The night UK Prime Minister Boris Johnson claimed a landslide victory by his Conservative political party, fellow Britons and anyone else around the globe with access to BBC News witnessed a real-world example of how and why Artificial Intelligence (AI) becomes even more useful when combined with Human Intelligence (HI).
The night UK Prime Minister Boris Johnson claimed a landslide victory by his Conservative political party, fellow Britons and anyone else around the globe with access to BBC News witnessed a real-world example of how and why Artificial Intelligence (AI) becomes even more useful when combined with Human Intelligence (HI).
As the leader of the company harnessing the superior intelligence that comes from the blending of HI with AI to help businesses help their employees perform their best, I found this effort by the fabled English journalism institution makes the case for HI+AI brilliantly.
The BBC reported on results for every constituency, voting district, across the British Isles. Utilizing AI, it produced nearly 700 stories in all, including several done in Welsh. This momentous occasion in UK politics was also the biggest use of machine-generated journalism thus far attempted by the BBC. All articles reflected BBC style, making use of certain phrases or structure, all of which was programmed in beforehand.
Leveraging AI to report on what’s largely an aggregation and distillation of data makes sense. Machines, given the necessary data to look for and how best to organize it, can present the desired information in a comprehensible manner. Articles can be prepared quickly that include which candidate to be a Member of Parliament got how many votes, what turnout was like, vote shares by political party, changes in vote shares from a prior election.
Before you think that this application of AI spells the end of the need for journalists, consider that the head of the project suggested the tech used was designed to enhance the service provided to those who look to the BBC for the latest news — like election night results — not to replace humans. Robert McKenzie, editor of BBC News Labs, said, “This is about doing journalism that we cannot do with human beings at the moment. Using machine assistance, we generated a story for every single constituency that declared last night with the exception of the one that hasn’t finished counting yet. That would never have been possible [using humans].” He did note that journalists at BBC offices checked the articles before publication.
Further, he indicated that if the story is solely grounded in data, AI can work fine on its own. However, in those constituencies where analysis needed to be added to the articles, to look beyond the data, HI must come into play to add necessary context. None of the stories prepared using AI had any quotations or any analysis of what happened in voting or why. McKenzie notes, “that’s quite a downside in terms of quality journalism.” In effect, AI-based journalism can currently provide the who, what, and when of the craft but not the how and certainly not the why. That requires HI.
McKenzie said, “this clearly only works on stories that are grounded in data. It is not a technology that allows you to do any kind of analysis. None of the stories have any quotations in, none of them have any analysis of the event or its significance. It is purely a written version of what has happened based on the data
Other news organizations have made use of AI, but always with the aid of HI. The Associated Press looks at AI as a way to improve its services, from news gathering to production, and distribution. It’s made use of AI to report on quarterly earnings news from publicly-held companies for a few years. But like the election that brought Mr. Johnson a majority, those are largely data-centered stories.
As long as a human can see beyond the data and bring insight and analysis, as well as an innate sense of what will spark readers’ interest, adding HI to AI will be to everyone’s benefit be it’s involving election results or the performance of employees.