Revolutionary Measures

The news – written by robots

The march of technology has radically changed many jobs. Factory work has become increasingly automated and roles that involved processing documents have been swept away. And as the cost of processing power falls, artificial intelligence improves dramatically and more and more information is available online, machines are becoming cleverer. From delivering online learning to scanning legal documents for relevant information or automated trading of shares, computer-based algorithms are increasingly capable of replacing people in more traditionally white collar roles.

Partial front page of the Los Angeles Times fo...

Partial front page of the Los Angeles Times for Monday, April 24, 1922, displaying coverage of a Ku Klux Klan raid in an L.A. suburb (Photo credit: Wikipedia)

Judging by recent stories, the next profession under attack is professional journalism. Already hit hard by the free model of the internet and the rise of citizen reporting, journalists now have to fight off robots with their eyes on their jobs. AP has just announced that it will use software from Automated Insights to produce 4,440 robot-written corporate earnings reports every quarter. The company argues that by letting the software write basic stories that essentially cover the financial details of the earnings announcement it frees up human reporters to write more detailed analysis pieces – and also ensures it can cover more companies without expanding its staff.

AP is not alone. An increasing number of news outlets are using software to write up reports of matches in minor sports which aren’t popular enough to justify the attention of a ‘real’ reporter. And in California, the LA Times uses a program to analyse data from the US Geological Survey to provide a first report on earthquakes. It is also using separate software to analyse incoming lists of arrests from the police to flag those that look newsworthy to reporters. This essentially works by thinking like a journalist and looking for potential signs of interest, such as particular names or occupations and high bail amounts that can then be followed up.

So, should journalists (and by extension, other writers such as PR people), be worried? There may be a lot of hand wringing about these developments, but I think there are three reasons that the human hack will survive, and even thrive.

1          Robot journalism is all about the facts
At present software is very good at searching for information, collating it and presenting it in a way that can be easily read. If you look at the AP or LA Times pieces they are never going to win any prizes for journalism, as they are basic stories that journalists would knock out because they had to, without really getting out of second gear. Software can do it much faster, freeing up their time for more interesting pieces.

2          Opinion and context is key to retain readers
We are bombarded by facts. What we crave are journalists who can put the facts in context, create a logical narrative, and most importantly add experience and opinion. While, technically, software could try and mimic this by analysing the complete works of Caitlin Moran and regurgitating it, it can’t get across personality in the same way. So real opinion will always beat computer journalism that stitches together opposing quotes when it comes to engaging the eyeballs of readers.

3          Investigative journalism is alive and well
Reporters continue to work tirelessly to expose scandals, often with very little official data to work with. Whether it is uncovering irregularities with the expenses of MPs or child abuse in care homes, a large number of nationally and internationally important stories have only been published because of long term research and work by journalists or newspapers. Computers may well have been used to help in analysis and getting to the truth, but they did not lead the investigation.

So, before reporters worry too much about R2D2 stealing their jobs, it is worth understanding what readers actually value in an article. Yes, they want the facts, but more to the point they want opinions, colour and context, especially if it is personal and random – and at the moment, this is beyond the scope of software.

September 17, 2014 Posted by | Creative, Marketing, PR | , , , , , , , , , , | 3 Comments

Robots vs serendipity

A swarm of robots in the Open-source micro-rob...

By bringing the world together the internet opens up pretty much unlimited possibilities. You can discover completely new topics and interests, communicate with people across the globe and access a myriad of content that was previously unavailable.

More and more of what we read, watch and listen to comes via the internet – and this is only going to increase as previously analogue services such as TV go digital. On one hand this widens choice, but how do we navigate and find things we are interested in? And more to the point, is just watching what we’re interested in necessarily a good thing?

Showing my age, when I was growing up I had a choice of three TV channels (I remember the excitement of the Channel 4 launch), and video recorders were in their infancy. So you watched what was on – or switched the TV off and did something (less boring) instead. That meant there was a greater chance of stumbling upon a programme or subject that you wouldn’t have chosen to watch but actually widened your knowledge. I’m not saying the 1970s was a golden age of TV but you were likely to see a broad range of subjects in your daily viewing.

Now we have a plethora of channels and there’s always that nagging fear that there’s something better on the other side. Navigating this maze is difficult – how do you choose what to watch when there are thousands of alternatives? The way I see it there are essentially three ways of making a choice:

Robots – like Amazon Recommendations your TV/Set Top Box or PC sees what you have watched and enjoyed in the past and comes up with more of the same. However this essentially narrows your viewpoint – you’ll potentially end up watching programmes very similar to those you’ve seen before. The same goes for search – after all, you’ve got to know what you’re looking for before you type something into Google.

Friends – personal recommendations work, provided they come from people you trust. And given pretty much every programme is available on catch-up TV, you can view what your friends on like after the fact. And social media provides a quick way of gathering recommendations. Better than robots, but still likely to keep your watching within a relatively constricted area – after all we’re governed by a herd mind.

Editorial choice – what does the newspaper/TV guide say is good and worth watching? TV previews tend to cover a wide range of subjects so can highlight programmes that you wouldn’t normally watch. All good, but even with glowing reviews some programmes may not sound like your cup of tea and you won’t watch them.

Ironically the digital world can give us too much choice and make us flee back into watching a tiny fraction of its range. So, what’s the solution – or does there even need to be one? I’d argue that we should rely less on robots or even our friends and trust to serendipity – switching on the TV to a random channel and giving the programme 10 minutes to make an impression. Yes, it might mean seeing some duds but it also gives the chance of finding a new area that will change your life. Now all we need is an app to help us do that……………

September 11, 2013 Posted by | Creative, Marketing, Social Media | , , , , , | Leave a comment