Robot journalism: the end of human reporters? - The Listening Post (Feature)

Robot journalism: the end of human reporters? – The Listening Post (Feature)



this is one of those stories that since it's about automated journalism robot journalism really should write itself but the technology at least for television is not quite there yet it is however already working its way into some types of reporting whether you've realized it or not some of the data-driven journalism you've seen online may have been produced by robots computers financial reporting sports updates or earthquake alerts are the kinds of stories already being produced without human involvement one piece of software word Smith created by a company called automated insights produces thousands of earnings reports every quarter another video creation platform called watch it means that hundreds of news videos can be generated and uploaded to news websites and social media automatically the upside for news organizations is obvious more stories produced more quickly with fewer people to pay but what are the downsides for news reporting the listening-posts Meenakshi Ravi now on the implications of robot journalists breaking news about an earthquake in California a first look at Apple's quarterly earnings and a report from the field moments after a game of college baseball three stories you'd think at least three journalists would have been involved how about none software aided reporting has been dubbed robot journalism and it conjures up images of metal machine men roaming the newsroom in reality automated content is the work of just a piece of software that instead of helping you with word processing or virus protection helps you produce news reports these technologies use artificial intelligence and natural language processing to convert data into news stories they use algorithms that produce these story forms now these are not sophisticated stories they're stock quotes they may be sports scores so really stories that don't require a lot of thinking you can do a lot more a lot quicker it's a lot cheaper to do it and in essence you know in some cases you might even be making fewer mistakes if the data is qualitative it doesn't matter if it's a Senate subcommittee hearing or a meeting of parliament as long as that data is tracked and documented somewhere we can pull that data and write a story from it so if you're a mere human journalist the information for a new story would be facts quotes may be a personal angle the robots in the newsroom need much less and they don't run on coffee give them numbers a set of pre-programmed narrative parameters for instance an earthquake measuring between 2.0 and 4.9 on the Richter scale is classified as light whereas 6.02 9.0 varies from strong to great add a dictionary full of words and specialist terms think tensional stress main shock plate tectonics and your computer can spit out story after story on a topic don't expect verbal flourishes or any emotional insights the copy will be cut and dried this is algorithmic journalism and companies like narrative science in Chicago watch it in New York and tel-aviv and automated insights in North Carolina have spotted the potential in the market since July 2014 automated inside software wordsmith has been used in the newsrooms of The Associated Press AP to produce corporate earnings reports it's one of the first large-scale implementations of software produced journalism and according to the AP it's a Productivity bonanza each quarter we were spending about 20 to 25 percent of our time during earnings season writing these earnings reports and universally my reporters pretty much hated them really it wasn't the best use of our time and we found a way we could automate it and free up that time before wordsmith The Associated Press would write about around four hundred and forty companies per earning season after wordsmith they can write close to five thousand articles on these companies per season and that number is growing we have introduced any jobs because of this no one's lost their job but we put that time back into our journalists hands and by doing that they are doing the things they went to school to do and got into journalism to do which is to report and dig and develop sources technology can be both good or bad for an industry in terms of robots and jobs for journalists what you have to think about is we're gonna stay of change and a lot of the journals losing their jobs is because the industry is changing not because some algorithm was developed to write stories one concern with automated journalism is that it really privileges this idea of the types of data you can use this is usually quantitative data that's pretty easy to feed into the machine you can have it spit out of the story and so what we'll get is certain types of stories with easy to get data where other types of stories that are much more difficult to get won't be fed into the machine it won't become stories but for those who evangelize about automated content their main selling point isn't the type of stories their software can produce it's about the quantity they can churn out what they call the quantity multiplier effect according to automated insights its wordsmith program can juice 2,000 stories per second a standard eight-hour day would mean 57 million-plus stories whether there is that much news or not is a separate question watch it the online video creation platform whose clients are amongst others USA Today Germany's Spiegel Online and euronews in lyon france says you can source footage and edit video packages of about a minute to two minutes long in half an hour or so in a live newsroom like Al Jazeera's editing each minute of reported news takes on average about three times the watchit Edit time an hour and a half well you know with any kind of automation you're gonna get more quantity whether it's making cars or you know plowing fields for fruit Automation produces productivity you look at your Twitter feed and there's 400 stories going on one so I don't see it as a bad thing about putting more content into the world with automated journalism being able to produce so many stories what happens is that we shift from this idea of journalism telling us some very most important stories of the day to this idea of lots and lots of stories and they may only be available or interesting to a few people all the time and this really shifts what journalism does because journalism has been about establishing the news agenda and when we shift to automate journalism we have this much more pluralized environment where there are lots and lots of stories one of the main reasons that robot journalism will my opinion become popular is because it fits the economics of how online news works a minute of broadcast television it can be up to $2,000 right it's insanely expensive so we can help drive those costs down from a monetization perspective on average you make as much money selling or putting ads against a watch it clip as opposed to something that you have produced say using your traditional workflow so it makes sense commercially it makes sense on the revenue side but it also makes sense on the cost side given the advertised cost and time efficiencies of online video creation platforms we thought why not give the technology a try and it worked once we filmed the interviews written the script and worked out how to use watch it it took us just about three hours to put together this report you can watch that version on our Facebook page we decided not to use it on the show because what we learned in our experiment is that services like watch it or its competitor ribbits aren't aiming to serve TV networks with their exacting broadcasting standards they are geared primarily towards online outlets that are in a hurry and want to cash in on video that's on trend and can be produced within hours of a hashtag going viral so the quality isn't quite broadcast standard but that isn't the point this is about clicks and instant views and ultimately about money I don't think what we're gonna see is robots writing the front page of the New York Times for us so for time I think we'll see more robot journalism fit into regular journalism but what I think we'll see is a hybrid approach where robots are writing part of a story and humans are writing part of a story and what we'll see is journalism that speaks to the strengths of both so robots are very good at analyzing big sets of data finding some narratives and those but humans have a creative edge they can see different kinds of patterns if a robot reporter were writing the story it will tell you what has happened what has already changed what robots wouldn't be able to tell you is where this trend is going how far the integration between technology and in-depth journalism will go even if they could make that prediction maybe they'd rather not tell us you

One thought on “Robot journalism: the end of human reporters? – The Listening Post (Feature)

  1. That's Tony G. at 4:12 , he dont need to work, the robot computer do his job and he can go play poker full time !  Looser !!!

Leave a Reply

Your email address will not be published. Required fields are marked *