Tag Archives: AI

Google and PA put pressure on local journalists with ‘robot reporters’

Google yesterday announced an $800,000 project with the UK’s Press Association news agency, in which PA will develop ‘robot reporters’ that will write 30,000 articles a month for local newspapers.

That Orwellian piece of news got me thinking about local journalism and whether an algorithm, let loose in council databases, really can replace the reporter. Is it good for news journalism, is it good for reporters and what effect could it have on the role local papers play as a check and balance on local democracy?

My first contact with local journalism was when I did some work experience on a local weekly newspaper in the UK in the summer holiday of 1977 working with reporters and photographers. In many ways it was the age of the dinosaur; papers then were produced with huge production inefficiencies and the printing end of the operation was a tough closed shop with endemic restrictive practices and over-manning.  But the local paper, as it hit the street, did have the effect of keeping public authorities honest and often exposed them if they weren’t.

The newspaper office of the ‘70s had that heavy stationery-cigarette combination smell created by stacks of newspapers, copy pads, spikes, brimming ashtrays, wastepaper typewriterbaskets and rancid coffee cups. It was a world that hadn’t changed much since the 1920s – all reporters used shorthand, clattered away on manual typewriters and phoned-in copy (stories) from red phone boxes on street corners.

Fierce news editors demanded the facts and wielded a ‘blue pencil’, striking out comment or sloppy writing: words like ‘recently’, ‘quite’, ‘very’, and ‘big’ were outlawed. Did the story answer the where, when, what, why and how? If not – go away and do it again.

There were stacks of local and county council committee papers that reporters had to read and mark up to ensure a good story wasn’t missed – and they attended the meetings, known as ‘night jobs’, compensated for by an expense allowance of a quid or so for a pie and a pint for your evening meal.

Story tips came from indiscrete coppers, court ushers, friendly council officers and from blokes in pubs.  There was a weekly duty list for court reporting, ensuring all courts were covered or at least the charge sheet was collected and scanned for important cases.

Parcels of news stories marked ‘Urgent, useless if delayed’ were taken to the bus station and given to a driver who delivered it by bus to a neighbouring town. A messenger would meet the bus and deliver the parcel to the print works where a team of compositors would re-type the reporters’ stories into a type-setting machine.

Compositors cut, waxed and pasted paper columns of type onto a page layout, just like cutting an article out of a newspaper and gluing it into scrapbook. Printing plates were made from these layouts and fitted to an ancient Goss web-offset printing press, which thundered away producing the following morning’s paper.

Screen Shot 2017-07-07 at 10.07.05.png

Printing press from circa 1905

These inefficient ‘dinosaur’ newspapers and human news reporters frequently produced exclusive, investigative breaking news about ‘rotten boroughs’, often picked up by national papers, radio and TV news. The result was that a staff of half a dozen journalists often delivered a big helping of accountability for those in power locally.

And today, what of those newspapers in the summer of 2017 – if they haven’t already closed? On the positive side, the operations are far more efficient from a production point of view – email and shared drives are clearly more efficient than putting news stories on a bus…. There’s less paper and no ashtrays in the office and fewer reporters who rarely leave their desks.  That reduced staff have to fill paper and web editions, copy is transmitted in milliseconds, mobile phones that aid the job of reporters and photographers, citizen journalists and bloggers.

Those local papers that have survived can’t spare reporters to go to council meetings or read the committee papers and there’s no court reporting. There’s less time to dig around and tease out a really strong story, while press releases are being copied and pasted, unedited, into papers to fill more white space.  Great for the PR company writing press releases, but not great for an independent press. Ironically, the original Google/PA roboreporter announcement contained a typo in the first line (yes, really) – which is faithfully reproduced in many news stories on the topic. QED.

Google’s robot reporters project – in their words – ‘supports journalism’, but not journalists (my words).  In the final analysis, though, churning out 30,000 stories a month is less of a news operation and more intended to net more readers for the new content, alongside which Google can place its paid-for ads.  Follow the money…

The result of automation will be fewer reporters, fewer investigative, exclusive splashes on sharp practice in public organisations and those in authority are less likely to be held accountable. The local newspaper – paper and web – becomes an increasingly endangered species as their output is automated and quality diminished. The ‘food chain’ of stories from local news teams through to the national media has all but ceased.

And while the internet has brought information to billions of people, a team of Google-sponsored robot reporters isn’t going to be enough to provide sufficient checks and balances on democracy locally, or even nationally. Because even robots require someone to create real news stories and content in the first place before they can ‘scrape’ it and pass it on.

News isn’t so much being suppressed, it’s not even being created. As a result, there’s never been a better time to be a malevolent or incompetent politician, con man, dodgy businessman, criminal or corrupt public official.

Photo credits: Typewriter by Ian Livesey

Goss Printing press: By Unknown (Life time: Unknown.) [Public domain], via Wikimedia Commons

Making America Great Again: risks and opportunities for the new POTUS

I wouldn’t want to be Mr. Trump. Not for all the tea in China (although from Friday, I hear he will be mandating that ‘all that tea must be made in the US from now on’).

I wouldn’t want to be Mr. Trump, because I’ve just read the 2017 Global Risks Report, released last week by the World Economic Forum.  Managing the multitude and magnitude of risks in the report will need thought, care, attention, leadership and policy responses from him as head of the most powerful nation on earth.  As an aside, I use the reports to provide strategic background to crisis and resilience scenarios that I develop for exercises and workshops for clients. In the past, they’ve also provided excellent context for Enterprise Risk Management (ERM) work.

So what are the risks that may be keeping POTUS Donald Trump up at night?

The top four strongest risk trends in the 2017 report are ‘rising inequality of income and wealth’, ‘polarisation of society, especially among older generations’, ‘climate’ and ‘cyber dependency’. Risk trends pull together a basket of specific risks and the higher the number and importance of those contributory risks, the stronger the trend.

Looking at the impact of specific risks (and there’s certainly reference to a ‘water crisis’, but that’s only risk number three, in terms of impact). Top and second place go to ‘weapons of mass destruction’ and ‘extreme weather events’, fourth is ‘major natural disasters’ and finally, ‘failure of climate change and mitigation’.


Watson and the Shark, photo of detail of original painting by John Singleton Copley. Photo: Adam Roscoe

In terms of likelihood, the top risk is ‘extreme weather’, second is ‘large-scale involuntary migration’, then ‘major natural disasters’ followed by ‘large-scale terrorist attacks’ and finally ‘massive incident of data fraud or theft’.

It seems to me that many of the environmental and social risks, could be rooted in, affected by or the result of climate change.  Extreme weather, water scarcity and some major natural disasters and some triggers for migration are all connected with climate change, so it is concerning that the new POTUS intends to roll back on the Paris Climate agreement.

When combining most likely/biggest impact risks, then ‘interstate conflict’ and ‘unemployment or underemployment’ are consistent features in BOTH the 2016 report and 2017 report. These are both big enough issues to sit firmly on Mr. Trump’s desk, where a predecessor displayed a plaque stating ‘The buck stops here’.

Interstate conflicts and proxy wars are already flaring in the Middle East and are drawing in US and other western (NATO) forces, whether Mr. Trump thinks NATO is up to the job or not. 

‘Making America Great Again’ is the theme of this presidency and there seems to be a risk is that the ‘greatness’ could be made at the expense of relations with other countries like China and Mexico.

25 million jobs a year

A previous Chinese president in the 2000’s said what kept him up at night was the ability of the economy to grow at a sufficient rate that it created 25 million jobs a year. Economic growth and the ability to create those jobs affects the government’s ability to maintain order and close inequality gaps in the country. But as China moves from a so-called ‘cheap labour arbitrage country’ to a high-tech, high automation exporter, they will need to cut a win/win deal with the US that saves face for both parties and doesn’t precipitate a mutually damaging trade war.

86% of all US jobs lost in the decade from 1997 were lost to productivity, not trade

And it’s a war that really needn’t be fought.  According to the economists Michael Hicks and Srikant Devaraj, 86% of manufacturing job losses in the United States between 1997 and 2007 were the result of rising productivity [a part of which was achieved through automation], compared to less than 14% lost because of trade (see page 20 of the 2017 Risk Report). Making America Great Again need not to be a regressive step into Luddite protectionism, even if AI, robotics and biotech will require regulation.

If it is to remain competitive, the US will also need to actively manage the inevitable disruptive effects of automation that the Fourth Industrial Revolution brings.  The promise of on-shoring or re-on-shoring jobs to the US, allegedly at the expense of lower cost economies is a delicate juggling act – if not purely smoke and mirrors.  Create too many blue- and white collar jobs in the US that can be done by automation and robotics and the economy risks being uncompetitive.

A partial response to this challenge could be regular retraining of older workers in industry and commerce – so they continue to contribute to taxes, rather than become a drain on the social security system – should become the standard operating practice for all companies and could be supported by tax breaks.  High value-added work will always be needed and will be well-paid, but people will need to be trained on an ongoing basis to do it.

Win, win, win…

Part of the solution should include investment in low carbon technologies, including solar, wind, hydro and electric mobility, all of which are getting huge attention and investment in China.  The US could also improve its energy resilience by making additional investment in alternative sources of power, aside from fossil fuels, and retraining older workers to install, service and run new energy projects.  Low carbon tech could create a ‘win, win, win’ for the economy, jobs and the climate and in doing so positively mitigate some of the key risks highlighted by WEF.

Regardless of whether the future is low carbon or high carbon, you just have to look outside the top 10 headline risks to find some consequences of getting the response to these key risks wrong or too late or weak.  The ‘failure of a financial mechanism or institution/fiscal crisis’, ‘failure of national governance’ and ‘profound social instability’ all vie for a position in the top 10 risks and could jump to the top spot.

But who knows what will happen, when and how? After all – in 2016, few pundits or risk reports predicted that Brexit would become a reality, or that Donald Trump would be elected President, or that 5,000-1 outsiders Leicester City (a UK football club, round balls) would win the Premier League title.