Hacks, some really bad news for you

 

The Washington Post has adopted Artificial Intelligence technology to report key information from the 2016 Rio Olympics at its live blog and Twitter account. The newspaper used its in-house software Heliograf which automatically generates short multi-sentence updates, including daily schedule of events, results and medal tallies etc. The procedure quickly creates simple but useful reports.

Jeremy Gilbert and Sam Han lead a team of three engineers working full time on developing Heliograf. They are supported by a group of product analysts and news room staffers for shaping the software. Gilbert is a former Northwestern University journalism professor associated with Narrative Science which experimented with machine-made news reporting.

“We’re not trying to replace reporters”, Gilbert says “we’re trying to free them up. In 2014, the sports staff spent countless hours manually publishing event results. Heliograf will free up Post reporters and editors for analysis, colour from the scene and real insight to stories in ways only they can”.

The Post plans to use Heliograf to cover the US presidential elections in November. Gilbert remarks that, during the 2012 elections, the newspaper’s coverage was up to 16 hours behind and this time the software can change the whole scene.

The use of Artificial Intelligence in journalistic practices was pioneered by, among others, Chicago-based Narrative Science followed by North Carolina-based Automated Insights. While companies like Forbes and Deloitte hire the services of Narrative Science, Associated Press has tied up with Automated Insights. Both of the agencies have developed tools to turn financial data into news stories

Although the use of Artificial Intelligence so far has been mostly confined to the processing of data into narratives, the promoters of such companies have predicted that 90 per cent of news stories will be produced by the use of Artificial Intelligence by 2030. The claim appears to be a little unrealistic at this point in time but if we look back to review the impact of earlier technology in the media - for example role of the Internet in transforming journalistic practices during the past three decades - one cannot completely ignore these prophesies.

In China, the Beijing Times experimented with producing stories for its news website through its software Dreamwriter during September 2014. It has been told that the software could churn out a 1,000 word piece in just one minute.

Besides the two American companies mentioned earlier, more than 10 other companies are engaged in automated content creation. They include five in Germany (AX Semantics, Text-On, 2txtNLG, Retreco and Textomatic), two in France (Syllabus and Labsense) and one each in the United Kingdom (Arria), China (Tencent) and Russia (Yandex).

The interesting aspect of such companies is that none of them considers themselves as journalistic organizations, nor are their products specifically geared towards providing journalistic solutions. They develop technology which can be applied to any industry such as manufacturing companies for product descriptions or patient summaries for hospitals.

Worldwide, efforts are being put into developing software solutions for generating automated news. Though automated journalism is mostly limited to basic sports reports and business news at the beginning, it is expected to enter other areas of social activity. It has the capacity to generate news stories for repetitive topics for which structured data is available.

Algorithms can generate clean and accurate copy faster on a larger scale. At the same time, they can produce stories in multiple languages and different angles, even answering users’ questions by creating a personalized environment. However, algorithms rely on data and assumptions which may be subjected to error and bias.

It is expected that the technique will bring whole new areas in building-up news stories which are being neglected by conventional journalism. Kris Hammoud, co-founder and chief scientist of Narrative Science, says, “Storytelling is one of the great gifts we have. When you go home, what is the question your wife or husband ask? ‘How did your day go?’ That’s not going to stop. But what the machine will be able to do is to communicate with us about a lot of other stuff that goes on in the world and explain how it relates to us.”

In India we may not expect Artificial Intelligence to enter journalism so soon; even in technology-rich countries, the practice is in the nascent stage. But, as I said earlier, if we review earlier trends such as the use of smart phones, Internet penetration and the success of application-based start-ups in our country, can we completely ignore the possibilities?

At the same time, when we recollect the experience of hand-drawn dummy sheets for page make-up during the hot-type era in the pre-emergency and manual placement of matter during the introduction of the cold-type wave in the post-emergency era, nobody thought about the availability of thousands of templates for page layout in the computer screen at a click of the mouse which is a simple practice even in the printing plants located in the remotest corners of the country today.

Leading economist Raghuram Rajan has attributed the populist anti-immigration upsurge in the US and Great Britain to the loss of middle-class jobs to technology. During the second week of June, he remarked, “People at the upper end will do well in an economy that supports innovation. People at the lower end will manage because the jobs of the gardener and security guard will still be around, but not the middle class jobs of clerks and factory workers”.

When we apply the logic to journalism, the use of Artificial Intelligence in reporting and editing may wipe out job opportunities for lower-end beat reporters who gather information and junior copy editors who re-write. As a result, two issues will arise out of the situation. On the one hand, there will be no recruitment of a less talented work force in newspaper establishments and, on the other, there will be no in-service space for new recruits to learn the tricks of the trade. Newspapers will recruit only highly talented professionals and, to be eligible for employment, job seekers will have to make their own arrangement at journalism schools or open-source learning courses and practise their skills on  social media platforms. A whole new eco-system may emerge in the media landscape.

 

 

Dr Pradeep Mahapatra is a media critic, teacher and consultant based in Odisha. He retired as the Head of the Department of Journalism & Mass Communication, Berhampur University.