journalists of jobs. But media organizations that are harnessing AI and its benefits say it is designed to complement journalists, not to replace them.
The scope of AI is limited to simple tasks as of now, said Choi of GIST.
Lisa Gibbs, the director of news partnerships for the Associated Press, was quoted by Forbes as saying, “The work of journalism is creative, it’s about curiosity, it’s about storytelling, it’s about digging and holding governments accountable, it’s critical thinking, it’s judgment — and that is where we want our journalists spending their energy.”
Her remark implies that the kind of work demanded of journalism cannot be done by AI-based algorithms, at least not in the near future.
저작권자 © WIRED Korea 무단전재 및 재배포 금지
- 최초작성 2020.05.04 16:11
- 수정 2020.05.06 09:51
저작권자 © WIRED Korea 무단전재 및 재배포 금지
이 기사를 공유합니다