“It is impossible that AI is going to challenge me, as it is not able to get the emotional connection and it might always repeat itself as it can’t understand the beauty of language.” If you still hold this opinion in your mind, you have to be careful of being dropped behind by this era.
What you might not realise is that media companies have already used artificial intelligence in many different ways. The Washington Post, for example, has been using Bot Heliograf, its artificial intelligence technology, to analyse data, record trends and then enrich templates in order to create short stories. Other media outlets like Associated Press, CNBC and Al Jazeera are also increasingly using AI while producing news.
Surprisingly, the history of automated storytelling is more than 40 years. The first story written by the algorithm was created by Yale University, according to Tow Center for Digital Journalism. Since the 1960s, automatic text summaries have already been applied on weather forecasts, and later, in 1990s, it has been used on sports, financial and medical reports as well.
(Automated forecast for 6 March 1970)
The algorithm, which also is mentioned as automated journalism, or robot journalism (sort of misleadingly), can be used for news stories when clear, organised and reliable data are obtainable. The algorithm is capable of creating content on a large scale and even with fewer errors than human journalists. The articles written by it can be personalised to meet various demands of an individual reader in a quicker cheaper way.
So, how exactly automated journalism works?
It can roughly be split into five procedure, from collecting data to publishing stories.
In this process, the software is based on a set of predesigned rules according to particular problems at hand. It is a cooperation work from engineers, journalists and computer linguists. Moreover, domain experts are going to define the criteria of newsworthiness, depending on the evens which the algorithm has found, and ranks them by importance. Lastly, computer linguists will use sample texts to identify basic semantic logic and then translate them into system langue which the computer is able to understand and construct sentences.
It is said that once the algorithm is developed, each step of the news production process will be automated. It has shaken journalism industries for its ability to write mountains of news stories in a short time. Maybe you should start worrying about automated journalists might gradually replace human beings in this area.
What can be confirmed now is that the algorithm is still in its learning phase right now. There are still lots of limitations in using AI in a newsroom.
“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,”
said Lisa Gibbs, the director of news partnerships for The AP.